EPA-600/2-80-111
August 1980
FATE AND EFFECTS OF PARTICULATES
DISCHARGED BY COMBINED SEWERS
AND STORM DRAINS
by
Richard D. Tomlinson, Brian N. Bebee,
Andrew A. Heyward, Sydney G. Munger and Robert G. Swartz
Water Quality Division
Municipality of Metropolitan Seattle
Seattle, Washington 98104
Steven Lazoff and Dimitris E. Spyridakis
Department of Civil Engineering
University of Washington
Seattle, Washington 98195
Michael F. Shepard, Ronald M. Thorn,
Kenneth K. Chew and Richard R. Whitney
College of Fisheries
University of Washington
Seattle, Washington 98195
Grant No. R805602010
Project Officer
John N. English
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
This study was conducted
in cooperation with
Municipality of Metropolitan Seattle
Seattle, Washington 98104
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
-------
FOREWORD
The Environmental Protection Agency was created because of
increasing public and governmental concern about the dangers of
pollution to the health and welfare of the American people
Noxious airf foul water, and spoiled land are tragic testimony
to the deterioration of our natural environment. The
complexity of that environment and the interplay between its
components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its
impact, and searching for solutions. The Municipal Environ-
mental Research Laboratory develops new and improved technology
and systems for the prevention, treatment, and management of
wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize
the adverse economic, social, health, and aesthetic effects of
pollution. This publication is one of the products of that
research; a most vital communications link between the
researcher and the user community.
This report provides the details of an evaluation of the
distribution and biological impacts of particulate materials in
combined sewer and storm drain discharges in the Seattle,
Washington region, and presents the extent of the urban runoff
problem in terms of statistics and observed and anticipated
impacts on water quality.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
111
-------
PREFACE
The quality of semi-natural waters in urban areas is directly
related to the amount of local human activity. Efforts to
protect urban waters for recreational uses require a knowledge
of discharge sources of pollution, pollutant loadings and the
effects of these pollutants on the receiving water, the benthic
communities and the public health of the local populace.
The Municipality of Metropolitan Seattle Water Quality
Laboratory contributes to this knowledge through environmental
programs designed to:
1) monitor surface waters within King County for
biological and chemical parameters,
2) assess the effects of stormwater runoff and treat-
ment plant discharges on water quality,
3) monitor recreational waters and shellfish for
microorganisms of sanitary significance, and
4) investigate the impacts of industrial waste
discharges on water quality.
This report details an investigation of the fate and ecological
effects of particulates in stormwater runoff and combined sewer
discharges entering Lake Washington and Puget Sound. The
potential public health risk related to enteric viruses
associated with such particulates is also addressed.
IV
-------
ABSTRACT
The distribution and biological impacts of discharged
participates were evaluated for selected combined sewer outfalls
(CSOs) and storm drains (SDs) in the Seattle, Washington region.
Intensive studies were done on one CSO and one SD discharging
into Lake Washington and having residential drainage basins of
comparable size and incident rainfall. The mean storm discharge
concentrations of suspended solids and most particulate con-
taminants were greater for the CSO than for the SD. However,
due to the SB's greater discharge volume (runoff + baseflow)r
its annual particulate discharge load was greater for Cu, Pb,
organic carbon and chlorinated hydrocarbons. Human enteric
viruses were also detected in the CSO discharge, but were not
found in storm drainage or in any near-outfall sediments.
Prevailing circulation patterns implied a negative sanitary
impact of CSO discharges on adjacent recreational areas.
Light transmission measurements of discharge plumes identified
extensive additional inputs from neighboring CSOs, SDs and
construction sites. Particulate distributions were influenced
by various dispersion processes, including water density
layering, near-bottom offshore streaming and longshore advection.
Oligochaete numbers and biomass were found to be substantially
enhanced near two CSOs and two SDs studied in Lake Washington.
Near-outfall depletion of other taxons at both CSOs and SDs
also provided evidence of effluent toxicity and/or substrate
alterations. Impacts of discharges on the freshwater benthos
raised concern relative to the feeding success of sportfish.
Discharge concentrations of selected contaminants were also
assessed for a marine CSO. On the basis of six different
biological indicators sensitive to water and sediment quality,
the nearshore area within 150 m of that outfall was
characterized as polluted.
This report was submitted in fulfillment of Grant No.
R805602010 by the Municipality of Metropolitan Seattle under
the partial sponsorship of the U.S. Environmental Protection
Agency. This report covers a period from October 27, 1977 to
October 23, 1979, and work was completed as of July 31, 1979.
v
-------
-------
CONTENTS
Foreword
Preface ......... f ....... \ !!!!!!!! iv
Abstract .................... !!!!!! v
Figures ..................
Tables
Abbreviations and Symbols
Acknowledgement
3
.
xvi 11
1. Introduction .................... -,
Project origins ................ j_
Project objectives .............. 5
2. Summary and Conclusions .............. 7
Freshwater studies ............. [ 7
Marine studies . ............... n
3. Recommendations ................. [ 13
4. Method of Study ....... ...... ! ! ! ! ! 15
Selection of sampling sites .......... 15
Preliminary freshwater studies ...... 15
Intensive freshwater studies ....... 17
Marine studies .............. 17
Description of sampling sites ......... 17
Preliminary freshwater studies ...... 17
Intensive freshwater studies ....... 17
Marine studies. ... .......... 18
Field sampling methods ............ 21
Combined sewer and storm drain discharges 21
Settling particulates .......... 23
Sediments ................ 25
Benthic biota .............. ~>c
__. •••••• ^o
Viruses ................. 28
Analytical techniques ............. 29
Combined sewer and storm drain discharges 29
Settling particulates .......... 30
Sediments ................ 31
Benthic biota ......... ..... 32
Viruses . . ............... 34
Data reduction ................ 36
Combined sewer and storm drain discharges 35
Settling particulates and sediments ... 37
Benthic biota .............. 37
vii
-------
CONTENTS
Page
5. Results 40
Freshwater studies 40
Preliminary studies 40
Intensive studies 45
Marine studies ... 99
Discharge monitoring 99
Turbidity distributions of discharge plumes 101
Sediment particle size distributions. . . .108
Benthic biota m
6. Discussion 133
Effects of discharges on littoral food chains. .133
Freshwater environment „ 133
Marine environment „ 2.37
Extrapolation of present findings to other
outfall sites 139
Extent of nearshore waters affected by
discharge particulates , 140
Freshwater environment , 140
Marine environment
Distribution and effects of discharge-borne
viruses 142
References -^5
Appendix 154
Glossary -, cq
Vlll
-------
FIGURES
Page
1 Locations of major water bodies and sewage
treatment plants in the Seattle area ......... 2
2 Locations of sampling sites .............. 16
3 Principal set of intensive sampling sites in
Lake Washington ................... 19
4 Secondary set of intensive sampling sites in
Lake Washington ................... 20
5 Details of sampling transects around the Denny
Way Regulator Outfall, 1978 ............. 22
6 Details of sediment trap configuration ......... 24
7 Configurations of the sediment trap arrays moored
at the principal sampling sites in Lake Washington. . 25
8 Relative enrichment of the nearshore surface
sediments of Lake Washington with normalized
categories of parameters ............... 41
9 Summary of rainfall and overflow response at
Combined Sewer Outfall 023 and Storm Drain 7
from March 1978 to February 1979 ........... 52
10 Aerial perspectives of contours of percent light
transmission at Combined Sewer Outfall 023 —
1533-1645 hrs., 3/7/78 ................ 57
11 Longitudinal sections of contours of percent light
transmission at Combined Sewer Outfall 023 —
1533-1645 hrs., 3/7/78 ................ 58
12 Longitudinal section through the Combined Sewer
023 outfall, showing contours of percent light
transmission — 0856-0954 hrs., 4/4/78 ......... 59
IX
-------
FIGURES
Number Page
13 Aerial perspectives of contours of percent light
transmission at Combined Sewer Outfall 023—
1330-2330 hrs., 2/6/79 60
14 Transverse section through a point 34 m S of
Combined Sewer Outfall 023, showing contours of
percent light transmission—1330-2330 hrs.,r 2/6/79. . 61
15 Aerial perspectives of contours of percent light
transmission at Storm Drain 7—1915-2014 hrs.,
2/6/78 62
16 Aerial perspectives of contours of percent light
transmission at Storm Drain 7—2244-2348 hrs..
2/6/79 63
17 Transverse sections through a point 61 m N of Storm
Drain 7f showing contours of percent light trans-
mission on 2/6/79 , 64
18 Transverse section through a point 344 m N of
Combined Sewer Outfall 023, showing contours of
percent light transmission—2244-2348 hrs., 2/6/79. . 65
19 Aerial perspectives of contours of percent light
transmission at Control Site 3—0506-0549 hrs.,
2/12/79 66
20 Transverse sections showing contours of percent
light transmission at Control Site 3—0506-0549
hrs., 2/12/79 67
21 Aerial perspectives of contours of percent light
transmission at Combined Sewer Outfall 044—
2210-2350 hrs., 2/11/79 68
22 Aerial perspectives of contours of percent light
transmission at Storm Drain 19—0020-0250 hrs.,
2/12/79 69
x
-------
FIGURES
Number
23 Summary of results of particle size distribution
analyses for Combined Sewer Outfall 023 and
Storm Drain 7
24 Dry weight distributions of lead, zinc, copper and
total carbon in the surface centimeter of sedi-
ments collected near Combined Sewer Outfall 023 ... 73
25 Dry weight distributions of lead, zinc, copper and
total carbon in the surface centimeter of
sediments collected near Storm Drain 7. . . 79
26 Dry weight distributions of lead, zinc, copper and
total carbon in the surface centimeter of sedi-
ments collected at Control Site 3 80
27 Net-change in numbers of-organisms and biomass per
core, as a function of distance from the combined
sewer and storm drain outfalls in February and
September, 1978 93
28 Summary of rainfall and overflow response at the
Denny Way Regulator outfall from March 1978 to
February 1979 1Q2
29 Aerial perspectives of contours of percent light
transmission around the overflow outfall of the
Denny Way Regulator—0517-0746 hrs., 4/16/78 104
30 Aerial perspectives of contours of percent light
transmission around the overflow outfall of the
Denny Way Regulator—2050-2243 hrs., 2/24/79 106
31 Transverse sections of contours of percent light
transmission around the overflow outfall of the
Denny Way Regulator—2050-2243 hrs., 2/24/79 109
32 Summary of particle size distribution analyses for
the Denny Way Regulator 110
xi
-------
FIGURES
Number pag€
33 Variation in the condition indices of mussels and
oysters kept in pots at four sites near the Denny
Way Regulator outfall during 1978 . . . 112
34 Size data for Enteromorpha collected near the Denny
Way Regulator outfall during April and August, 1978 .115
35 Density of infaunal individuals and taxa found in the
subtidal samples collected near the Denny Way
Regulator outfall during April and August, 1978 . .
36 Species-richness curves for molluscs collected near
the Denny Way Regulator outfall during April and
August, 1978
120
37 Species-richness curves for polychaetes collected
near the Denny Way Regulator outfall during April
and August, 1978 ............. 1
38 Dendrogram of subtidal infauna samples collected
near the Denny Way Regulator outfall during April
and August, 1978 ............. '..... . 122
39 Positions of subgroups of sites from cluster analysis
of samples of subtidal infauna collected near the
Denny Way Regulator outfall during April arid
August, 1978 ............
123
40 Proportions of the total number of annelids,
arthropods and molluscs collected near the Denny
Way Regulator outfall during April and August, 1978 .125
41 Proportions of total numbers of polychaetes within
each of three feeding-type categories, for samples
collected near the Denny Way Regulator outfall
during April and August, 1978 127
42 Species-richness curves for periphyton samples
collected near the Denny Way Regulator outfall
during April 1978 128
xn
-------
FIGURES
Number Pag<
43 Position on canonical variables of periphyton samples
collected near the Denny Way Regulator outfall
during April 1978 128
44 Species-richness curves for macroalgae collected near
the Denny Way Regulator outfall during April and
August, 1978 13Q
45 Position on canonical variables of samples of boulder
wall taxa collected near the Denny Way Regulator
outfall, 1978 131
xm
-------
TABLES
Number Page
1 Correlation Matrix for CSO Sediments 44
2 Correlation Matrix for Storm Drain Sediments 44
3 Correlation Matrix for Control Sediments 44
4 Summary of Rainstorms Monitored for Quantity and
Quality of Discharges into Lake Washington 45
5 Summary of Estimated Total Pollutant Loads and
Fraction of Particulates in Storm Discharges
Monitored at Combined Sewer Outfall 023 and
Storm Drain 7 , 47
6 Interstation Comparison for Storm Drain 7 arid
Combined Sewer Outfall 023 of Mean Discharge
Concentrations and Pollutant Loadings 48
7 Summary of Virus Analyses of Discharges from
Combined Sewer Outfall 023 and Storm Drain 7 .... 50
8 Estimated Total Annual Masses of Selected
Constituents in Discharges from Combined Sewer
Outfall 023 and Storm Drain 7 53
9 Summary of Field Trips for Measuring Turbidity
Distributions of Discharge Plumes 55
10 Correlation Coefficients for Nearshore
Sedimentation Rates vs. Volume of Discharge
from Combined Sewer Outfall 023 and
Storm Drain 7 70
11 Summary of Results of Two-Way Analysis of Variance
for Sediment Trap Collections from Combined Sewer
Outfall 023, Storm Drain 7 and Control Site 3. ... 71
12 Mean Sedimentation Rates and Dry Weight
Concentrations of Selected Constituents Analyzed
in Sediment Trap Solids Collected at Combined
Sewer Outfall 023, Storm Drain 7 and Control
Site 3 71
xiv
-------
TABLES
Number
Page
****«B_JKHM
13 Summary of Virus Analyses of Receiving Waters near
Combined Sewer Outfall 023 and Storm Drain 7 .... 74
14 Significant Differences Between Mean Station
Concentrations for Selected Parameters Analyzed
in the Top 0.5 Centimeter of Sediment Cores
Collected at Combined Sewer Outfall 023,
Storm Drain 7 and Control Site 3 77
15 Mean Concentrations and Concentration Ratios
for Selected Metals in Particulates Sampled
at Combined Sewer Outfall 023, Storm Drain 7
and Control Site 3 01
•••••••• oJL
16 Upper Confidence Limits Expressed as Percentages
of the Means for 2, 4, 6, 8 and 10 Cores per
Sampling Location for Chironomid, Oligochaete
and Copepod Counts 85
17 Coefficients of Determination for Linear and
Multiple Regressions Using Total Organism
Counts in February as the Dependent Variable .... 89
18 Coefficients of Determination for Linear and
Multiple Regressions Using Total Organism
Counts in September as the Dependent Variable. ... 90
19 Coefficients of Determination for Linear and
Multiple Regressions Using Total Organism
Weight in February as the Dependent Variable .... 91
20 Coefficients of Determination for Linear and
Multiple Regressions Using Total Organism
Weight in September as the Dependent Variable. ... 92
21 Summary of Rainstorms Monitored for Quantity
and Quality of Discharges into Puget Sound
from the Denny Way Regulator Outfall 99
22 Statistical Summary of Estimated Pollutant
Loads and Concentrations in Storm Discharges
Monitored at the Denny Way Regulator 100
xv
-------
TABLES
Number Page
23 Estimated Mass of Selected Constituents in
Discharges from the Denny Way Regulator Outfall . . .'101
24 Rainfall, Overflow Volumes and Tidal Variations
for Overflow Turbidity Distributions Monitored
Around the Denny Way Regulator Outfall 103
25 Tissue Burden of Heavy Metals in Mussels Kept in
Pots at Four Sites near the Denny Way Regulator
Outfall 114
26 Tissue Burden of Heavy Metals in Oysters Kept in
Pots at Four Sites near the Denny Way Regulator
Outfall 114
27 Characteristics of Subtidal Sediments Sampled
near the Denny Way Regulator 116
28 Characteristics and Annelid Densities of Intertidal
Soft Sediments Sampled near the Denny Way
Regulator 129
xvi
-------
LIST OF ABBREVIATIONS AND SYMBOLS
a
Al
BGM
C
C1HC
CI
CSO
Cu
g
Hg
H2S
kg
km
m
mE
MG
mg
MLLW
mN
mS
mW
NA
ND
O&G
P
Pb
PCB
ppb
ppm
r
a, Sx
SD
TOC
TPO4-P
X
Zn
volume attenuation coefficient
aluminum
Buffalo Green Monkey cells (for virus analysis)
carbon
chlorinated hydrocarbons
Condition Index (for shellfish)
combined sewer outfall
Copper
grams
Mercury
hydrogen sulfide
kilograms
kilometers
meters
meters east of line through outfall
million gallons
milligrams
mean low low water
meters north of line through outfall
meters south of line through outfall
meters west of line through outfall
not applicable
not determined
oils and greases
phosphorus
lead
polychlorinated biphenyls
parts per billion
parts per million
correlation coefficient
standard deviation
storm drain
total organic carbon
total phosphate phosphorus
arithmetic mean
zinc
xvii
-------
ACKNOWLEDGMENTS
The authors gratefully acknowledge the aid of the following
individuals and organizations in providing outfall locations
and engineering specifications for the project's base map:
Mr. Bruce Jones and Ms. Pat Flynn of the City of Seattle
Department of Engineering, Mr. Miles Fuller of the City of
Mercer Island Utilities Department, Mr. Fred French of the
City of Kirkland Public Services Department, Ms. Pam Bissonnette
and Mr. Hector Gyre of the City of Bellevue Department of
Public Works, the King County Department of Public Works,
the City of Renton Engineering Division and the N.E. Lake
Washington Sewer District.
We also wish to thank Mr. Harry Truitt of the Light House
Dive Shop in Lake City and Mr. Don Bloye of the Silent World
Dive Shop in Bellevue for supplying some of the equipment
and expertise that aided in the success of the diving portion
of our sampling program. The divers themselves, Mr. Steve
Aubert, Mr. Paul Farley, Mr. Dan Sturgill and Mr. Bob Dutton,
all of the Metro Water Quality Division, have our considerable
gratitude for the extensive effort they expended under
difficult circumstances to provide the best samples possible.
An expression of our appreciation is also offered to
Dr. Douglas G. Chapman, Dean of the University of Washington
College of Fisheries, for his suggestions in the formulation
of the statistical analyses for the freshwater biota, and to
Mr. John Buffo of the Metro Computer Services Division for
the time and skill he applied to the turbidity contour
analyses.
XVlll
-------
SECTION 1
INTRODUCTION
PROJECT ORIGINS AND RESEARCH FOUNDATION
The City of Seattle is surrounded by a multiplicity of both
natural and modified water bodies (Figure 1) . It is bounded
on the west by the central basin of Puget Sound, part of a
great complex of channels and inlets having a maximum depth
of 267 m just three kilometers northwest of West Point. To
the east of the city lies Lake Washington, which extends 29 km
on a north-south axis. Although the natural outlet for the
i^f ^ original?-y at ^s southern end, the flow patterns
*nH Q«^er! ^ ^e Garly 1900s with the construction of a canal
and saltwater locks between Lake Washington and Puget Sound;
andSnrov?^ay became ^he principal outlet for Lake Washington
c^.?r ! a navi9ai>le link through Lake Union to Puget Sound.
Seattle makes heavy commercial and recreational use of both
±jus east-west canal and its primary north-south watercourse,
the Duwamish River, whose estuary on Elliott Bay hosts
extensive saltwater port facilities.
Seattle's concern for the maintenance of good quality for its
bounty of waters led to the formation in 1958 of ^Municipality
wfthe^P ^ SKanttle (Metro>' a P^lic corporation vested Y
with the responsibility for areawide sewage disposal.
Previously, both treated wastes and raw sewage had been
discharged to fresh and marine waters throughout the region.
Metro, however, consolidated the facilities of some 30 agencies
and provided for treatment and discharge of all sanitary wastes
at four treatment plants (POTWs) on Puget Sound and one on
most oTth (F±gUr 1} ThlS aPProach s^ved to eliminate
noK impacts of haphazard waste discharge - most
notably the dense algal blooms consistently appearing in Lake
Washington. But the peak flow volumes associated with rain
nn t0° ^reau- f°r GVen the revised treatment system to
accommodate and combined sewer overflows continue to occur
abaT^nf nf City Under SUCh conditi°ns. The assessment and
abatement of any consequent environmental impacts has been
given high priority by Metro as part of its responsibility for
maintaining good regional water quality
-------
RICHMOND BEACH
POTW
BOTHELL
Figure 1. Ix^cations of major water bodies and sewage treatment plants
in the Seattle area.
-------
The extent of the present combined sewer overflow problem in
Seattle can be explained in terms of statistics, locations
and observed and anticipated impacts on water quality.
Statistically speaking, there are about 160 overflow locations
in the Seattle area, of which 30 are under the jurisdiction of
Metro and the remainder are the responsibility of the City of
Seattle. A brief, intense rainstorm can gorge city collector
sewers, causing basement backups and discharges to" Lake
Washington. In contrast, a milder storm of longer duration
will tend to flow more smoothly to the larger trunk sewers
where a buildup of flow beyond treatment and interceptor trunk
capacity will lead to programmed overflows to the lower
Duwamish River, Elliott Bay and Lake Union. These longer
storms, typical of the winter season, may result in less
than 50 percent of municipal wastewater actually reaching the
West Point Treatment Plant during the actual storm period.
Operation of Metro's Computer Augmented Treatment and Disposal
(CATAD) System, which provides in-line storage in some of
the larger sewers, permits some flexibility as to discharge
location, with highest priority given to protection of
freshwater bodies (Lake Washington, Lake Union) where possible
Frequency of overflow at present is estimated to average about
40 occurrences per outfall per year, of which perhaps five
or six are summer occurrences.
Geographically, combined sewer and storm drain outfalls are
scattered throughout the area, with discharges of varying
volumes to all of the major water bodies around the City of
Seattle. Some of these discharges are over shellfish beds,
or as in Lake Washington, near public bathing beaches or
spawning areas for anadromous fish; others are in the harbor
area where water contact sports or potential public health
hazards are minimal.
In wet weather, sewer flows comprise varying mixtures of
street and rooftop drainage, infiltration/inflow and raw
sewage. This untreated wastewater can be generally characterized
as a dilute sewage with typical BOD values of 100 to 200
mg/1. Suspended solids values also vary widely: 50-525
mg/1 for combined wastewaster (Metro, 1976) and 54-112 mg/1
for storm runoff (Farris et al., 1974) (Metro data show 36-
96 mg/1 and 2-24 mg/1 respectively for primary and secondary
treatment plant effluent).
Studies done by Dalseg and Leiser (1970) of the quality of
wastewater entering Lake Washington from two combined sewer
overflows and two storm drains yielded solids concentrations
five to ten times higher for the combined wastewater than
for the storm drainage. However, physical separation of one
of the combined systems had resulted in an estimated
fivefold annual increase of suspended solids discharge at that
location due to the increased number of storm drain overflows.
-------
Further calculations indicated that each inch of rainfall
resulted in 1330 lb of suspended solids loading from one of
the storm outfalls; the comparable figure for one of the
combined outfalls was 650 lb. The Seattle area has an annual
average rainfall of about 100 cm.
During August of 1976 Metro's Water Quality Division monitored
four combined sewer outfalls in the Seattle region. The
results of the study (Tomlinson et al., 1976) indicated
potential problems from discharged particulates and their
associated toxins. First-flush suspended solids concentrations
at three of the stations reached levels exceeding 500 mg/1.
In addition, deposits of black, oily sediments were found at
the ends of three lake and river outfalls. The highest
concentrations of heavy metals were found at or near the
outfalls. High concentrations of pesticides (up to 1.95 ppm
dry wt.) were also found in the sediments near the single
outfall monitored in Lake Washington.
It appears that much of the particulate fraction of wastewater
overflows may be deposited near the outfalls. Summer season
dye studies indicate that the nearshore surface circulation
at two lake stations is comparatively sluggish, providing a
maximum dilution of 2.7:1 (Lake Washington) and 75:1 (Lake
Union) during the first three hours following release. Such
conditions might be expected to favor the near-shore settling
of much of the sediment injected into the environment by
combined sewer outfalls and storm drains.
Very little information is available concerning the effects
of combined sewer overflows and storm drainage on freshwater
benthic communities. White (1975) has discussed the potential
influence of fluctuations in organic particulates on the
community distributions of many of the benthic organisms;
population increases of these organisms were generally
related to increases of organic material. On the other
hand, alterations in particle size distributions and toxicity
increases in the community substrate as a result of urban runoff
might be expected to be detrimental to such populations. The
importance of many of these taxonomic assemblages as primary
or secondary elements of the freshwater food chain has been
documented by Tomlinson et al. (1977) .
Further, there is a strong need for information relative to
the effects of discharged particulates and their associated
contaminants on the benthic biota around marine sewer outfalls
and storm drains. Armstrong et al. (1978) conducted a pre-
liminary investigation on the effects of discharges from one
of Metro's principal marine CSOs on the adjacent hard- and
soft-bottom benthic biota; based on their analyses, they
were able to delineate a zone of impact attributable to the
CSO. These marine investigations were limited to the Spring,
-------
oneu^
due to variuons in discharge characteristics • J^xtxonal
studies would need to consider the impacts of the seasonally
of discharges on seasonal biological processes.
Beyond the concern for aquatic biota, there is considerable
apprehension on the part of public health workers over the
Presence of infections viruses in CSOs . A.l«ge port £» °f
these pathogens has been shown to be associated with the
'
to the health of recreationists.
wii-h the recent advent of quantitative virus detection
techniques if has become apparent that the concentrations
of viruses 'of human origin in sewage usua lly vary seasonally ,
reaching a peak during the late Summer .
variation reflects the higher infection rates that
during the warmer months of the year (Grabow, 1968) .
their9extensive qualitative potential for h^an infection
(Fenner and White, 1976) a quantification of the vxruses
measures.
Therefore, in view of their potential impact on the health
of both the endemic aquatic biota and human Creationists
there is an urgent need for more data on the loading, distra.
bution and ef flats of combined sewer and storm drain Articu-
lates and their associated contaminants entering Settle s
nearshore waters. The present project was designed to provide
this type of information.
PROJECT OBJECTIVES
The principal project objectives were as follows:
1 TO determine the distribution patterns and fate of
' -
aea To determine seasonal differences and to corre-
late quantitative in-situ observations with suspended
solids loading factors and current patterns.
-------
2. To determine the effects of the settled particulates
and associated contaminants on the population distri-
butions of benthic organisms.
3. To determine concentrations of viruses in CSO discharges
and their concentrations and persistence in fresh receiving
waters and near-outfall sediments.
4. To determine the distribution patterns and ultimate fate
of selected particulate discharge contaminants, including
copper, lead, zinc, carbon and phosphorus.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
FRESHWATER STUDIES
A preliminary study of sediments near 29 combined sewer
outfalls (CSOs), storm drains (SDs) and control sites in Lake
Washington provided the following observations, which served
as a guide for the selection of the principal (intensive) study
sites:
1) A composite enrichment index (including metals, pesticides,
PCBs, organic carbon, total phosphorus and oils and
greases) indicated that the southwestern portion of the
lake had the most highly contaminated nearshore sediments.
These pollutants were from combined sewers and storm
drains serving the City of Seattle. Several stations
along this shore had a composite sediment enrichment of
more than 16 times background levels. The sediments at
the northern end of Mercer Island were also found to be
highly contaminated.
2) The measured pesticide content of the sediments was predom-
inantly DDT and its degradation products. Contamination
levels of pesticides along the Seattle shoreline of Lake
Washington were up to 37 times background concentrations.
Pesticides were present in the sediments around both
CSOs and SDs and gave evidence of being associated
predominantly with storm water in both types of systems.
Intensive studies of contaminant loading, discharges plume
circulation, the characteristics of settling particulates,
and sediment contaminant distributions were carried out all
or in part at 2 CSOs, 2 SDs and 2 control sites, with the
following conclusions:
1) Discharge plume distributions were studied only in the
absence of thermal stratification of the lake. At all
study sites, including one CSO, one SD and one control,
major secondary inputs were detected and identified as
construction runoff or discharges from neighboring CSOs
and SDs. The discharge plumes from the study outfalls
rose through the water column and formed surface lenses
-------
of turbid water covering up to 10.1 hectares while still
intact as a definable entity. A fraction of the dis-
charged particulates remained thus for as much as a day
or more following an overflow, (depending on local
circulation dynamics); another portion was found to
quickly settle around each outfall and then move off-
shore along the bottom.
2) Circulation patterns at some of the study sites were
defined as unexpectedly complex. Turbid storm drainage
was observed entering a control site via longshore
advection.
3) The representative CSO and SD sites had significantly
higher mean sedimentation rates than did the control area
by an average factor of 2.9 times. Data from a related
study indicate that sedimentation rates just offshore from
the outfalls are 2.5 times the control values, reflecting
the near-bottom offshore movement of discharge par-
ticulates. The total flux (input per unit area) to the
control sediments of all particulate contaminants was
determined to be significantly lower than for the CSO or
the SD.
4) Sediment and discharge particle size distributions were
determined for the representative CSO and SD. The median
particle size for the discharge particulates was appreciably
smaller for the SD than for the CSO, indicating a higher
potential for transport away from the outfall. There was
evidence that the turbulence caused by the discharges
washes the finer particulates away from the outfall area
at both sites. There were indications of the breakdown
of particulates and the dissipation of organics near the
outfalls during dry periods.
5) The general trends observed with respect to wastewater
impacts on the benthic infauna near 2 CSOs and 2 SDs
implied enhancement of oligochaetes(aquatic earthworms)
at all sites, with the greatest enrichment of the worm
communities occurring at the CSO locations. This
difference between site types was assumed to be due
to the fact that much of the particulate matter in the
CSO discharges had been predigested in human guts, making
the carbon more readily available to microorganisms,
which are in turn the principal food of oligochaetes.
On the other hand, there was also evidence of near-
outfall depletion of chironomids (aquatic insect larvae)
at three of the four study locations, implying toxic
effects and/or impacts of substrate alterations. For
chironomids, there were indications that a permanent zone
ot depletion (coinciding with an area of visible discharge
-------
debris) existed around the outfalls; populations
within these areas remained depressed the year around,
whereas those farther away varied inversely with the
seasonal rate of discharge. The impacts on the populations
of copepods (microcrustaceans) and nematodes (roundworms)
were more variable and less extreme; these organisms
were negatively affected in most instances. The statistical
analysis of the pelecypod (freshwater mussel) population
indicated only slight and varied reactions to the presence
of the CSO or SD discharge.
6) Using regression analyses of infaunal biomass as a function
of distance from an outfall, estimates were derived for
biomass increase or decrease due to CSO and SD discharges.
The impacted area around a representative CSO was determined
to be 0.6 hectares (1.4 acres). The total mean oligochaete
biomass determined for two control areas of similar size
and depth was 57.9 kg (127 Ib) fresh weight. The impacted
outfall area had an estimated additional biomass (enrich-
ment) of 47.3 kg (104 Ib) of oligochaetes. It was further
estimated that this weight of prey organisms implied a
potential increase of about 5 kg, or 10 Ib of consumer
organisms (fish) per year. Oligochaetes constituted
over 90% of the total infaunal biomass at this station,
and the described discharge impact was the most extreme
one observed. The effects of consumption of tainted
organisms by fish were not assessed. Also it was
acknowledged that potential negative effects of the dis-
charges on fish reproduction, not measured for this
project, constitute a pollutional influence of unknown,
and possibly serious magnitude.
7) Human viruses were readily detected in the CSO discharges.
None were found in storm drainage. The level of viruses
in a given sample of combined discharge was linked to
time of day as related to the cyclic use of local restroom
facilities.
8) No human viruses were found in receiving water samples
collected near the SD. During an overflow at the CSO,
however, viruses were detected at a level representing an
estimated end-of-pipe dilution of 32:1. No viruses were
found near the CSO 24 hours after the overflow. It was
concluded that combined sewer overflows may constitute
a health hazard for people using Lake Washington beaches
soon after overflows, as in the summer months. This
conclusion was based on literature indicating that
viruses present at detectable levels can cause infection.
9) For a total of 42 separate samples collected near the CSO
and SD outfalls soon after overflows, no viruses were
detected in the top 3 cm of the bottom sediments.
-------
10) The fraction of storms resulting in overflows was determined
to be approximately 76% for the one CSO and the one SD
monitored, with 48 overflows at each site during the one
year study period, for which the total incident rainfall
was 59 cm. The total overflow volumes for the two
systems during the one year period were quite similar,
being 25,600 m3 (6.76 MG) for the CSO and 27,700 m3 (7.32
MG) for the SD. The non-storm base flow (near-surface
soil "interflow" plus groundwater infiltration) at the
SD added another 50,200 m3 (13.3 MG) to its total discharge.
11) For one CSO and one SD with similar size drainage basins
(120-acre average), land-use types (residential), and
rainfall, the mean storm overflow concentrations for many
of the measured parameters (including suspended solids,
copper, zinc, total phosphorus, oils and greases and
selected chlorinated hydrocarbons) were lower at the SD
than at the CSO. This was true for both the particulate
and total (i.e. particulate + soluble) contaminants.
Lead concentrations were much higher at the SD. However,
because CSO flows are intermittent by nature whereas those
from storm drains are continuous, the mean pollutant load
discharged by the SD during a given storm period was
typically similar to that from the CSO for all parameters
except lead, total phosphorus and oils and greases. The
storm loading for Pb was generally much higher at the SD,
whereas the total mass discharge for TP04-P arid O&G was
lower.
12) The total mass of solids discharged by the principal
CSO and SD monitored during the one year study period was
approximately three metric tons each, with 20 percent of
the SD solids being contributed by non-storm base flow.
Of the particulate fraction, the respective annual mass
loadings from the CSO and the SD were 10.7 kg and 5.2 kg
of phosphate-phosphorus, 5.2 kg and 10.3 kg of heavy metals
and 1.3 g and 1.2 g of pesticides.
13) Contaminant distributions for the surface layer of the
bottom sediments around the representative CSO and SD
outfalls (representing residential land use) indicated
localized enrichment, with apparent modifications from
current action and near-bottom downslope streaming. The
concentrations of metals in the sediments were generally
only 20-50% of those measured for settling particulates,
implying selective removal of the contaminated particulates
to deeper areas. Statistical analysis of the relative
surface sediment distributions for copper, lead and zinc
indicated different transport characteristics for each
metal due to their association with particulates of
different sizes and/or weights. Sediment enrichment with
lead, zinc and copper, respectively, was determined to
10
-------
average 9 times, 2 times and 3 times near the outfalls and
7 times, 2 times and none at the control site.
MARINE STUDIES
The conclusions for investigations carried out at and near
Metro's Denny Way Regulator, a computer-controlled combined
sewer overflow facility, were as follows:
1) Determination of the discharge plume distributions was
complicated by a 1.5-3.0 m (5-10 ft) thick surface layer of
highly turbid water from the Duwamish River moving past the
Denny Way facility. However, the river water was generally
excluded from the nearshore area just north of the outfall
by the CSO plume. Following one or more complete tidal
cycles, the CSO plume assumed a more symmetrical
distribution around the outfall. The area measurably
impacted by settling particulates extended 200-300 m along
the shore in both directions from the outfall.
2) Compared to the particulates emitted by the study
outfalls in Lake Washington, the solids discharged at Denny
Way had a much higher organic content and smaller median
size. The surface sediments near the outfall showed
evidence of heavy inputs of CSO particulates. As opposed
to the drainage basins of the CSO and SD studied in Lake
Washington, which are less than 5% commercial/industrial,
over one third of the Denny Way drainage area is
commercial/industrial. The total area of the marine
system is more than 5 times that of either of the fresh-
water basins.
3) With respect to substantial impacts on the local aquatic
biota, the nearshore area within 150 m of the Denny Way
outfall can be referred to as polluted. This was confirmed
by a sampling regime used to quantify six different
biological parameters known to be sensitive indicators of
water and sediment quality. The relative effects were
evident at the 9 m depth contour, but somewhat diminished
farther (approximately 100 m) offshore at 13 m depth.
There was some evidence that during periods of few over-
flows the shallow subtidal infauna does undergo a small
degree of recovery toward more natural conditions.
The results of the biological studies correlate well with
previous work done at the same site.
4) The fraction of total storms resulting in overflows at
Denny Way was 51% for the study year, the total number of
overflows being 36. The total overflow volume was 6.60
X 105m3 (175 MG) for an annual total rainfall of 99 cm.
The associated mass of suspended solids emitted by the
11
-------
outfall during this period was 85.1 metric tons. Of this
particulate fraction, 7.6 metric tons was organic carbon,
approximately 0.4 metric ton was heavy metals, and less
than one gram was pesticides.
5) Compared to the combined sewer and storm drain discharges
monitored in Lake Washington, the mean concentrations of
heavy metals in both the particulates and the particulate +
soluble wastes discharged at Denny Way were higher,
reflecting the large component of industrial and commercial
inputs to the latter system. However, its mean concen-
tration of particulate chlorinated hydrocarbons was much
lower, implicating residential use as the principal source
of pesticides.
12
-------
SECTION 3
RECOMMENDATIONS
Further work is needed to relate the localized findings of this
project to Lake Washington as a system. It is suggested that
mesoscale studies be conducted that cover at least one extended
nearshore section (perhaps 2 km long) substantially inundated by
CSOs only, and likewise one affected by storm drainage only; such
areas have been identified in the lake. Toxicant distributions
in these areas (including those for the EPA organic priority
pollutants, which have recently been found in substantial concen-
trations in local freshwater and marine sediments) could be
defined from analysis of sediment grid samples and related to
detailed biotic community distributions determined by numerical
classification analysis. One major advantage of mesoscale
studies is the statistical elimination of the variability asso-
ciated with single drainage systems.
The virus data described herein constitute a good foundation, but
survival measurements and more detailed nearshore circulation
studies would permit public exposure estimates as an adjunct to
epidemiology measurements presently underway.* Virus measure-
ments also should be made in marine waters in locations such as the
Denny Way Regulator, which adjoins a public park having some water
contact recreation. In addition, our failure to locate viable
viruses in the near-outfall sediments presents a major conflict
with work done elsewhere (Gerba et al., 1977); the reasons for
this apparent contradiction need to be resolved.
Although the present work provided estimates of wastewater im-
pacts on standing crop numbers and biomass for infaunal communi-
ties, it was noted that reproduction and toxicant tissue burden
information on fish constitute very important deficiencies in
our knowledge of the biological impacts on higher organisms.
*The Municipality of Metropolitan Seattle is presently (1979-1980)
involved in an epidemiology study in conjunction with EPA and the
University of Washington that is designed to 'identify relation-
ships between disease incidence in swimmers and levels of indi-
cator bacteria. The determination of a good indicator organism
would permit local authorities to set meaningful standards for
the safe use of recreational waters.
13
-------
It is suggested that investigations of this sort be done in Lake
Washington using the prickly sculpin, an important primary
carnivore and principal food item for many of the larger fish
species. This fish exhibits only limited migration tendencies
and might be expected to remain within impacted or non-impacted
areas for the duration of such a study. A commercially-harvested
crayfish species also abounds in the lake and could be investi-
gated for effects of CSO and SD wastewater discharges.
Light transmission measurements were shown to be useful indica-
tions of general circulation trends. Their utility might be
extended through correlations with grab sample analyses of
various particulate pollutants, comparisons that would provide
further information relative to the differential particle sorting
patterns and solids dispersion characteristic of each snecies
Light transmission measurements are also needed as indicators'of
wastewater circulation patterns under stratified conditions in
the lake.
Near-bottom time series of light transmission data would help to
clarify the velocity and dimensions of turbidity plumes streaming
offshore from an outfall. Sectioned cores collected along an
onshore-offshore transect identified by this approach would
provide age-dated profiles of pollutants to further our knowledge
of particulate deposition rates and patterns.
Finally, it is suggested that study plans be made to take
rnn^ag^-°f ovefflow abatement work imminent in Lake Washington.
o? ?hfUr£n°n9? <-5e?tn S°°n On * new PumPin9 system downstream
of the CSO 023 outfall, the principal freshwater combined sewer
outfall investigated for the present project. The effect will
be that overflows will cease entirely at CSO 023 in about two
years. The opportunity to study the response of the environment
to this alleviation of pollutant stress is rare and should be
turned to good account.
14
-------
SECTION 4
METHOD OF STUDY
SELECTION OF SAMPLING SITES
Preliminary Freshwater Studies
P3:ior to tne Present study, no comprehensive map was available
giving the locations and dimensions of the sewer outfalls in
Lake Washington. On the lake's western shore the Seattle sewer
system and outfalls represent a composite of ancient and modern
facilities that have grown with the city since its founding
in 1851. Overall, the documentation for these structures
is extensive. However, much of the eastern shore was settled
in the last two decades; all of the sewers there are separated,
and the details of storm drain facilities (including locations)
had not been comprehensively plotted.
Drawing information from a number of sources, as much of the
available outfall information as feasible was compiled to provide
a comprehensive choice of study sites. A base map was con-
structed using sub-basin drainage charts (RIBCO, 1974).
Further details were obtained through visits to the files of
pertinent agencies representing the nine sewer districts sur-
rounding the lake. The resultant map (Figure 2) included 23
emergency outfalls (activated only by power failures), 34
CSOs, 56 pump stations and 240 SDs. For Lake Washington's
115 km of shoreline, this gave a conservative average of three
outfalls per km. The map, although not complete, was felt to
include all of the major drainage systems.
Using the compiled information, candidate sampling sites (13
CSOs, 20 SDs and 15 control locations) for the preliminary
study were selected. In this, we were guided by two principal
criteria: 1) that the locations be as free as possible of con-
tamination from other outfalls, and 2) that the selected
facilities have a high probability of large and frequent storm
overflows or direct drainage responses.
Outfall interferences were judged from map information. Data
on estimated (modelled) overflow frequency, volume and duration
for each of the CSOs were obtained from the City of Seattle
15
-------
N.E. LAKE WASHINGTON
C13
:*:*:*:•:•:*:•:*:•:•:*:•:•:•:•:•:•:£ L L i OT
I I FINAL STUDY SITE
• COMBINED SEWER STUDY SITE
0 STORM DRAIN STUDY SITE
O CONTROL STUDY SITE
•COMBINED SEWER OUTFALL
- STORM DRAIN OUTFALL
SANITARY SEWER
EMERGENCY OUTFALL
Of- PUMP STATION
LOCAL JURISDICTION
• METRO PUMP STATION
SEWER DISTRICT BOUNDARY
Area not mapped
RENTON
Figure 2. Locations of sampling sites.
16
-------
Engineering Department; because similar information was un-
available for the SDs, those facilities having the largest
drainage basins were chosen.
Further on-site inspection resulted in elimination of some of
the sites due to uncharted interferences or predicted diffi-
culties with equipment installation, servicing and protection.
The remaining stations were rated on the basis of the various
concerns noted, and ten in each category were chosen for the
preliminary field studies (Figure 2) .
Intensive Freshwater Studies
Based on the results of the preliminary survey, six stations
(two CSOs, two SDs, two control sites) were selected for inten-
sive study. To this end, two additional criteria were added to
those previously mentioned: 1) that the CSOs and SDs have
strong localized indications of contamination by wastewater
effluent, and 2) that the control areas be geographically close
and physically similar to the relative CSOs and SDs. The
locations of the final six stations (CSOs 023 and 044, SDs 7 and
19 and Controls 3 and 4) are shown in Figure 2 with their code
numbers enclosed in rectangular boxes.
Marine Studies
^ marine studies were carried out at one of the largest (in
terms of both size and discharge "volume) combined sewer overflow
facilities in the Seattle area. The Denny Way Regulator Station
is a computer-controlled CSO that discharges mixtures of raw
sewage and street runoff into Elliott Bay on Puget Sound
(Figure 2) . This facility was selected for study on the basis
of its size, frequency of overflow (average of 38/year) ,
accessibility, and its location in the midst of a large public
park, which extends 350 m to the south and 1100 m to the north
along the shore of the bay.
DESCRIPTION OF SAMPLING SITES
Preliminary Freshwater Studies
The descriptions given below for the intensive freshwater study
sites may be considered generally representative of conditions
observed at the 14 additional preliminary sites.
Intensive Freshwater Studies
The locations ultimately chosen for the intensive freshwater
studies (Figure 2) were the best available but did not fully
meet all of our specified criteria. For any of the four out-
falls selected, there was a substantial potential for discharge
17
-------
interference from other, nearby systems. This is particularly
evident in Figures 3 and 4, which show the site configurations
for the two CSOs. In each case, a large storm drain was found
within several hundred feet of the combined sewer outfall.
However, the approximate dimensions of the zone of debris accumu-
lation observed by the divers indicated that the separation dis-
tances were probably sufficient to minimize interferences; also,
probable current patterns deduced from dye studies in the general
vicinity of these two locations (CH2M/Hill, 1974) indicated that
the discharge plumes were likely to be parallel in most instances
and not convergent.
The two combined sewer outfalls were found to have scoured pits
at the ends, with built-up piles of debris somewhat further off-
shore. The pits were approximately half a meter deep and one
meter wide; the mounds were as much as 1.5 m high and 14 m across
(somewhat smaller at CSO 044 than at CSO 023). The discharge
points for the two storm drain systems were nearshore (SD 7) or
onshore (SD 19) in localized riprap, and opened onto much steeper
terrain than did those of the CSOs, thus minimizing scouring and
debris buildup.
Evidence of scouring and deposition was found to be common at
the many outfalls investigated for the preliminary studies; at
each such location, the sediment samples were taken just past
the crest, on the back side of the debris mound. Active outfall
structures were also found to have transient contiguous areas
of visible debris accumulation, as large as 0.6 hectare (1.4
acres), as indicated in Figures 3 and 4. The dimensions of the
biological sampling arrays, the locations of the sediment traps
and the transmissometer grid configurations are also shown in
these diagrams. The pipe diameter for all four outfalls was
61 cm. The typical substrate for the various areas was: CSO 023
- fine sand, silt, light debris; CSO 044 - pebbles, sand, silt,
clay, light debris; SDs 7 and 19 - fine sand, silt, some clay,
light debris; C 3 - sand; C 4 - sand, some silt, light debris.
The service areas associated with the four outfalls were: CSO
023 - 47.3 hectares, CSO 044 - 67.2 hectares, SD 7 - 50.6
hectares, SD 19 - 42.9 hectares. The CSO 044 sewers were partly
separated, with street drainage being discharged to the lake
through a nearby storm drain (Figure 4); rooftop drainage was
discharged through the combined system.
Marine Studies
The marine studies were carried out around Metro's Denny Way
Regulator CSO, where discharged materials enter the receiving
waters through a conduit located in the intertidal zone. The
sampling program was conducted in the intertidal and shallow
subtidal (down to 13 m below MLLW, Mean Lower Low Water) regions
in an area bounded by Pier 70 to the south and Pier 91 to the
18
-------
*30m W
OUTFALL
LINE
< 30m E
(a) Combined Sewer Outfall 023
OUTFALL
107m E
SED. TRAP
ARRAY
23m E
T SEDIMENT TRAP
O OUTFALL TERMINUS
-6.1"' DEPTH IN METERS
*~XOUTSIDE DIMENSIONS OF BIOLOGICAL SAMPLING ARRAY
E23ZONE OF VISIBLE EFFLUENT DEBRIS
• TRANSMISSOMETER SAMPLING LOCATION
Figure 3. Principal set of intensive sampling sites in Lake Washington.
19
-------
a) Combined Sewer Outfall 044
SEDIMENT TRAP «
O OUTFALL TERMINUS
*"* OUTSIDE DIMENSIONS OF BIOLOGICAL SAMPLING ARRAY
USA ZONE OF VISIBLE EFFLUENT DEBRIS
• TRANSMISSOMETER SAMPLING LOCATION
,6.1- DEPTH IN METERS
•169m
Figure 4. Seoondary set of intensive sanpling sites in Lake Washington.
20
-------
north (Figure 5). The intertidal region is primarily made up of
a steeply sloping man-made boulder sea wall which extends from
a point above Higher High Water down into the subtidal. This
wall is interrupted by two small coves. The coves consist of
a gently sloping beach comprised of mixed sand, silt and cobble
substrata. The CSO conduit is located in the southernmost of
the two coves. The base of the conduit is positioned at about
+2 feet above MLLW (tidal heights will be referred to hereafter
as feet above MLLW). An area of scouring due to flows from the
CSO is evident and runs seaward from the base of the conduit to
the waters edge at low tide (i.e., at least +0). This scoured
area was termed the "wash out" zone by Armstrong et al. (1978).
The sediment is finer grained there, and debris is usually piled
at the base of the conduit in this zone. The boulder wall and
coves form the shoreline of a public park.
In Figure 5 are shown the locations of all of the sample sites.
These are the same as those sampled by Armstrong et al., with
some exceptions: 1) the subtidal sites at 24 m depth were not
sampled by us because no impact was evident in the previous
study; 2) the transect numbers for the subtidal sites at 9 m
and 13 m were made consistent with their relative position with
respect to the CSO (see Armstrong et al., Figure 4); and, 3)
new sites were established for sampling Enteromorpha.
FIELD SAMPLING METHODS
Combined Sewer and Storm Drain Discharges
Quantity and Quality —
For selected rainstorms, the overflow discharges at Combined
Sewer Outfall (CSO) 023, Storm Drain (SD) 7 and the Denny Way
Regulator (refer to Figure 2 for locations) were sampled for
loading analysis of suspended solids, metals, total organic
carbon, total phosphorus, oils and greases, and total chlorinated
hydrocarbons. Automatic sequential samplers triggered by
pressure switches were used in conjunction with Arkon pressure-
driven flow recorders to provide 2 1 samples at regular intervals
during the overflows. The sampling intervals were chosen
according to the speed of flow response in each system, so that
samples would be taken that were representative of all of the
major high and low points on the overflow hydrographs.
Plume Distributions --
During a given storm a grid of up to 42 stations (at Denny
Way) was covered at each sample site; at each grid point a
Martek Model XAS Transmissometer was lowered to the bottom,
giving a real-time plot of percent light transmission
versus depth on a shipboard XY plotter. Each cast took a
maximum of two minutes (to a maximum depth of 24 m (80 feet),
with a complete grid coverage requiring about one hour in
21
-------
N
E ENTEROMORPHA SITE
P PERIPHYTON SITE
B BIOASSAY SITE
M MACROALGAL SITE
) (J3)SUBTIDAL SITES
(DEPTH IN METERS)
a—.INTERTIDALSITE
TRANSECT 1
(CONTROL)
TRANSECT 6
TRANSECT 7
PIER 70
Figure 5. Details of sampling transects around
the Denny Way Regulator Outfall, 1978,
22
-------
the lake and two and one-half hours at Denny Way. Whenever
possible, each station was visited more than once to measure
various stages of the plume development and dispersion.
For grid-site location, a one-time survey was carried out
wherein onshore-offshore transects were established visually
using landmarks. Rangefinder distances were then determined
along these transects using buoy markers when needed. Depth
readings from the shipboard depth finder were recorded for
each point, and used for all subsequent outings to determine
station locations along the transect. The estimated maximum
positional error for the stations farthest offshore was 15 m
and was somewhat less for inshore stations.
Temperature and oxygen probes were mounted on the transmissometer
and were used to periodically check the stratification of the
water column. These sensors were interfaced to a Martek Mark V
Water Quality Monitor and an X-Y plotter. Both units were lab-
calibrated prior to use in the field.
Particle Size Distributions --
The solids grain-size samples of overflow discharge were
collected from a single overflow at CSO 023, SD 7 and Denny
Way. Single, manual grab samples were taken from the CSO
023 and SD 7 facilities, whereas that from Denny Way was
sampled over a five-day period using three automatic composite
samplers. Nearly 750 1 of effluent were collected at
each station.
Settling Particulates
Settling particulate matter was sampled continuously from
1/30/78 to 1/29/79 near the wastewater outfalls at CSO 023
and SD 7 and from 10 to 60 m offshore at the Control 3 site
(Figure 2). The sampling devices, or sediment traps, consisted
of 30 cm X 30 cm polyvinyl chloride platforms, which held four
10 cm diameter funnels joined to 50 ml centrifuge tubes (Figure
6). Each trap collected a total area of 314 cm^, all parts of
the platforms (including screws) were made of plastic to limit
metals contamination. Rigging lines were passed through the
center of each trap platform and connected to two polyurethane
floats above the collection surface and to a concrete anchor
below. This configuration kept the collection surface 2 m
from the lake bottom. Thin nylon support lines radiating
from the central line to each corner of the trap platforms
served to keep the collection surface level in the water column.
For each collection period, the traps were serviced and
cleaned, with retrieval and reset being done by SCUBA divers.
This method of collection was designed to minimize stirring
of the lake bottom sediments.
The sediment trap arrays used for CSO 023, SD 7 and C 3 are
23
-------
POLYURETHANE
FLOATS
CENTRAL LINE
SUPPORT LINE
PVC LATCH
PVC PLATFORM
FUNNEL
TYGONSLEEVE
50 ml CENTRIFUGE
TUBE
BRASS SNAP LOOP
EXCESS LINE
CONCRETE ANCHOR
SEDIMENT-WATER
INTERFACE
Figure 6. Details of sediment trap configuration
24
-------
shown in Figure 7. Sediment trap placement was determined using
the visual and photographic observations of the divers. The
rather distinct line of transition between the two types of
sedimentary material was denoted the plume boundary.
OUTFALL PIPE
(7.9m)
GO (9.1m)
(12.2m)
[Dili.5m)
'GQ(11.3m)
a) COMBINED SEWER OUTFALL 023
fl OUTFALL P»f
(3.0m) Q] *""^ 70'
(2.4m)
1 (D (4.0m)
[J|(8.8n
(7.6m) [4]
b) STORM DRAIN 7
(1](11.3ml
10m
(5.2m) [I]
(7.9m) [3]
c) CONTROL SITE 3
[D (3.7n
[5]<6.7r
[3 (7.6m)
Figure 7
Configurations of the sediment trap arrays moored
at the principal sampling sites in Lake
Washington. The positions of the traps are
indicated by boxed numbers. The values in
parentheses are the location depths.
Sediments
Surface Samples —
Sampling for preliminary analyses of the freshwater sediments
was carried out by SCUBA divers. Each station was initially
surveyed as to depth and bottom contours, nature and condition
of the outfall pipe, extent of apparent debris accumulation
and proximity and nature of interfering structures. At each
CSO and SD, duplicate samples of the surface 3 cm of the
(apparently) most highly contaminated sediments were collected
in specially-prepared 110 cm3 glass cylinders closed at each
25
-------
end with rubber stoppers. For the sake of chlorinated
hydrocarbon/PCB analyses, the stoppers of one of each set
were covered with aluminum foil. For the aluminum analyses
the stoppers of the other were left bare.
Cores —
Lake sediment cores for analysis of sediment profiles of
metals and carbon were also collected by divers at the
transmissometer grid sites. Thirty cm lengths of clear
tenite butyrate tubing (3.5 cm ID) were used for this purpose.
Cores from depths greater than 15 m were taken with a 7 kg
modified Phleger gravity corer.
Particle Size Distributions —
Particle size distribution samples were collected by the
divers just outside the area of visible overflow influence
at the ends of the outfall structures at Denny Way and
Madison Park (CSO 023).
Benthic Biota
Freshwater Studies --
The freshwater benthic biota were manually sampled by divers
using white, polyvinyl chloride core tubes. This method of
sampling meets all of the requirements of the ideal sampling
device as described by Brinkhurst (1974) .
Several important questions were resolved in the process of
refining the sampling methods, including the selection of
the most appropriate core dimensions and the number of cores
that would be required at each sampling site to provide
statistical confidence in the data. Brinkhurst (1974) feels
that only a small percentage of animals live in sediment
deeper than the top 6 cm. In order to sample nearly 100% of
the organisms at the site, 15.3 cm was chosen as the depth
of core penetration. The core tube diameter was 2.9 cm.
An adaptation of the techniques used by English (1964) was
used in a preliminary study to determine the number of cores
to be taken at each sampling site. For this work, 20 cores
were collected from each of several sites within a selected
nearshore area. Means for chironomid, oligochaete, and copepod
counts were calculated for each core and confidence intervals
were calculated for each number of cores ranging from one to
the maximum number taken. By expressing the confidence intervals
as percentages of the sample mean we were able to choose the
most cost-effective yet statistically acceptable number of cores
to be taken at each site.
Twenty cores of each of two experimental diameters (29 and
38 mm) were taken from each of three depths - 1.5, 4.6 and 7.6
meters. Core locations in an 81-division, 0.5 m2 wire sampling
26
-------
grid were chosen for each depth with the aid of random number
tables. The contents from each tube were preserved with 10%
formalin, and stained with Phloxine B dye.
Due to documented seasonal differences in the populations of
benthic organisms (Patten et al., 1976), sampling was carried
out during two seasons. The February sampling period was
chosen as a period optimum for detecting pre-emergence
conditions. The late summer or fall period was chosen as
the optimum time for post-emergence observations.
In all areas, station depths and distances from the outfall
were located by divers. The transect line was generally
anchored at the center sampling station and placed at the
desired angle from the shoreline. The divers proceeded
along the transect line to the desired station depth where
the distance was recorded. The 0.5 m2 (81-division) sampling
grid was placed at each sampling station and eight core tubes
(to assure six cores) were inserted into the predetermined grid
compartments. Core tubes were marked with grid location
designations selected from random number tables. On the boat,
the samples were emptied into plastic bags and preserved with
full strength formaldehyde mixed with the water in each sample
to form a standard 10% formalin solution. In the laboratory,
organisms from the cores were counted, sorted and preserved in
70% isopropanol.
Marine Studies —
In situ bioassay studies were started at the marine CSO (Denny
Way) during February, 1978. Two hundred individuals each
of the blue mussel, Mytilus edulis, and the Pacific oyster,
Crassostrea gigas, were place in four dome-shaped plastic
crab pots (22 cm high X 60 cm across), and anchored at
Mean Lower Low Water (MLLW) at various distances from the
outfall (indicated by Bs in Figure 5).. Eight to ten individuals
of each species were removed monthly for condition index
analyses (refer to laboratory methodology for details). In
addition, extra specimens were collected during February and
April, 1978 for trace metals analysis.
The length-width relationship of the green alga Enteromorpha
intestinalis was studied because previous observations from
samples along the boulder wall had indicated a correlation
between distance from the CSO and the length-width ratio as
a measure of growth. Increased growth rates of a closely-
related species (E. prolifera tubulosa) have been observed
in areas inundated by wastewater (Tewari, 1972). Collections
were made at each site by carefully scraping a number of
individuals off of the hard substrata. These were placed in
plastic vials and preserved in 5% formalin. The sampling
was conducted in April and August of 1978.
27
-------
To sample periphyton communities, blocks made of half pumice
and half concrete with a flat surface area of at least 450 cm2
were placed at Mean Low Low Water (MLLW) at various distances
from the CSO. At this tidal level, diatoms are generally the
most abundant alga by cover. The periphyton that attached to
the blocks was sampled in April after about four weeks of
exposure. The periphyton was brushed into glass jars with a
stiff toothbrush, and preserved in 5% formalin.
The epibenthic community that attached to the boulder wall
at MLLW was sampled at various sites in April and August. A
0.06 m2 plexiglass plate with 20 randomly-located points was
tossed non-selectively 20 times in the region around each
site. The taxa that occurred under each point within each
quadrat were identified and given a score equal to the
number of points which covered the entity. Taxa that occurred
within the quadrat but not under a point were identified and
given a score of one.
Intertidal soft substrata infauna were sampled during April and
August at seven sites in the CSO cove and three sites in the
control cove to the north (Figure 5). A cylindrical plexiglass
tube (31.2 cm2 in cross-sectional area and 15 cm deep) was
used to extract four sediment cores at each location. The
sample of sediment was placed in a plastic bag and was then
preserved in 10% formalin and stained with Rose Bengal
stain. At each site, a 3-5 g sediment sample was taken for
volatile organic analysis.
The subtidal infauna was sampled at two depths along seven
transects during April and August (Figure 5). Collections of
bottom sediments were made with a 0.1 m2 Van Veen benthic
grab. Two grabs were taken from each site. A 3.5 g sample
of the sediment was obtained from the first replicate at
each site to be analyzed for volatile organic content. The
samples were screened through a 1 mm mesh screen and the
material retained on the screen was preserved and stained in
10% formalin and Rose Bengal Stain.
Viruses
For each overflow sample, approximately 285 1 of the discharge
were pumped from the pipe into a 380 1 polypropylene container.
Two equal-volume aliquots (ranging from 76 to 190 1 ) were
subsequently taken from the larger sample for analysis.
Collection methods for the receiving waters were similar.
The overflow plumes were first located by light transmission
measurements. Approximately 285 1 of the most turbid water
were then pumped into large polypropylene containers on board
the boat and was transported to shore for further processing.
Follow-up samples were collected in the same locations 24
28
-------
hours later.
Three sets of sediment samples were collected at the sediment
trap locations at CSO 023 and SD 7; one set was collected at
the Control 3 site. The sample material was scooped from
the top 2-cm layer with sterilized glass tubes. The first
set was collected by divers 7 days after a storm, whereas
the second and third and the C 3 samples were taken from Van
Veen grabs.within one day following a storm.
ANALYTICAL TECHNIQUES
Combined Sewer and Storm Drain Discharges
Quality Analyses —
Individual analyses were performed as follows:
Total Suspended Solids - determined in conjunction with the
total chlorinated hydrocarbons analyses, as specified
below.
Metals - split sample, acidified one part with concentrated HNO-,
Filtered the other through a .457 urn membrane filter that
had been prerinsed with 1% HN03 and deionized water.
Acidified the filtered sample with concentrated HNC>3. Both
samples analyzed for Cu, Pb, Zn, Al on IL Model 453 Atomic
Absorption Spectrophotometer, and for Hg on Anti-Pollution
Technology Corp. Model 2006-1 Mercometer. Only the
unfiltered sample was digested prior to analysis.
Total Phosphorus - split sample, filtered one part through
Whatman #40 filter. Acidified both fractions with 0.5
ml 5 N H-SO./SO ml sample. Samples digested as for Cu,
Pb, and Zn (aliquots were taken for analysis from the same
digestion mixtures). PO4 measured as orthophosphate by
the ascorbic acid method (APHA, 1975).
Chlorinated Hydrocarbons - filtered all of sample (for
analysis of particulates only) through muffled, tare-
weighed Whatman GF/A glass fiber filters. Filter then
dried at 105°C and cooled in desiccator. After final
weight was taken the sample was extracted by Soxhlet
with acetone for eight hours. Extracts dried, then
infused with petroleum ether and concentrated. Cleanup
and fractionation of dissolved components done in
Florisil columns by elution with various mixtures of
ethyl and petroleum ether. Analyses were done by gas
chromatography, using a Tracer Model 222 GC fitted with
a Ni high-temperature electron capture detector.
29
-------
Oils and Greases - Hexane/Soxhlet extraction method
(APHA, 1975).
Total Organic Carbon - Potassium persulfate/phosphoric acid
digestion of blended samples in nitrogen-bubbled,
sealed ampoules. Resultant CO2 was conveyed by nitrogen
carrier gas through a Model 865 Beckmann Infrared
Analyzer. Strip chart peaks were compared to those of
similarly-treated standards.
Particle Size Analyses --
The 750 1 samples collected from the CSO 023, SD 7 and
Denny Way overflow structures were reduced to a total volume
of approximately 4 liters each by a succession of settlings
and decantings over a period of about one week. Following
the initial volume reduction, the samples were refrigerated
at 4 C to minimize deterioration.
The testing of the consolidated overflow particulates consisted
of organic content and grain-size determinations. The
former was estimated by burning off the organic material
after first removing the sample moisture by oven drying at
65 C. Grain-size determinations were performed in accordance
with ASTM Test Designation D422-63, and included combined
coarse sieve and hydrometer analyses. This work was done
for Metro by the Seattle laboratory of Converse Davis Dixon
Associates, Inc., geotechnical consultants.
Settling Particulates
The sediment trap samples were centrifuged at 9000 rpm for 7
minutes, decanted and dried at 60°C for 36 hours. The
samples were then weighed individually and the four samples
from each trap were combined and comminuted prior to chemical
analysis. One hundred mg subsamples were then digested
using HF-HNO3-HC1O4 by a procedure similar to that of Bortleson
and Lee (1972), except that the entire digestion was done in
a single 10 ml teflon crucible (Birch, 1976). An aliquot of
this digestrate was used for determination of P, Cu, Pb, Zn
and Al.
The analyses were done as follows:
Phosphorus - ascorbic acid molybdenum blue method (APHA,
1975). The mean of duplicates was reported for each
sample. The mean coefficient of variation for 14 sets
of duplicate digestions was 4.8%.
Copper, Lead and Zinc - digestrate analyzed on an IL Model
353 Atomic Absorption Spectrophotometer. Precision,
including subsampling, was within 20%, and generally
better than 10%.
30
-------
Aluminum - two methods: as for Cu, Pb, and Zn, and analysis
by neutron activation. For 28 samples, with one sub-
sample analyzed by each method, the mean coefficient of
variation was 4.3%.
Carbon - analyses done on 20-100 mg of dried sediment using
a Leco induction furnace and carbon analyzer. The
results given are the means of two replicates. The
mean coefficient of variation for five samples (four
replicates each) was 6.0%.
Sediments
Surface Sample Analyses --
The excess water was drained off each of the surface sediment
samples collected for the preliminary study, and the sediment
was transferred to glass (for PCB and chlorinated
hydrocarbon analysis) and polypropylene (for all other
analyses) containers. The samples were weighed wet and then
oven-dried at 105°C. Aliquots taken for Hg analysis were
dried at 5QOC to minimize loss through vaporization. All
samples were weighed dry when cool.
Individual analyses were performed as follows:
Copper, Lead and Zinc - as for settling particulates.
Analyses were done by atomic absorption spectrometry
(APHA, 1975) on an Instrument Laboratory Model 353
Aluminum - as for settling particulates.
Mercury - dried sediments digested for 1 hour in concentrated
H2S04/concentrated HN03 mixture, infused with a KMnO3/KoS0OQ
mixture and digested for an additional 2 hours. 2 8
Analyses were done by cold vapor flameless atomic
absorption spectrometry using an Anti-Pollution Tech-
nology Corp. Model 2006-1 Mercometer.
Total Phosphorus - digestion and analysis as described for
discharges.
PCBs/Chlorinated Hydrocarbons - wet sediments extracted and
analyzed as described for discharges.
Oil and Greases - Hexane/Soxhlet extraction method (APHA, 1975).
Total Organic Carbon - Preliminary Study: inorganic carbon
driven off as C02 by acidification of samples to pH = 2 0
with HC1. Residual (organic) carbon measured by a Leco
Model WR-11 Total Carbon Analyzer by infrared analysis of
31
-------
CO2 produced through sample combustion. Intensive studies:
as for settling particulates.
Core Analyses —
The cores were first examined for gross physical characteristics
and then extruded and sliced at 0.5 cm and at 1 cm intervals to
12 cm depth. The slices were weighed, dried at 103°C for 24
hours, reweighed, and ground in a Diamonite mortar. Digestion
and analysis were done as described for settling particulates
on the 0-0.5 cm, 0.5-1 cm, 3-4 cm and 7-8 cm sections. Analysis
for chlorinated hydrocarbons was as for discharges. Due to the
large number of samples involved, some cores were stored at 5°C
for up to one week before processing.
Particle Size Analyses --
The sediment grain size determinations were done using the
methodology specified above for the discharge particulates.
Benthic Biota
Freshwater Studies —
For each sampling site at each station, six of the eight cores
were randomly selected for benthic organism counts. Four of the
six were randomly selected for dry weight analysis. The
cores were first washed in a #50 (.297 mm) screen to reduce
the amount of sediment. Ten cores from the September sampling
period were also selected to test the number of organisms
that would pass through a .500 mm screen but not through a
.297 mm screen. The results from this test can be used for
comparisons between this study and other benthic studies using
.500 mm screens.
The February samples were first screened and then placed in
bottles of 70% isopropanol. Sorting and counting took place
within several days. Organisms from approximately half of
the samples were counted and identified at the same time
that the organisms were separated from the sediment. For
the remaining cores the organisms were separated from the
sediment and stored for identification and counting at a
later date.
The weighing of the February samples was delayed until all
counting and sorting was completed. Most of the organisms
from the September samples, however, were counted, sorted and
weighed immediately after screening. Whereas it was recognized
that significant weight loss can occur for benthic organisms
while stored in preservatives, each of the two seasonal lots
received consistent treatment within itself, and no between-
season weight comparisons were made.
Organisms were separated from the sediment using binocular
dissecting microscopes. Identification was done with the
32
-------
aid of taxonomic keys by Pennak (1953). No attempt was made
to identify organisms to the species level.
Oligochaetes (aquatic earthworms) and nematodes (roundworms)
were counted only if the organisms were intact because head
and tail pieces were not clearly distinguishable. Worm
parts, however, were included in the dry weights. Although it
was anticipated that such a dichotomy of quantifying procedures
could weaken count-to-weight comparisons, this was considered
unavoidable due to time and task constraints. No eggs, egg
cases, empty shells or empty exoskeletons were counted or weighed
except for mature Oligochaetes found in egg cases.
Some organisms present in the samples were not counted or
weighed. Ostracods (seed shrimps) were recorded as either
present or absent, but were neither counted nor weighed
because they occurred in high numbers in most samples, were
difficult to pick out, and are usually neglected by benthic
researchers in freshwater. Terrestrial insects and larval
fish found in some of the core samples were not quantified
because they are not normally considered benthic organisms.
Organisms in the Order Cladocera (water fleas) were likewise
ignored because they are generally described as zooplankters
and not benthic organisms.
Some members of the Family Spongillidae (freshwater sponges)
were also observed in the benthic cores. These were not
counted or weighed because identification was difficult and
the size and thickness of a single sponge was highly variable.
Pennak (1953) states that sponges are generally associated
more with hard substrates than with muddy substrates.
Therefore, changes in the numbers or weights of sponges in
our core samples would probably be related more to substrate
than to pollution impacts.
Dry weights were obtained for samples dried at 80°C for 24
hours. The samples were held in a desiccator between drying
and weighing. The dry weights for each core were divided
into three categories: Oligochaetes, chironomid and ceratopogonid
larvae, and the rest of the benthic organisms combined.
Weight data were important for Oligochaetes because many
oligochaete parts, representing a significant part of the dry
weight, were not included in the oligochaete counts. Weight
data were important for both Oligochaetes and chironomid
larvae because they generally vary in size more than the
other benthic organisms and they are both important food
items for fish. In general, ceratopogonid and chironomid
larvae were counted separately, but both groups were combined
and weighed together because of their taxonomic and physical
similarities. The dry weights of all groups were combined
and used as an indication of the total biomass at each
sampling station.
33
-------
Marine Studies —
A condition index (CI) (Westly, 1961; Quayle, 1969) was
determined periodically for composite samples of 8-10 individuals
of each of the two mollusc species, Mytilus edulis (blue
Mussel) and Crassostrea gigas (Pacific oyster). The index
describes the plumpness of bivalves, and has been used to
monitor environmental conditions (Wass, 1967).
The condition index analysis was first done at the time of
experiment initiation, and monthly thereafter. For the
February and April samples, the concentrations of Pb, Zn,
and Cu were also determined for both species; this was done
by atomic absorption spectrometry following hot digestion in
successive baths of concentrated HNO3 and HC104.
For each periphyton sample, a wet microscope slide mount was
made of a small subsample of the material. The first 200
cells that were encountered on the slide at a magnification of
200X under a compound microscope were counted and identified
(usually to genus). Three subsamples were analyzed from
each sample.
The length-width ratios of the samples of the green alga,
Enteromorpha intestinalis, were determined for 50 randomly-
selected individuals per sample using a dissecting microscope
with an ocular micrometer at 120X.
All intertidal infauna samples were rescreened through a 1 mm
screen and rinsed with tap water to remove the formalin.
All organisms retained on the screen were identified and
counted. Subtidal infauna were sampled in April and August.
Volatile organic contents of both intertidal and subtidal
sediments were determined using the 3-5 g sample obtained at
the infauna sample sites. Each sample was dried for 30 hr
at 100°C, weighed, burned for three hr at 550°C, and reweighed.
The loss of weight, expressed at a percentage of the dry
weight, gave a rough estimate of the organic carbon present
in the sediments (Morgans, 1956).
Viruses
From each of the original discharge samples, two equal-volume
aliquots (ranging from 76 to 190 liters) were taken concurrently
for virus analysis. One of these samples consisted only of
the overflow water. To the other sample a known level of
seed virus was continuously added, and both samples were
subsequently passed through the filtration apparatus.
From these two samples the percent recovery of virus was
calculated.
Viruses were concentrated by a modified version of the
tentative standard method of finished waters described in
34
-------
APHA (1975). Sample water was first passed through an orlon
prefilter (Orion Depth Filter No. 019R10S, Commercial Filters
Division, The Carborundum Company, Lebanon, Indiana) for
removal of abrasive particulate matter. The pH of the water
was adjusted to 3.5 by addition of HCl solution with a
Johanson proportioner. Aluminum chloride was concurrently
added by proportioner to a final concentration of .005 M.
The conditioned water was then passed across a series of
virus-adsorbing filters consisting of a fiberglas-wound
depth filter (No. K27R10S, Commercial Filters Division,
Carborundum) and a series of three epoxy-fiberglas filter
tubes of 8.0 urn porosity (Grade C filter, Balston, Inc.,
Lexington, Massachusetts).
After collection of each sample, residual sewage was evacuated
from all of the filters with forced air. The orlon prefliter
was then rinsed under pressure with 800 ml of sterile physiologi.
cal saline adjusted to pH=3.5 with HCl. Virus-adsorbing filters
were rinsed with 1,200 ml of an identical saline solution.
Viruses were then eluted from the prefilter with 800 ml of a
sterile .05 M glycine solution adjusted to pH=11.5. Viruses
were eluted from the virus adsorbing filters with 1,200 ml of
an identical glycine solution. The pH of each eluate was
adjusted to approximately 8.5 with sterile .05 M glycine
(pH=1.5) immediately after passage through the filters.
This latter step helped prevent inactivation of viruses at
the high pH. The eluates from the prefilter and virus-
adsorbing filters of each sample were then combined for
reconcentration of viruses.
The virus reconcentration procedures followed were a modified
combination of those suggested by The Carborundum Company
and Farrah et al. (1976). Aluminum chloride was added to
each sample eluate to a final concentration of .003 M. The
pH was raised to 7.0 by addition of 1.0 M sodium carbonate.
A floe formed that trapped or adsorbed viruses and settled
to the bottom of the container. The floe was allowed to
settle for one-half hour and then collected and centrifuged
at 4,000 X g for 20 min. The supernatant was discarded, and
the viruses were eluted from the floe with fetal calf serum
adjusted to pH=11.5 and containing .01 M EDTA. Enough of
the calf serum was mixed vigorously with the floe until a pH
of 9.5 was obtained. This mixture was, centrifuged at 15,000 x g
for 10 min. The virus-containing supernatant was saved and
adjusted to pH=7.4, Hanks' balanced salts were added and
the resulting solution was stored at 70°C until time of
assay. The range of sample volumes was about 50-100 ml.
The procedures used in processing the sediment samples for
viruses were as described by Smith et al. (1978). At the
time of the virus assay the samples were rapidly thawed in a
water bath and chloroform added to 10% v/v to rid the sample
35
-------
of bacteria and fungi. Prior to inoculation onto cells, we
removed a portion of the sample and allowed it to stand at
37°C for one hour to facilitate evaporation of any residual
chloroform; this aliquot was then used as the inoculum.
Buffalo Green Monkey (BGM) cells were used for the virus
assays. Stock cultures of cells were grown for seven days
with the medium changed on the fourth day. The growth
medium was comprised of Minimal Essential Media (Eagle)
with Hanks' salts, L-15 (Leibovitz medium), 1.5 mg/ml NaHCO3,
0.5 mg/ml amphotercin B, 10% fetal calf serum, 100 units/ml
penicillin G and 100 g/ml streptomycin sulfate. Each of
twenty 25-cm flasks containing BGM cells grown for four
days at 37°C was inoculated with 0.2 ml of sample (4.0 ml
total per sample). Viruses were allowed to adsorb to the
cells for 1.5 hr at 37°C prior to overlay with an agar
medium. The overlay medium consisted of Minimal Essential
Media (Eagle) with Hanks' salts, 1.5 mg/ml NaHCOo, 0.5 g/ml
amphotercin B, 100 units/ml penicillin G, .025 g/ml MgCl2,
0.6 percent neutral red (1:300), 1% purified agar and 1%
fetal calf serum. Flasks were incubated for seven days at
37°C, and plaques were counted on days 3 through 7.
DATA REDUCTION
Combined Sewer and Storm Drain Discharges
Pollutant Loading —
Estimates of total loading for each measured parameter for
each outfall for each storm were made with the aid of a
computer program designed to calculate a time integral
product of concentration times flow for each storm hydrograph.
To this end the flow data were entered as values measured at
regular intervals throughout. The quality analyses were
less abundant, being limited specifically to the significant
features of the hydrograph, i.e., the highs, the lows, and
the initial and final ascending and descending limbs.
Between these points, values were interpolated to complement
the flows given at regular intervals. A storm was defined
as a minimum total of .08 cm of rain separated from preceding
and succeeding measurable rainfall by a minimum of three
hours.
Single quality samples contained insufficient particulate
material for chlorinated hydrocarbon analyses. As a result,
sequential samples were composited to represent specific
features on the hydrographs. With the exception of the
final storm monitored at each of the three principal outfalls,
CSO 023, SD 7 and Denny Way, the sediment loading values
were likewise calculated from storm segment composites.
36
-------
For the storm drain, where a dry weather base flow is discharged,
pollutant mass loadings were calculated for both the base
flow and the storm washoff—the latter being determined as
the difference between the total storm load and the base load
estimated from preceding flows and concentrations. Total
rainfall volume on each service area was also determined.
Plume Distributions —
The vertical profiles of light transmission versus depth for
storms monitored were digitized and keypunched for computer
contouring. For contouring purposes, an interpolative contouring
routine was used to fit a bivariate elastic spline surface
to the initially coarse raw data grid, resulting in smooth
contour output. This program was a modified version of a
program described in Numerical Plotting System User's Manual
No. W00053 (University of Washington, August, 1977). Beyond
this, the plots were graphically enhanced for report presentation
Settling Particulates and Sediments
Means, standard deviation, correlation coefficients and analyses
of variance were calculated with the exclusion of measurements
more than three standard deviations from the mean because a few
very extreme values obscured any meaningful comparisons. Two-way
analyses of variance required equal numbers of measurements in
each cell and therefore Monte-Carlo simulation techniques (Wallis
et al., 1974) were used to insert non-biased data points.
Comparisons between station means for sediment core surface
segments were made using Scheffe's procedure for linear contrasts
(Scheffe, 1959). The levels of significance of correlation
coefficients for sediment core segments and sediment trap samples
were found using the Student's t distribution.
The information provided by the sediment particle size distri-
bution analyses was reduced to four values for each sample:
organic content, median particle diameter(M2), sorting coeffi-
cient (S0) and skewness(Sfc). As described by Sverdrup et al.
(1942), the latter three parameters collectively describe
the shape of a size distribution curve.
Benthic Biota
Freshwater Studies —
The primary purpose of the preliminary study done relative
to the benthic biota was to determine the most time- and
cost-effective number of cores to be collected at each
station for biota distribution analyses. To this end, a
complete statistical analysis was performed on chironomid
larvae, oligochaete and copepod counts for stations located
at 1.5, 4.6 and 7.6 m depths. These organisms were
selected on the basis of their high numbers and importance
37
-------
in the littoral food chain.
For each station sampled (20 cores/station), a log (X+l)
transformation was made on the raw counts (Xs), because the
variance (S2) greatly exceeded the mean, indicating a negative
binomial distribution. The 95% confidence limits were
calculated for each n (number of cores) from one to twenty,
using the formula
* * ^05(11-1)
The resultant values of X and the confidence limits were
then transposed back to the original scale and the confidence
limits for values of X representing 1-20 cores were expressed
as fractions of X. The number of cores to be collected per
station was determined as the most useful balance between
statistical reliability and sample handling time.
For the principal project studies, linear regression analyses
were run for the organisms that were either important in the
food chain of fishes, or found in high numbers in our samples
(Nie et al., 1975). For Control Stations 3 and 4, linear
regressions were run using weights and/or transformed counts
(log (X+l)i) as the dependent variable (Y) and the depth of
the sampling station as the independent variable (X). This
was done to establish the strength of the depth effect in a
relatively unaffected or control area. For CSOs, however,
distance from the source of discharge was used as the independent
variable (X). This was done to test for possible effects of
pollution that might be stronger near the discharge source.
For SDs, stepwise multiple regressions were run using weights
or transformed counts as the dependent variable (Y) and
depths and distances as the independent variables (X-L and
X2). With the aid of multiple regression techniques, the
relationships between variables were separated or combined as
necessary to investigate all possible relationships. Multiple
regressions were not run with the CSO because the majority of
sampling stations in each area were the same depth.
When the regression equations were not statistically significant
(a = .05 level), the means of the raw counts and dry weights
were used for comparisons. The predicted values used in
this report represent the average value at each site (with
95% confidence) that would be obtained by multiple sampling
at each site. This provides a statistical measure of the
increases or decreases in the numbers and biomass of benthic
organisms relative to depth and distance from the study
outfalls.
Correlation coefficients (r), coefficients of determination
(r^), and the significance level of the regression equations
38
-------
were used as measures or indicators of the strength of the
relationships between the count and weight data vs. the
depth and distance factors. The "r" value indicates the
"degree of association" between variables (Zar, 1974) ., An r of
1.00 or -1.00 indicates a perfect association between variables
whereas an r value of 0.00 indicates no association whatsoever.
The value of r2 is the fraction of the total variation in Y
(counts or weights) that is explained by the fitted regression.
Marine Studies —
Species-richness curves (Hurlbert, 1971) were calculated on
the combination of data from all replicates at a site for
the periphyton, boulder wall algae, subtidal polychaetes and
subtidal molluscs. The multivariate technique of discriminant
analysis was used to statistically evaluate sites using
periphyton and boulder wall community data. Discriminant
analysis weighs and combines the discriminating variables
(i.e., species counts) in such a way as to force the groups
to be as statistically distinct as possible. The species
that are most important in discriminating among sites are
indicated by tests of significance based on a linear combination
of all species values. A computer program (BMD07M, Dixon,
1973) performed this analysis. The part of the output used
consisted of (1) the mean and standard deviations of
species values at each site, (2) the proportion of variation
explained by the canonical variables associated with the
analysis, (3) the rank order of species according to their
entrance into the discriminant function, and (4) a plot of
the samples along canonical variables one and two.
Classification analysis was used to compare subtidal samples
within each season. After the dissimilarity of the sites
(Bray-Curtis coefficient, Clifford and Stevenson, 1975) was
calculated based on square root-transformed species-abundance
values, the sites were clustered using the group-average
sorting strategy (Sokal and Michener, 1958). The 20 most
abundant species were used in the analysis each season. The
results of this analysis are presented in the form of dendrograms
The proportion of individuals within the phyla Arthropoda,
Mollusca, and Annelida at each of the subtidal sites was
determined and graphically compared using the methods of
Snee (1974). The proportions of feeding types (after Jumars
and Fauchald, 1976) of polychaetes at each site were also
compared using this method.
39
-------
SECTION 5
RESULTS
FRESHWATER STUDIES
Preliminary Studies
To aid in the selection of CSO and SD sites for more intensive
study, levels of contaminants in near-outfall sediments were
assessed for the Preliminary Study stations (Figure 2). A
summary of the resultant data is presented in Appendix A,
Table A-l. in all cases, the samples represent surface sedi-
ments collected by divers at the near-outfall locations thought
to have the maximum potential for contamination by the discharge
particulates. y
Figure 8 represents an attempt to composite the information
presented in Table A-l. The "sediment enrichment factors"
are given for three categories of parameters: (Inorganics
(total organic carbon plus oils and greases) and total phos-
phorus, (2) metals (Cu, Pb, Zn, Hg), and (3) chlorinated hydro-
carbons and PCBs. To calculate these relationships mean values
were derived for each parameter at each station, and the
resultant figures were normalized by parameter, i.e., the lowest
value found for each parameter was divided into each of the cor-
responding values from the other 28 stations. These figures
were then added for each station for each category, and the
sums were in turn normalized within the categories!, Thus, the
total length of the three histogram bars at each station divided
by 3.0 would give the overall enrichment of that station
compared to a hypothetical composite of the most pristine
sediments found in the lake. The sediments from CSO 046,
which does not appear to overflow, approach this quality
overall. The actual background standards used for the Figure 8
calulations were: organics and total phosphorus - Station
CSO 046, metals - Station C 10, and chlorinated hydrocarbons
(ClHCs) and PCBs - Station C 3. These values were all comparable
to or lower than those determined for pre-1900 core segments
collected in the deep portion of the lake by Birch (1976) and
Spyridakis and Barnes (1976). In terms of low concentrations,
Station CSO 046 was also second in both metals and ClHCs and
PCBs.
40
-------
41
-------
Sediments around the sewer outfalls on the eastern shore of
Lake Washington generally appeared to be only one-third to
one-fourth as contaminated overall as those between the two
floating bridges and south of Seward Park (denoted by C 4 and
C 5) on the western shore. Several stations in this area had
sediment enriched more than 16 times over background levels.
The area between the bridges had particularly high concen-
trations of metals, which are typically associated with both
CSOs and SDs. It should be noted that the pipe at Station
SD 6 has separated in two places and that only a portion of
the wastewater plume is emitted from the end, where the
samples were taken.
On the western shore, chlorinated hydrocarbons (predominantly
DDT) were widespread and up to 37 times background concen-
trations. PCBs were found only at Stations SD 11 and SD 13
in concentrations of 118.6 and 49.8 ppb dry weight, re-
pectively.
The sediments in the northern third of the lake were relatively
clean, with the maximum enrichment being about three times
background.
The storm drains around Mercer Island had created some
problems. Sediments sampled at a 3 m depth just off SD
17 (SD 17-3 m) were over twice as contaminated with
metals and DDT (75 times background levels)_ as any others
found in the lake. These sediments had been enriched by
particulates from a drainage system that serves most of the
Mercer Island business district. They were rapidly filling
in the portion of the bay around the outlet channel; aerial
photos of the area taken on February 6, 1978 show a large
surface film of oil or gasoline there.
Analyses of sediment samples collected at the mouth of the
drainage channel (SD 17) indicate that most of the contaminants
were swept a little further offshore before settling. These
observations are in accord with dye circulation studies done
at this site by CH2M/Hill (1973), who found the circulation
in the bay to be dominated by a large eddy—a pattern which
would tend to promote sedimentation. Unfortunately, the
shoaling was found to be so extensive that the sediments
were stirred by propellers and boat hulls, thus precluding
more intensive study as part of the present project.
Two other Mercer Island storm drains—SD 19 and SD 20—were also
associated with sediments that were higher in DDT than those
at any site sampled on the eastern lake shore. Storm Drain 19,
which enters the lake at the northern edge of a large public
swimming area, was subsequently selected for more intensive
study. SD 20 drains a large area (> 200 acres) in the center
of the island and runs into the lake next to a private dock.
42
-------
Over a ten-year period it has built up a sediment delta next
to the dock that extends approximately 20 m into the lake.
As a result the landowner has been obliged to build a second
pier for boat moorage. It should be noted that the metals
contamination of these sediments was also found to be equivalent
to or greater than at any station sampled on the eastern
shore.
In order to define the nature and strength of any relationships
that might exist among the various sediment parameters,
correlation matrices were calculated for the data grouped by
station type (CSOs, SDs, controls). The results are presented
in Tables 1-3. The square of the correlation coefficient
for any two parameters indicates that fraction of the total
variation in the dependent variable Y that is explained by
fitted regression; confidence levels have been denoted in
the tables for the most significant correlations.
It is apparent that the greatest number of significant cor-
relations was found for the SD sediment parameters. In this
sense the CSO sediments were only slightly less diverse than
the control sediments, which is surprising considering the
partial stormwater content of the CSO effluents. The most
closely correlated parameters were: Cu/Zn (SD) , r = .888;
Cu/Zn (CSO), r = .887; Pb/O&G (SD), r = .870; Cu/Zn (control),
r = .841; TOC/O&G (CSO), r = .835; TOC/C1HC (SD), r = .821;
Zn/TPO4 (CSO), r = .809; Hg/ClHC (SD), r = .768; and O&G/C1HC
(SD), r = .767.
Thus, it may be seen that Cu and Zn were strongly related in
all sediments studied in the lake, indicating common source(s)
and possibly extensive longshore circulation. Recent studies
done by Metro (unpublished data) have shown the primary
source of these two metals in Seattle's combined sewers to
be domestic plumbing; their origin in storm drainage is less
well defined. The strong relationship between Pb and O&G in
the storm drain sediments was not unexpected, motor vehicles
being the prevalent source of both; this relationship was
also present (but weaker) in the CSO sediments. TOC and O&G
also had multiple significant correlations with other parameters
at the CSO and SD stations.
It is of interest to note that chlorinated hydrocarbons (ClHCs)
were significantly correlated with various other parameters
measured in SD sediments (TOC, Hg, O&G, Pb and Zn), whereas
CSO sediments showed virtually no evidence of such relationships
This would seem to indicate that the ClHCs enter the combined
sewer wastes predominantly in the storm water fraction, a
supposition that is commensurate with the fact that the
primary component of the C1HC values was DDT, which until
recent years was a widely-used pesticide.
43
-------
TABLE 1. CORRELATION MATRIX FOR COMBINED SEWER SEDIMENTS.
^ "a Pb Zn TOC Q & G C1HC 3K?4_
Al -.330 (10) -.345 (10) .185 (10) -.324 (10) -.529 (9) -.357 (9) -.279 (10)
Cu 0255 (18) -.051 (18) .887*tl8) -.088 (17) .115 (16) ^239 (8) .588*1.18)
Hg -022 d8) .414*(18) -.029 (18) .123 (17) .426 (9) .439*(18)
Pb -014 (18) .224 (17) .491*(16) .184 (8) .014 (18)
Zn -.152 (17) .093 (16) .106 (8) .809**18)
TOC .835*tl6) .515 (9) -.022 (17)
° & G - .592 (8) .207 (16)
C1HC .609 (8)
* 95% confidence level. ~~~~ ~ ——
** 99% confidence level.
0 Number of samples.
TABLE 2. CORRELATION MATRIX FDR STORM
£u Hg pb zn TOC O & G C1HC TPO4
Al .326 (13) -.347 (11) -.070 (13) .100 (13) -.270 (11) .061 (12) .253 (13)
Cu .211 (21) .478*(23) .888**23) .327 (21) .561**22) .519 (10) .497**23)
Hg -069 (2D .239 (21) .545*t21) .292 (21) .768*tlO) ,514**21)
Pb .481*(23) .167 (21) .870**22) .688*(10) .039 (23)
Zn -162 (21) .549*t22) .602*(10) .355 (23)
nrv-v-1
^ .217 (22) .821*tll) .584*t22)
O & G **
.767*tll) .245 (23)
C1HC *
•539 dD
* 95% confidence level. ~ ~
** 99% confidence level.
() Number of samples.
TABLE 3. mERRr-ATiON MflTRIXFOR CONTROL SEDIMENTS
— ^2 Pb Zn TCC Q & G C1HC TPO
Al .273 (10) .418 (10) .223 (10) .639*110) .400 (10) -.087 (10) ,416 (10)
Cu -.024 (19) .299 (21) -.841**21) .279 (19) .243 (20) -.036 (9) .521**21)
Hg .391*(19) .146 (19) -.280 (20) -.280 (20) .157 (10) .574**19)
Pb -I23 (21) .086 (19) .173 (20) .436 (9) .429*(21)
•I54 (19) .137 (20) -.050 (9) .478*(21)
TOC
.035 (20) .186 (10) -.259 (19)
0 & G
.244 (10) -.051 (20)
C1HC
:z=z———-—— . -.096 (9)
** 99% confidence level.
() Number of samples.
44
-------
Intensive Studies
Discharge Monitoring —
Discharge Loading Estimates for Single Events — The combined
sewer overflows and storm drainage generated by several storms
were sampled quantitatively and qualitatively at CSO 023 and
SD 7 in order to determine typical particulate inputs to the
freshwater environment by a representative CSO and SD. These
systems were selected in preference to CSO 044 and SD 19 because
they were easier to sample. Table 4 is a summary of the storms
monitored.
TABLE 4. SUMMARY OF RAINSTORMS MONITORED FOR
QUANTITY AND QUALITY OF DISCHARGES
INTO LAKE WASHINGTON FROM COMBINED
SEWER OUTFALL 023 AND STORM DRAIN 7,
MARCH-SEPTEMBER, 1978
Date
Station
Total Overflow Discharge Discharge
Rainfall Duration Volume Fraction
(cm) (hr) (m3)
3/6
3/24
4/15
8/31
- 3/7
3/23
- 3/25
- 4/16
- 9/2
SD 7
CSO 023
SD 7
CSO 023
SD 7
CSO 023
SD 7
CSO 023
SD 7
1.75
0.43
0.43
0.74
0.74
2.67
2.67
2.11
2.11
23
2
5
23
29
20.5
21
38.5
32.5
1001
53
213
51
835
951
1849
663
567
10.8
2.5
9.3
1.3
20.8
7.4
13.4
6.7
4.4
(a) The fraction of the total basin-incident rainfall that was
discharged by the outfall during the period noted.
For the Seattle region the months represented include the
end of the rainy season and most of the dry season. Because
the drainage basins for the two stations sampled were only
2.4 km apart, with similar orientation and no intervening
ridges, a single rain gauge was used for both.
The similarities in relative size of the CSO 023 and SD 7
drainage basins and incident rainfall for any given storm invited
comparisons of their pollutant loading and runoff characteristics
45
-------
The major distinction affecting the relative quality of runoff
for these two areas was land-use distribution; the CSO 023
basin was comprised of approximately 75% single family residences,
with the rest multiple family residences, whereas the SD 7 basin
had an estimated 90-95% of its area devoted to single family
residences (normally associated with lower pollutant accumulation
rates than multiple family residences). For three of four storms
an appreciably lower percentage of the total rainfall on its
drainage basin was discharged by CSO 023 than by SD 7; this was
to be expected, since a CSO discharge is typically intermittent,
whereas that of an SD is continuous. On the other hand, the
relative magnitudes of the mean storm discharge concentrations
for the various pollutants varied from one storm to the next,
as will be discussed below. A statistical summary of the
contaminant mass discharges is offered in Table 5.
Mean ratios (SD 7: CSO 023) of per-storm mean concentrations
and total loadings are also presented here for each parameter,
in support of the ensuing discussion (Table 6). Whereas the
data from four storms were available for calculation of the
values given in Table 6, those from only two storms were so
used. The annual cycle of SD 7: CSO 023 ratios of per-storm
discharge volumes indicated that the SD is disproportionately
affected by groundwater inputs during the wet season; therefore
it was decided to eliminate two of the three wet season storms
to reduce the bias in the calculation of annual mean pollutant
concentrations. Mercury, aluminum and organic carbon data were
not available for both of the storms used and consequently no
annual means were calculated for these parameters.
Particulate Discharge — With reference to Table 6, the SD 7:
CSO 023 ratios of mean particulate concentrations were appreciably
less than 1.0 for suspended solids, Cu, Zn, total P, and
chlorinated HC. Only Pb had a significantly higher solids
concentration at SD 7, this at least partially due to dilution
by low-lead sanitary wastes at the CSO. However, the mean
discharge volume for SD 7 was 1.5 times that of CSO 023, and
consequently the mean storm particulate loadings for suspended
solids, Pb, and Zn at SD 7 were greater than those at CSO 023.
The mean loading values for C1HC were similar for the two sites,
giving a SD-.CSO ratio near 1.0.
For the three storms monitored at both stations (Table 5), there
were consistently higher mass loadings of Al from SD 7 than from
CSO 023, indicating a greater input of inorganic terrigenous
materials to the lake by the storm drain. The percentage of the
total Al found to be associated with particulate matter was high,
invariate and nearly identical (approximately 96%) for the two
outfall sites, implying that Al was present in the overflows
predominantly as the relatively insoluble aluminosilicates, i.e.,
inorganic terrigenous material. As such it was used as an
inorganics tracer in the subsequent water column and sediments
46
-------
TABLE 5. SUMMARY OF ESTIMATED TOTAL POLLUTANT
LOADS AND FRACTION OF PARTICULATES
IN STORM DISCHARGES MONITORED AT
COMBINED SEWER OUTFALL 023 AND
STORM DRAIN 7, MARCH-SEPTEMBER, 1978
CSO 023
MASS RANGE:
kg
No. of
Storms
SD 7
kg
No. of
Storms
Suspended Solids
Total Cu
Total Hg
Total Pb
Total Zn
Total Al
Total Organic C
Total TP04-P
Total 0 & G
Particulate ClHC(b)
6.40 -
.001 -
.0000 -
.001 -
.003 -
.059 -
.865 -
.059 -
.584 -
.316 -
142
.039
.0008
.041
.113
2.56
18.0
1.28
9.41
100 mg
4
4
4
4
4
3
2
4
4
4
27.4 -
.020 -
.0000(a)
.098 -
.044 -
.762 -
11.0 -
.120 -
.740 -
1.02 -
142
.049
.0004
.316
.156
6.62
16.2
.417
8.06
86.1
5
5
5
5
5
4
2
4
5
5
(c)
Percentv 'of No. of
MEAN PARTICULATE MASS: Total Mass Storms
(c)
Percent of
No. of
Storms
Suspended Solids
Cu
Hg
Pb
Zn
Al
Organic C
TPO4-P
O & G
C1HC
100.0
78.1 + 10.2
88.4 + 6.9
69.1 + 14.2
68.6 + 19.6
96.0 + 2.0
42.1 + 22.9
28.5 + 10.3
NA(d)
ND(e)
4
4
2
4
4
3
2
4
100.
64.0 +
ND^
88.1 +
64.1 +
96.6 +
56.0 +
52.7 +
NA
ND
0
8.7
a)
2.7
5.0
0.9
8.1
13.1
5
5
5
5
5
4
2
4
(a) Below detection limits.
(b) Selected chlorinated hydrocarbons: a-BHC, lindane,
heptachlor, heptachlor E, aldrin, dieldrin, endrin
and DDT (ODD + DDE + o,p DDT + p,p DDT).
(c) x + la.
(d) Not applicable.
(e) Not determined.
47
-------
TABLE 6. INTERSTATION COMPARISON FOR STORM
DRAIN 7 AND COMBINED SEWER OUTFALL 023
OF MEAN DISCHARGE CONCENTRATIONS AND
POLLUTANT LOADINGS ESTIMATED FOR TWO
STORMS
-------
studies.
There was a large variation from one storm to the next in the
mean concentrations and total mass loading of chlorinated
hydrocarbons at both sites (Table 5). Trends at the two
stations were similar in that DDT accounted for a very large
portion (approximately 75-95%) of the total loading of
particulate pesticides during the spring storms, and only
about half of the August-September storm total, which was
itself appreciably lower than that of the other storms.
Lindane was the only other pesticide consistently present at
both stations. The persistent occurrence of DDT and lindane
conforms to the observations of Brenner et al. (1978) for
runoff from the Juanita Creek watershed in the northeastern
segment of the Lake Washington basin. The occurrence of these
contaminants in both the CSO and SD systems indicates that they
were present in the storm drainage, rather than in the sanitary
wastes. Low concentrations of two other pesticides, benzene
hexachloride (BHC) and aldrin, were also detected in the fall
storm discharges; lindane, BHC and aldrin levels in the
sediments were comparable at the two stations.
Total Discharge — With two exceptions, the SD:CSO ratios
calculated for mean pollutant concentrations in the unfiltered
(total) storm discharges were similar to those estimated for
the particulate fractions alone. The ratios for Pb and TP04~P
in the total discharges were notably lower, indicating that in
comparison to the CSO analysis, an appreciably larger fraction
of the Pb and TPO4-P in the SD discharges was determined to be
particulate. This observation agrees with the data presented
in Table 5 for the larger storm set, values which include the
above-mentioned wet weather bias.
Viruses — The results of virus analyses performed on CSO 023
and SD 7 discharges are summarized in Table 7. As evidenced
by these data, it seems unlikely that there are any human
viruses present in typical storm drain discharge. This
finding is logical in view of the expected absence of human
fecal wastes in storm drainage. With rare exceptions, human
viruses only infect, and are excreted by, humans. Therefore,
their presence in a storm water system would be indicative of
its contamination by human fecal wastes from leaking sanitary
sewers, septic tanks or landfill runoff. Whereas these conditions
do occasionally occur, they represent deviations from the norm.
On the other hand, viruses were readily detected in the
discharges of CSO 023. From the substantial difference
in the two estimates made for that system, one is led to
consider those factors which may characteristically influence
the virus levels of combined wastewater. Generally, the
virus concentrations are a function of the concentrations of
fecal matter in the municipal wastewater; this in turn is
49
-------
dependent upon two variables—rate of input and dilution within
the system.
Total Recovery
Overflow Vol. Efficiency
Station
CSO 023
SD 7
Date
7/16/78
10/23/78
9/22/78
11/08/78
(m3)
454
625
2120
16
No. Viruses/m3
1.32xl06
0.21xl06
0
0
(%)
2.1
3.4
9.4
1.1
The rate of input of fecal wastes typically varies with time
of day. If an overflow occurs at a time of day when toilets
are heavily used, the result will be a correspondingly high
concentration of feces (and therefore viruses) in the combined
discharge. Flow data from Seattle Metro's wastewater treatment
plants indicate two principal peaks during rainless days,
one occurring at about 6-11 a.m., and the other at about 4-8
p.m. With reference to Table 7, the sample with an estimated
1.32 x 106 viruses/m3 was pumped from the CSO 023 system at 10
a.m., during the period of highest morning inputs. The sample
containing 0.21 x 10° viruses was collected at 11 p.m., well
after the evening peak.
The other principal factor influencing virus concentrations
is dilution within the sewer system. During an intense rain
the proportion of storm water in the combined sewage would
be greater than during a lighter storm. Thus, given equal
fecal loading to the system, virus concentrations might be
expected to be higher during the latter event. It is important
to note, however, that the total virus loading to the receiving
waters is basically a function of total overflow volume, and
in this sense the dilution effect of a more intense storm
tends to be compensated by a correspondingly greater discharge
volume. Consequently, the most influential factor regarding
virus loading is probably the time of day at which an over-
flow occurs.
50
-------
Rainfall, Flow and Loading Summary — The total Seattle rainfall
for the March, 1978 - February, 1979 study period was near the
annual average. The rainfall at the Denny Way combined sewer
outfall totalled 98.8 cm (38.9 in), whereas a 30-yr mean
measured for a site just 2.7 km to the SE and at a similar
elevation (Phillips, 1968) was 86.6 cm (34.1 in). The 1978
spring and summer were somewhat wetter than usual, but the 1978-
1979 winter was drier, compensating for most of the observed
difference.
The rainfall records for the CSO 023 - SD 7 - CSO 044 study area
show a substantial variance with the aforementioned totals
for the downtown area. The total rainfall recorded for this
portion of the western shore of Lake Washington was 58.9 cm
(23.2 in), a value confirmed by three separate gauges (near
CSO 020, SD 7 and CSO 046 in Figure 2). It seems probable that
this annual difference of 28 cm of rain represents a shielding
effect, or rain-shadow, with the hills of the city blocking
part of the precipitation from the storm fronts, which
typically come from the southwest.
The protected areas on the northeast sides of the region's
hills are apparently quite restricted - a gauge located ap-
proximately 1.3 km east of Mercer Island's easternmost
promitory collected 90.0 cm (35.8 in) of rain during the study
period, indicating very little protection from the island's
90-120 m hills. No rainfall data were collected for SD 19,
but on the basis of its location on the eastern side of
Mercer Island, it seems likely that the total for the one
year study period was less than the 90.9 cm (35.8 in) recorded
across the channel.
The overflow response to the indicated rainfall pattern is
shown in Figure 9. The values given for SD 7 represent flow
above baseline. For a total of 62 periods of measurable
rainfall (separated by one or more dry days), there were 48
overflows at each of the two stations, CSO 023 and SD 7.
Taking into account the fact that multiple overflows occurred
during some of the long storms, the fraction of those storms
resulting in increased discharge at one or both stations was
76%; the average total rainfall for those storms that failed to
generate increased discharge was 0.1 cm (.04 in).
The total discharge volumes for the 12-month monitoring period
were: CSO 023 - 25,600 m3 (6.76 MG); SD 7 - 27,700 m3
(7.34 MG), + 50,200 m3 (13.3 MG) base flow. In combination
with the mean discharge concentrations given in Table 5 and
with concentrations measured for non-storm discharge at SD
7, these values were used to calculate total annual loads
for the various parameters of interest. The results are
presented in Table 8.
51
-------
3 >
^ o
0 <
oc
CO
MOTdbiAOIVlOl
52
-------
TABLE 8. ESTIMATEDva; TOTAL ANNUAL MASSES
OF SELECTED CONSTITUENTS IN
DISCHARGES FROM COMBINED SEWER
OUTFALL 023 AND STORM DRAIN 7—
MARCH 3, 1978 TO FEBRUARY 28, 1979
Parameter
CSO 023
Storm (Total)
(kg)
Storm
(kg)
SD 7
Non-Storm
(kg)
Total
(kg)
Suspended Solids
Total (b) Cu
Particulate Cu
Total Pb
Particulate Pb
Total Zn
Particulate Zn
Total TPO4-P
Particulate TPO4-P
0 & G
Particulate C1HC
-------
For all constituents except Pb, the annual storm loading of
particulates was less from the storm drain than from the
combined sewer. However, with the non-storm loading of the
storm drain added to its storm inputs, the totals for suspended
solids and Cu also exceeded those of the combined system.
The parameter relationships were similar for the total
(particulate + soluble) loading, except that the total Zn
loading from the two systems was essentially equal. The annual
mass of Cu, Pb, Zn and TPO4-P discharged by the CSO were all
60-65% particulate; the relative values at the SD varied
considerably being 67%, 83%, 55%, and 41%, respectively.
The difference in the percentage of particulate P estimated for
the two stations is undoubtedly due to the presence of high-P
sewage particulates in the combined system. The relative
difference recorded for Pb, however, implies the influence of
some process in addition to dilution by low-Pb municipal wastes
in the CSO. The presence of an unidentified source of high-Pb
particulates in the SD drainage basin seems unlikely. A more
reasonable conjecture is that a large fraction of the compara-
tively heavy, Pb-bearing particulates bypasses the overflow
barrier in the CSO system; by contrast, no such bypass exists
in the SD, so that a greater portion of the heavy particulates
are routinely discharged.
The total mass of solids discharged by the two outfalls
during the one year study period was 3.2 metric tons by CSO
023 and 3.3 metric tons by SD 7. Based on the respective
settling rates of .752 and .805 g/m2/day measured by the
sediment traps, a uniform blanket of this material would
cover 1.4 hectares (3.5 acres) at CSO 023 and 1.0 hectares
(2.4 acres) at SD 7. However, the biologically-affected
area around CSO 023 (based on oligochaete counts) was de-
termined to average 0.6 hectares (1.4 acres) for the two
seasons sampled, indicating that more than 60% of the solid
material was rapidly transported away from the outfall.
This fraction would be appreciably larger at SD 7; in fact,
so much of the solid material moved away from the
outfall and out of the sample grids that a comparable estimate
for the storm drain was not feasible. Such estimates are,
of course, very rough due to the extreme variability of
settling dynamics as functions of distance and direction
from an outfall.
Receiving Water Monitoring --
Turbidity Distributions of Discharge Plumes — The distributions
and movements of discharge plumes was monitored at CSOs 023 and
044 and SDs 7 and 19 during a number of storms, as summarized
in Table 9. These measurements were made using a submersible
light transmissometer system of the type originally described by
Petzold and Austin (1968). This instrument provides a
measurement of the volume attenuation coefficient, a, which
54
-------
is equivalent to the sum of the light scattering and absorption
coefficients. The volume attenuation coefficient may in
turn be related to common turbidity units using methods
described by Austin (1973), who determined the relationship
between <* and Jackson Turbidity Units (JTU) to be approximately
linear for natural water of moderate to good clarity.
TABLE 9. SUMMARY OF FIELD TRIPS FOR MEASURING TURBIDITY
DISTRIBUTIONS OF DISCHARGE PLUMES
Overflow Period D
Monitoring Period g
Time of Day
Date
Station
0000
0600
1200 1800
Overflow
Volume Rainfall
2400 (m3) (cm)
3/6/78 - 3/7/78 CSO 023
SD 7
C3
3/24/78 - 3/25/78 CSO 023
SD- 7
C 3
4/4/78 CSO 023
SD 7
2/6/79 CSO 023
SD 7
2/11/79 CSO 044
2/12/79 SD 19
C 3
C 4
1
•
•
•
C
•
•
i — a
1
i
•
^_
i
i
'.=
i 1
HT
J
— ™^^— — |
!)• •
303 1.75
1130 1.75
NA 1.75
56 1.02
711 1.02
NA 1.02
62 .46
201 .46
156 .99
299 .99
ND(fa) 2.13(c)
ND 2.2l(°)
ND 3.15(c)
ND 2.36(c)
(a) Not applicable.
(b) Not determined.
(c) Rainfall up to and including period of monitoring.
The other values given cover the entire overflow period.
For the purpose of plume tracing, the light transmission
(turbidity) data were reduced to sectional contour plots.
These included aerial perspectives at various depths, and
onshore-offshore and longshore transects showing light
55
-------
transmission patterns from the perspective of a diver looking
across the lake bottom. Representative plots were selected for
presentation here to show observed trends of participates move-
ment. In addition to data from the four outfall areas, the
results of storm measurements from the two control sites (C 3
and C 4) are also included.
Contrary to intentions, we were unable to obtain light trans-
mission data for periods of lake stratification. The data
presented here were collected during March and April of 1978 and
February of 1979 and in all instances show the discharge plumes
to have been intersecting the surface. At the two combined
sewer outfall sites (CSO 023 and CSO 044) this was due pre-
dominantly to the comparatively high discharge temperatures, e.g.
a difference of approximately 6°C was measured for a comparison
of discharge and ambient water at CSO 023 on 4/4/78. For the
two storm drains (SD 7 and SD 19) the shallow discharge depths
(2 and 0 m, respectively) were the foremost determinants.
CSO 023 — The various characteristics of the discharge plume
dispersion observed at CSO 023 are summarized in Figures 10-14.
Figure 10 shows the residual light transmission distributions
21.5 hr from the time the 3/7/78 overflow began. The plume was
seen as a lens of turbid surface water 150 m NE of the outfall,
2.7 m thick (Figure lie) and covering an area of approximately
4.5 hectares (11 acres). The extremely muddy water SW of the
outfall was temporarily uncontrolled drainage from a_nearby
building site (Figure lla). For the measured CSO discharge of
303 mj, the size of the discharge plume at that time represented
a dilution of approximately 400:1. The plume of 4/4/78,
measured 3 hr into that overflow (Figure 12), covered only 0.6
hectares (1.5 acres), but represented a similar estimated
dilution of 615:1. The dilution calculated for a lens of turbid
surface water found at point C (Figure 10) on 3/24/78 - 3/25/78,
however, had an appreciably higher value - 2000:1 - only 3.5 hr
into the overflow period.
This considerable discrepancy, in view of the comparatively good
agreement between the other two examples, implies an unaccounted-
for volume contribution from an additional source: the large,
nearshore storm drain indicated in the figure. Interference
from that system was much more obvious during the storm of
2/6/79 (Figure 13); a contour plot of an onshore-offshore
transect in that location (Figure 14) clearly indicates a tongue
of turbid wastewater moving offshore from the drain site. The
location of the 156 m3 of discharge known to have been released
by CSO 023 was obscured by the storm drainage.
The surface position of a discharge lens as a function of
time gives an incomplete indication of particulate circu-
lation patterns. Much of the solid material present in
56
-------
COMBINED SEWER OUTFALL 023
BUILDING UNDER
CONSTRUCTION
10 feet (3.0 meters)
Figure 10. Aerial perspectives of contours of percent light transmission at
Combined Sewer Outfall 023—1533-1645 hrs., 3/7/78. The lettered
symbols indicate surface positions of the discharge plume core
at time of overflow initiation plus (a) 5 hr, 3/6/78; (b) 21.5
hr, 3/7/78 (distribution shown); (c) 3.5 hr, 3/24/78-3/25/78;
and (d) 3 hr, 4/4/78.
these features apparently remains suspended for extended
periods of time, as evidenced by the high light transmission
(lack of turbidity) at mid-depths in Figure 11, and is
ultimately advected out of the area. However, Figures 12 and
13 indicate that a substantial local fallout and settling of
57
-------
1-rvfV--- • •
•/••.«•
,'*(b) OUTFALL LINE
50 15.2
Figure 11.
longitudinal sections of contours of percent light transmission
at Combined Sewer Outfall 023—1533-1645 hrs. , 3/7/78. The
perspective is that of a diver facing shore.
58
-------
Figure 12.
15.2
Longitudinal section through the
Combined Sewer 023 outfall, showing
contours of percent light transmission—
0856-0954 hrs., 4/4/78. The perspective
is that of a diver facing the shore.
particulates occurs in the outfall area during the overflow.
The settling pattern denoted by Figure 13 indicates a downslope
movement of materials near the bottom (refer to Figure 3 for
CSO 023 bathymetry). The observed combination of suspended
transport and downslope settling was also quite evident at SD 7
and SD 19, as described below. These observations correlate
well with the measured distributions of sediment surface con-
taminants presented elsewhere in this report.
SD 7 -- Considerable discharge interference was also observed
in the light transmission distributions measured for storm
discharges at SD 7. Substantial inputs from a 60 cm storm
drain 345 m north of the SD 7 outfall were recorded for the
storm of 2/6/79. The relevant areas of influence for the
two outfalls are strikingly apparent in Figures 15 and 16,
which give aerial light transmission distributions at one and
four hours elapsed time following a 10-hour storm discharge at
SD 7. During the three-hour interim period, the SD 7 plume
began to dissipate, and the pattern of discharge from
the other outfall became more prevalent. The small cove
between the two outfalls appears to have encouraged a local
eddy in the nearshore circulation, resulting in a southerly
movement of discharge offshore. There were also manifestations
of this clockwise circulation pattern, and of particulates
expelled by the second storm drain, in storm data collected
3/6/78 - 3/7/78.
59
-------
48 INCH STORM DRAIN—r
Figure 13.
40 feet (12.2 meters)
Aerial perspectives of contours of percent
light transmission at Combined Sewer
Outfall 023—1330-2330 hrs., 2/6/79
60
-------
DEPTH
Figure 14. Transverse section through a point 34 m S of
Combined Sewer Outfall 023 showing contours
of percent light transmission—1330-2330
hrs., 2/6/79. The perspective is that of a
diver with the shore to his left.
Offshore settling of discharge solids was also observed during
these storms. Figure 17 gives a time sequence of light trans-
mission distributions for the onshore-offshore transect
61 m north ("downstream") of the SD 7 outfall. The first distri-
bution was recorded immediately following cessation of the
2/6/79 storm discharge; the second was recorded three hours
later. The dissipation and settling of particulates offshore
is evident. An even more graphic portrayal of this effect
is given by Figure 18, which represents the light transmission
distribution along a transect 283 m farther north, and
immediately offshore of the second storm drain, at the time of
the distribution in Figure 17b. Such patterns, with turbid
plumes moving offshore along or near the bottom were observed
frequently during the several storms monitored at SD 7.
Control Site 3 — The C 3 area was found to be inundated with
suspended particulate matter from a variety of sources.
Light transmission contours drawn for storm data collected
on 2/12/79 show the local response to an estimated 3.15 cm
(1.24 in) of rain (Figures 19 and 20).
61
-------
20 feet (6.1 meters)
J
•34-^
38.
•36.
40 feet (12.2 meters)
Figure 15. Aerial perspectives of contours of percent
light transmission at Storm Drain 7—1915-2014
hrs., 2/6/78.
62
-------
30m W
A _OUTFALL
LINE
>15m E
j 46m E
107m E
40 feet (12.2 meters)
Figure 16. Aerial perspectives of contours of percent light
transmission at Storm Drain 7—2244-2348 hrs.,
2/6/79.
63
-------
(a) One hour after cessation of storm discharge by
SD 7 (1915-2014 hrs).
DEPTH
(b) Four hours after cessation of storm discharge by
SD 7 (2244-2348 hrs).
Figure 17. Transverse sections through a point 61 m N
of Storm Drain 7, showing contours of percent
light transmission on 2/6/79. The perspective
is that of a diver with the shore to his left.
64
-------
Figure 18. Transverse section through a point 344 m N of
Combined Sewer Outfall 023, showing contours
of percent light transmission—2244-2348 hrs.,
2/6/79. The perspective is that of a diver
with the shore to his left.
An uncharted source of storm drainage was indicated at the
surface, inshore and 70 m N of the center of the sediment
trap array. Considering the large volume of rain incident
on the adjacent hillside, and the mild response indicated,
the drainage system implicated must have been small. The
sampling was done approximately midway through the 36-hour
storm period.
A second major input of particulates is evident below the
9 m depth in Figure 20, and in the 15 m section in Figure 19
The presence of an extensive turbid layer moving at
depth through a comparatively natural area of the lake was
unexpected, and would seem to constitute a large-scale
phenomenon. Although the circulation in this area was found
to be complex, with multiple surface drift observations in-
dicating intermittant flow to the southwest and northeast,
the prevailing flow would necessarily be southwesterly,
carrying basin runoff toward the Lake Washington outlet to
Puget Sound (Figure 2). Accordingly, the deep, turbid layer
observed at C 3 on 2/12/79 is believed to have been a long-
shore flow of settling particulates moving from the northeast.
The most likely sources of this material were (1) a series
of 22 storm drains (including eight > 60 cm diameter) carrying
runoff from an area of active construction 0.7-3.8 km to the
65
-------
north, and (2) Thornton Creek, which discharges highly
turbid runoff from a 3130-hectare (7740-acre) basin another
0.4 km beyond that (Figure 2).
(/>
in
M
CM
CENTER OF
SEDIMENT TRAP
ARRAY
SURFACE
8m W
SED. TRAP
ARRAY
423m E
53m E
84m E
Figure 19. Aerial perspectives of contours of percent
light transmission at Control Site 3—0506-0549
hrs., 2/12/79.
The nature of turbid longshore advection through the C 3 area
seemed to be quite variable—an extensive particulate layer
observed entering from the north on 3/25/78 was on the lake
surface, whereas that mentioned above was at depth. Regardless,
the sediment quality analyses done in the course of this
project indicated only light settling of contaminated particles
in this area, and the measured settling rate was substantially
lower than for the CSO 023 and SD 7 areas. These concepts are
discussed in more detail below.
CSO 044, SD 19 and Control Site 4 — A second set of three
nearshore areas was monitored during a single storm to
provide support data for the biological studies and for the
66
-------
a) 245m N of sediment trap array
(b) 70m N of sediment trap array.
Figure 20. Transverse sections showing contours of percent
light transmission at Control Site 3—0506-0549
hrs., 2/12/79. The perspective is that of a
diver with the shore to his left.
67
-------
discharge dispersion trends observed at CSO 023, SD 7 and C 3.
By the time the light transmission measurements were made on
2/11 and 2/12/79, 2.1 - 2.4 cm (0.84 - 0.93 in) of rain had
fallen at CSO 044, SD 19 and C 4. Although the overflow
volumes at the CSO and SD were not measured, the 12-16 hr of
heavy rain recorded prior to station occupation suggest that
the measured plumes represented more than ten hours of discharge
At the CSO 044 site, a shallow, 10-hectare (25-acre) area
surrounding the outfall was found to be covered with discharge
plume (Figure 21).
Figure 21. Aerial perspectives of contours of percent light
transmission at Combined Sewer Outfall 044—2210-2350
hrs., 2/11/79.
68
-------
The isopleths of light transmission for this area were pre-
dominantly vertical from the surface to the bottom at 6 m; there
was some indication of near-bottom movement toward the north-
east, i.e., countercurrent to the apparent surface drift. The
magnitude of discharge contribution from the 137 cm storm drain
northwest of the CSO 044 outfall is unknown, but was probably
considerable.
The turbidity* patterns monitored around the SD 19 outfall during
the same storm showed an added dimension of complexity due to
an extremely dense load of particulates moving past from the
south. These solids undoubtedly were part of the storm load
from the Cedar River (Figure 2). The turbid surface layer
related to this phenomenon was homogeneous to a depth of 9 m
and had light transmission readings as low as 10% (Figure 22).
z
12m W
15 feet (4.6 meters)
Figure 22. Aerial perspectives of contours of percent light trans-
mission at Storm Drain 19—0020-0250 hrs., 2/12/79.
The SD 19 outfall was simultaneously injecting turbid discharge
into the nearshore waters. The longshore movement of this plume
to the northeast trapped a wedge of comparatively clear ambient
water between it and the river discharge farther offshore. Dense
clouds of discharge particulates were also detected
sliding offshore along the transect lines 10 m S, 55 m N and
220 m N of the outfall; the net effect of this movement under
the river layer was to isolate a second parcel of ambient water
with core depths ranging between 6 m and 12 m and approximately
60 m offshore. Storm drainage from SD 19 was not found on the
69
-------
transects 70 m and 230 m south of the outfall.
The light transmission measurements made at Control Site 4 during
the storm of 2/12/79 revealed no coherent structure. Values at
all depths ranged between 40.5% and 43.5% light transmission.
Quantification of Particulate Contaminants — Sedimentation
Rates and Chemical Constituents — A total of 18 sediment traps
were moored in arrays of six at CSO 023, SD 7 and C 3
(Figure 7). The settling particulates captured by these devices
were collected periodically between January, 1978 and February,
1979. and analyzed for total dry weight, total C, total P, Pb,
Zn, Cu and Al.
The degree to which the measured sedimentation rates reflected
the volume of discharge at CSO 023 and SD 7 for each sampling
period was estimated by linear regression. The correlation
coefficients (r) are given in Table 10.
TABLE 10. CORRELATION COEFFICIENTS FOR NEARSHORE SEDI-
MENTATION RATES VS. VOLUME OF DISCHARGE FROM
COMBINED SEWER OUTFALL 023 AND STORM DRAIN 7
Station Discharge Type Corr. Coeff
CSO 023 Storm .852
SD 7 Storm .290
Storm + Nonstorm .157
Storm Drain 7 flows continuously, and therefore sedimen-
tation rates were regressed against both storm discharge and
total (storm + nonstorm) discharge in order to define the
relative strength of these relationships. Neither value of
r determined for this station indicated a close relationship
between discharge volume and sedimentation rate. These
findings are in accordance with previous observations that much
of the discharge plume characteristically bypassed the SD 7
sediment trap array. The opposite was found for CSO 023,
where the comparatively high correlation coefficient of 0.852
implied a close relationship; the plume at this station was
typically over the sediment traps.
Table 11 is a summary of the results of a series of two-way
analyses of variance performed for the various sediment trap
parameters, using stations and collection periods as variables.
These results indicate that there was significant interaction
between stations and collection periods on sedimentation
rate. This finding is reasonable since collection periods
70
-------
TABLE 11.
SUMMARY OF RESULTS OF TWO-WAY ANALYSIS OF
VARIANCE FOR SEDIMENT TRAP COLLECTIONS FROM
COMBINED SEWER OUTFALL 023, STORM DRAIN 7,
AND CONTROL SITE 3, USING STATION AND SAMPLING
PERIOD AS VARIABLES
Element^)
Number of
Sample Sets
Significance
Between Between
Sed. Rate (g/m2/day)
Total C (%)
Total P (%)
Pb (mg/kg)
Zn (mg/kg)
Cu (mg/kg)
Al (%)
11
9
11
11
11
11
11
0.01
0.01
0.05
0.01
0.01
0.01
0. 01
0.01
0.05
0.01
0.01
1* L.CJ. 0.^ UJ.U11
0 01
0. 05
0.01
0.05
0.05
(a) All data based on dry weight of participates.
(b) Level of a .
TABLE 12.
MEAN SEDIMENTATION RATES AND DRY WEIGHT CON-
CENTRATIONS OF SELECTED CONSTITUENTS ANALYZED
IN SEDIMENT TRAP SOLIDS COLLECTED AT COMBINED
SEWER OUTFALL 023, STORM DRAIN 7 AND CONTROL
SITE 3, 1/30/78-2/2/79
Sed. Rate (g/m2/day)
Total C (%)
Total P (%)
Pb (mg/kg)
Cu (mg/kg)
Zn
Al
(mg/kg)
9.
CSO
X
752
58
.252
366
126
349
4.
26
023
sx(a)
.714
3.85
.192
104
108
152
1.07
SD 7
9.
X
805
29
.212
521
94.8
286
4.
70
S
4.
X
696
5?
.211
355
44.8
127
1.
13
9.
C 3
X
268
34
.273
245
70.8
244
3.
72
Sx
.199
6 08
.205
97.1
41.8
125
1.09
(a) One standard deviation.
71
-------
reflect seasonal variability of discharge, with CSO 023 and
SD 7 having their greatest increases in sedimentation rate
(compared to that measured for C 3) during periods of storm
runoff. Even though collection period and station are thus
shown not to be independent variables, it is clear that both
CSO 023 and SD 7 had significantly higher sedimentation
rates than did the control site.
The mean sedimentation rates in g/m^/day (Table 12) from 1/30/78
to 2/2/79 were .752 for CSO 023, .805 for SD 7 and .268 for
C 3. A relative value for the adjacent profundal area is
.661, which was determined by correcting a rate measured at
60 m depth (Birch, 1976), for the offshore movement of parti-
culates entering the lake in the littoral zone. This cor-
rection was based on the observation that the area of per-
manent sediment accumulation in Lake Washington is about 50%
of the total surface area, and measured profundal values must
therefore be halved to render them comparable to littoral
measurements. The offshore sedimentation rate is, then,
less than nearshore values measured in the vicinity of the
outfalls, but greater than that determined for the control
area. A significant fraction of the suspended particulates
in the study areas is transient material which eventually
contributes to the sedimentation rate in the profundal zone.
The flux, or quantity settling through a given cross-sectional
area per unit time, of particulate C, P, Pb, Cu or Zn was
computed as the product of the constituent concentration
(weight/total weight of solids) and the sedimentation rate.
Compared to CSO 023 and SD 7, the control site had similar or
lower concentrations of each of the measured constituents
(Table 12) and a significantly lower sedimentation rate. The
flux of all solid elements at C 3 was therefore significantly
lower than at CSO 023 or SD 7.
The analysis of variance for total carbon, Table 11, shows
that the most significant variability can be attributed to
collection periods, and that there were no significant
differences between stations. The major source of measured
carbon, for which the highest concentrations occurred during
the summer months, appeared to be periphytic algae growing
on the sediment traps. Comparatively high negative cor-
relations (r=-.461, -.423 and -.538 for CSO 023, SD 7 and C 3,
with a =.01 for each) were calculated for %C vs. sedimentation
rate. This relationship can be attributed to the effects of
dilution by inorganic material during periods of high sedi-
mentation, and concomitant low light intensity (meaning
decreased periphytic photosynthesis). These phenomena are
typical of rainstorm and high-discharge conditions.
The trend for phosphorus was similar to that seen for carbon,
with significant differences in concentration between col-
72
-------
lection periods, but not between stations. Whereas there
were moderate correlations determined between C and P for
CSO 023 (r=.264) and SD 7 (r=.291), the correlation for the
control site (r=.660) was appreciably stronger. This points
to photosynthetic production as the common major source of C
and P in particulates settling in the control area.
The average concentrations of Pb in the sediment trap solids
were 366 mg/kg at CSO 023, 521 mg/kg at SD 7 and 245 mg/kg at
C 3, with the differences between stations found to be signifi-
cant at the 99% confidence level. These values are in keeping
with the relative Pb discharge evaluations discussed previously,
which showed SD 7 to be much higher than CSO 023 in this
respect (Tables 5 and 6).
Cu and Zn had similar concentration hierarchies: CSO 023 > SD
7 > C 3. According to the statistical results listed in Table 11
the between-station differences were not significant for Zn,
although some station-period interaction was indicated; the
opposite was true for Cu, possibly indicating notably different
sedimentation mechanisms for these two metals in the nearshore
environment.
Aluminum was analyzed as a conservative tracer of erosional
inputs to the system because it is a major sediment constituent
(3-10% of solids dry weight), and there are no non-erosional
sources of Al in the Lake Washington drainage basin. Correlation
coefficients calculated for sediment trap concentrations of
Al vs. storm discharge rates were .909 for CSO 023 and .363
for SD 7, the latter value being lower due to the tendency
of the SD 7 plumes to bypass the sediment traps, as determined
by light transmission measurements - the plumes were typically
found near shore in water too shallow for sediment trap
moorings. The mean Al concentrations determined for settling
particulates collected near both outfalls were significantly
higher than for solids from the control site, indicating a
greater input of erosional matter at the outfall sites.
Viruses — Table 13 is a compilation of data representing
virus analyses performed on receiving water samples collected
at CSO 023 and SD 7 during and after storm discharges.
As was expected from the previously-discussed lack of viruses
in the SD 7 discharge, none were found in the receiving
waters either. During an overflow at CSO 023, however,
viruses were detected at a level representing an estimated
end-of-pipe dilution of 32:1, the ratio of discharge con-
centration (10/23/78 sample, Table 7) to concentration in the
receiving waters (2/6/79 sample, Table 13). These two samples
were both collected during times of (assumed) light use of
sanitary facilities; for this reason it is suspected that
higher concentrations might be observed in the receiving waters
during peak-hour overflows (refer to discussion of discharge
73
-------
X
u
UJ
5 15
> 15-
oc
Q
u.
0
I- 10-
z
Ul
U
cc
Ul
Q. c
0-
w -30°-
CC
Ul
f-
Ul
I
0-
9 nn.
0-
.200-
3 SnNII/Mb31 3J.VQ
•nvdino ONndw
OJ. 3AI1V13U
31IS 31dlAIVS
ORGANIC CONTENT
nllnn
MEDIAN
PARTICLE
DIAMETER (M2)
?x1
•'•'•' T 1 1
• x'x
vv
p
?
p
m n ,-,
p-TT-l
ji:nn: :
SORTING 11-8
COEFFICIENT (SQ) PZI
D, .
rn.
LL
• X.;.' SKhWNtSS (S.
:xx 1 — I
^
y/
y/
%
^
n - - -
^•^ i — i
1 J-L
UIUIUIUI UJ LU UJ uj
EE EEuiEE EE
01 o en a> o « o o i
M M . oc co PO oo
i i i in
" Q
NG 1 <-WELL POORLY*.
^ SORTED SORTED
THEORETICAL
LIMIT
A — DISTRIBUTION FAVOR
FINER PARTICLES
THEORETICAL SYMM. = 1.0
Y^ DISCHARGE
[i-i^] DEBRIS NEAR OUTFALL
L_J SEDIMENTS
© SITE SAMPLED TWICE
COMBINED SEWER OUTFALL 023
STORM DRAIN 7
Figure 23. Summary of results of particle size distribution
analyses for Combined Sewer Outfall 023 and Storm
Drain 7.
75
-------
washing away of the smaller ones could be expected to improve
the sorting. If the discharge participates are comparatively
small (as for SD 7), they will likely be carried further from
the outfall before settling.
Core Analyses — To help define trends of movement and accumu-
lation of discharge particulates in the receiving waters,
sediment cores were collected at the transmissometer grid sites
at CSO 023, SD 7 and C 3. A total of 56 cores were dried,
sectioned and analyzed for Pb, Zn, Cu, P and total carbon.
Contoured distributions of some of these constituents in the
1-cm sediment surface layer are presented here in Figures 24,
25, and 26. In view of the typically patchy nature of aquatic
sediment distributions, particularly in areas subjected to
current action and a variety of anthropogenic influences, this
information should be used as a means to visualize general
trends, and not for detailed comparisons. The contours show
enrichment of the sediments near the two outfalls, with apparent
distribution modification from current action and near-bottom
downslope streaming. Table 14 lists the significant (a = .05)
differences between the stations for each of the measured con-
stituents; these were determined by a one-way analysis of
variance followed by the application of Scheffe's procedure for
linear contrasts (Scheffe, 1959).
TABLE 14. SIGNIFICANT DIFFERENCES (a =.05) BETWEEN MEAN
STATION CONCENTRATIONS FOR SELECTED PARAMETERS
ANALYZED IN THE TOP 0.5 CENTIMETER OF SEDIMENT
CORES COLLECTED AT COMBINED SEWER OUTFALL 023,
STORM DRAIN 7 AND CONTROL SITE 3
Wet wt./dry wt. CSO 023 > SD 7
Total C CSO 023 > SD 7
CSO 023 > C3
Total P None
Pb CSO 023 > C3
Zn CSO 023 > SD 7 < C3
Cu CSO 023 > C3
SD 7 > C3
77
-------
, STORM DRAIN 7 OUTFALL
*^
Z
I
m
-• 30m W
ii OUTFALL LINE
15m E
i 46m E
107m E
LEAD (mg/kg)
ZINC (mg/kg)
COPPER (mg/kg)
TOTAL CARBON (%)
Figure 25. Dry weight distributions of
lead, zinc, copper and carbon
in the surface centimeter of
sediments collected near
Storm Drain 7.
79
-------
These comparisons indicate tendencies for CSO 023 to have the
highest mean concentrations of discharge contaminants found in
the thin, transitory sediment surface layer. Overall, SD 7 was
not found to be significantly different than C 3 in this sense,
this was predominantly due to the comparatively rapid movement
of discharge particulates out of the sampling area, as is
apparent from the distributions in Figure 25. Further, evidence
is given below to show that the C 3 control site itself has been
contaminated by particulates advected from remote sources.
Table 15 is a compilation of metals data for particulates
sampled at the three principal study sites. These numbers
provide a useful aid for tracing the movements of discharge
solids. The concentrations of metals in the sediment trap
TABLE 15.
MEAN CONCENTRATIONS^ AND CONCENTRATION
RATIOS FOR SELECTED METALS IN PARTICULATES
SAMPLED AT COMBINED SEWER OUTFALL 023,
STORM DRAIN 7 AND CONTROL SITE 3
Mean Concentration
(mg/kg)
Location
CSO 023
SD 7
C 3
Profundal(c)
Source
Discharge *k)
Sed. Traps
0-0.5 cm Sed.
7-8 cm Sed.
Discharge '*>)
Sed. Traps
0-0.5 cm Sed.
7-8 cm Sed.
Sed. Traps
0-0.5 cm Sed.
7-8 cm Sed.
0-1 cm Sed.
Pre-1900 Sed.
Pb
257
366
129
51
2377
521
89
46
245
86
21
192
12
Zn
560
349
168
92
785
286
61
35
244
131
51
192
59
Cu
265
126
39
40
333
95
47
42
71
20
7
46
17
Pb:Cu
1.0
2.9
3.3
1.3
7.1
5.5
1.9
1.1
3.5
4.4
2.9
4.2
1.0
Zn:Cu
2.1
2.8
4.3
2.3
2.4
3.0
1.3
0.8
3.4
6.6
7.1
4.2
3.5
Zn:Pb
2.2
0.9
1.3
1.8
0.3
0.5
0.7
1.4
1.0
1.5
2.5
1.0
3.5
(a) Dry Weight.
(b) Calculated from data in Tables 5 and 6, as mg/kg of
suspended solids in the discharge.
(c) From *l®Pb-dated cores collected in the deep central
basin of Lake Washington, between CSO 023 and SD 7
(Spyridakis and Barnes, 1976).
81
-------
particulate Pb contributions from aeolian transport (Spyridakis
and Barnes, 1976). The subsequent increase in the Zn:Pb
ratio between the traps and the sediments implies a shorter
local residence time for Pb. All of the sediments above 8 cm
depth at CSO 023 have been deposited since the addition of
tetraethyl lead to gasoline in the late 1920s, as is evident
from their substantial Pb enrichment relative to pre-1900
(background) concentrations.
As noted in previous discussions relative to SD 7, substantial
advective losses of wastewater particulates from the sampling
area result in a considerably different pattern of local
deposition compared to that seen at CSO 023. Near the SD 7
outfall, the sediment distributions of Pb, Zn, Cu and C were
all found to be quite similar (Figure 25), showing the dominant
influence of near-bottom downslope streaming just north of
the outfall.
Comparisons of median discharge particle diameter (refer to
discussion of grain size distributions, above) showed that the
wastewater particulates emitted by SD 7 were much smaller than
those from CSO 023, facilitating advective losses. This
observation helps to account for the comparatively precipitous
drop of particulate metals concentrations between the SD 7
discharge and the sediment traps; the concentrations of all three
metals in the discharge particulates were substantially higher
at SD 7 than at CSO 023, but sediment trap concentrations were
higher only for Pb. Based on the relative decreases in metals
concentration from outfall to the surface sediments (Table 15),
and on their sediment correlations with carbon concentrations
(Pb: r=.920, Zn: r=.893, Cu: r=.440), the dispersion hierarchy
for SD 7 was determined to be Pb> Zn> Cu.
Compared to metals-bearing solids expelled by CSO 023, then,
the particles from SD 7 were appreciably smaller overall, with
less difference in the relative size and motility of Cu con-
taminants than for those carrying Pb and Zn. The steeper
bathymetry of the SD 7 sampling site also helps aggravate the
offshore, near-bottom movement of discharge particulates.
Indeed, inspection of the concentration distributions in
Figure 25 implies that the sediment enrichment maxima for SD 7
may lie even further offshore.
The surface layer of the bottom sediments at control site C 3
also show signs of contamination by Pb and Zn, with Cu con-
centrations being similar to those measured in pre-1900
(background) sediments (Table 15). Regression matrices
calculated for the C 3 sediments (Table A-2, Appendix) show
significant (a =.01) correlations for all combinations of C,
P, Cu, Zn and Pb (range of r=.645 to .951) in the 0-0.5 cm
layer, whereas no correlations at that level of significance
for the 7-8 cm layer. The metals concentrations in the deeper
83
-------
TABLE 16. UPPER CONFIDENCE LIMITS EXPRESSED
AS PERCENTAGES OF THE MEANS FOR
2, 4, 6, 8 AND 10 CORES PER
SAMPLING LOCATION FOR CHIRONOMID,
OLIGOCHAETE AND COPEPOD COUNTS
No. Of
Cores/ Depth
Grid (m)
2 1.5
4 1.5
6 1.5
8 1.5
10 1.5
2 4.6
4 4.6
6 4.6
8 4.6
10 4.6
2 7.6
4 7.6
6 7.6
8 7.6
10 7.6
Core Diameter =
29 mm
Chironomids Copepods
Oligochaetes
167.6
144.1
135.9
129.6
126.2
233.1
182.4
163.2
153.1
146.3
299.0
219.3
190.6
175.4
165.4
337.9
239.9
205.3
186.9
175.4
267.3
200.4
176.2
163.3
155.2
315.7
231.4
200.6
183.6
173.2
325.5
230.4
197.7
180.4
169.5
706.3
405.5
316.2
272.3
245.4
340.7
242.6
207.9
189.2
177.2
Core Diameter =
38 mm
Chironomids Copepods
Oligochaetes
176.1
149.3
138.5
132.8
128.8
215.4
172.4
155.9
147.0
141.2
274.7
205.7
180.6
167.1
158.4
342.8
238.8
203.7
1.84.9
1.73.4
285.7
212.0
185.7
171.2
161.9
372.6
261.4
222.0
200.9
187.7
580.6
349.7
279.0
243.9
222.9
586.8
352.7
281.2
245.6
223.8
279.6
206.9
181.1
167.2
158.4
as percentages of the mean total sample counts for each group
for a range of numbers of replicate cores; two different core
volumes were thusly evaluated. In all cases, the statistical
confidence increased with an increasing number of cores per
grid placement. With less than six cores taken per placement,
however, the confidence decreased substantially. Collecting
more than six cores per placement, on the other hand, only
slightly increased the confidence. The final sampling was
85
-------
Phylum: Arthropoda
Class: Insecta
Order: Diptera
Family: Chironomidae (Tendipedidae; midges)
Family: Ceratopogonidae (Heleidae; biting midges)
Order: Collembola (springtails)
Order: Trichoptera (caddisflies)
Order: Ephemeroptera (mayflies)
Order: Plecoptera (stoneflies)
Class: Arachnoidea
Order: Hydracarina (water mites)
Class: Crustacea
Subclass: Ostracoda (ostracods, seed shrimp)
Subclass: Copepoda
Order: Eucopepoda (Harpacticoid copepods)
Subclass: Malacostraca
Order: Mysidacea (mysids)
Order: Amphipoda (amphipods)
Phylum: Annelida
Class: Oligochaeta (aquatic earthworms)
Class: Hirundinea (leeches)
Phylum: Oncopoda
Class: Tardigrada (water bears)
Phylum: Platyhelminthes
Class: Turbellaria (flatworms)
Phylum: Nematoda (nematodes)
Phylum: Coelenterata
Class: Hydrozoa (freshwater hydra and jellyfish)
Phylum: Molusca
Class: Gastropoda
Order: Pulmonata
Family: Planorbidae
Genus: Gyraulus (freshwater snail)
Family: Lymnaeidae
Genus: Lymnaea (freshwater snail)
Class: Pelecypoda
Family: Sphaeriidae
Genus: Pisidium (freshwater mussel)
87
-------
TABLE 17.
Independent
Variable
Depth
Distance
Depth+Di s tance
Taxonomic
Group C3 C4
Chironomids (-) .66 (-).61
Oligochaetes (-).50 (-).55
Copepods (-) .48 (-).69
Nematodes (-) .49 (-).63
Pisidium (-).OOOl -10
Total C-J.66 C-).79
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
Samplinq Area
CSO 023 CSO 044 SD 7
(-).52
(-).04
(-).15
(-).37
(-).21
(-).59
.15 .06 NA(a)
(-).56 (-).07 (-).12
.10 .0_2 .06
(-).ll .004 (-).27
.10 (-).Ol (-).13
(-J.15 .001 (-).46
ND(b)
.16
.18
.37
.21
.59
SD 19
(-) .008
(-).04
(-).oi
(-) .0004
.07
(-) .002
.16
.005
.07
.20
.18
.16
.28
.09
.14
.28
.18
.25
(a) No association.
(b) Not determined.
Minus signs indicate negative correlation.
Underlined values represent nonsignificant (a = .05) correlations
89
-------
TABLE 19.
COEFFICIENTS OF DETERMINATION (r2) FOR LINEAR
AND MULTIPLE REGRESSIONS USING TOTAL ORGANISM
WEIGHT IN FEBRUARY AS THE DEPENDENT VARIABLE
Independent
Variable
Taxonomic Sampling Area
Group C3 C4 CSO 023 CSO 044 SD 7 SD 19
Depth
Distance
Chironomids (-).17 (-).42
Oligochaetes (-) .0^ (-) .2_3
Copepods
Nematodes
Pisidium
Total (~).04 (-).36
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
.001 (~).002
(-).Ol (-).02
.06 .001
(-).21 (-).Ol (-).Ol
Depth+Distance Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
18
06
.04
.03
.05
.15
.06
17
.13
Minus signs indicate negative correlation.
Underlined values represent nonsignificant (o= .05) correlations.
91
-------
NON-SIGNIFICANT CORRELATIONS
COUNTS: NONE
WEIGHT: Ch
(Co. N, P NOT TESTED)
IOUNTS
( 1
WEIGHT (mg)
(...)
NON-SIGNIFICANT CORRELATIONS
COUNTS: Co. N. P. T
-10-L-1.0 WEIGHT: Ch. O, T
(Co. N. P NOT TESTED!
COUNTS WEIGHT (mg)
NON-SIGNIFICANT CORRELATIONS
COUNTS: 0
WEIGHT: Ch. O. T
(Co. N. P NOT TESTED!
NON SIGNIFICANT CORRELATIONS
COUNTS: T
WEIGHT: T
3.0 (Co. N, P NOT TESTED)
Figure 27. Net change in numbers of organisms and biomass per core, as a
function of distance from the combined sewer and storm drain
outfalls in February and September, 1978. All relationships
determined by regression analysis and significant (at the 95%
confidence level) unless otherwise noted. Ch: chironomids,
O: oligochaetes, Co: copepods, N: nematodes, P: pelecypods
(Pisidium spp.), T: total for all organisms.
93
-------
constant depth was due to the presence of the outfall. With
reference to Figure 27, the strongest distance relationships
determined for CSO 023 were those for the oligochaetes
(aquatic earthworms) in February and the nematodes (roundworms)
in September. The correlations in each case were negative,
suggesting population enhancement near the outfall; these
data reinforce the commonly noted association of detritus-
feeding aquatic worms with sediment deposits rich in organic
pollutants (Hart and Fuller, 1974; Wetzel, 1975). The total
population of both oligochaetes and nematodes was seen to
increase during the dry season, when warm weather stimulates
feeding and reproductive activity (Wetzel, 1975). The
principal difference between the two taxonomic groupings was
that, in general, the February-to-September increases for
oligochaetes occurred away from the outfall, whereas those
for nematodes occurred near the outfall, with the residual
population in each instance remaining essentially constant;
the reasons for this difference are unknown, but may relate
to feeding competition or a differential response to toxicants
or substrate alterations. Oligochaetes and nematodes dictated
the total population correlations in February and September,
respectively.
Figure 27 also shows a weak positive relationship at CSO 023
between distance from the outfall and the numbers of chironomid
(insect) larvae, copepods (microcrustaceans) and pelecypods
(mussels). These correlations, which were non-seasonal in
terms of magnitude, imply that the CSO discharges may have been
slightly toxic to these populations. Smothering by settling
particulates may also have contributed to the observed decreases
CSO 044 — As for CSO 023 in February, there were negative
correlations of numbers of oligochaetes with distance from the
CSO 044 outfall during both seasons. The fraction of the total
oligochaete population found in the visible discharge debris
near the outfall (representing 55% of the samples taken) was
68-73% for all three data sets, indicating enhancement by the
organic effluent (refer to Figure 4 for sampling array con-
figurations) . The reason for the lack of similar correlations
at the two CSOs in September is not clear, but may be a result
of inter-specific differences in toxicity tolerance.
The only other relationship that was consistent between CSO 023
and CSO 044 was the positive correlation with distance of
numbers of chironomids, during both seasons. The fraction of
these organisms in the outfall debris (half of the samples)
ranged from 23% to 36% for the two outfalls during the two
seasons. This numerial depletion near the outfalls probably
represents a reaction to toxic substances or substrate alter-
ations in the near-outfall deposits.
95
-------
with distance from the outfall) in both February and September;
the effect for copepods and nematodes was less definite, with
copepod enhancement in September being the most definitively
indicated relationship (based on the raw data); the pelecypod
correlations for both seasons showed no notable distance
relationships. Of these, the trends denoted for oligochaetes
agreed with those at the CSOs, but enhancement of chironomids
and copepods was distinctly different from the depletion effect
associated with the combined wastewater systems. Based on the
present analysis, however, the trend disagreement was lacking
statistical verification. There was no apparent difference in
the mean concentrations or loading of the discharge constituents
at these stations (Tables 5 and 8) that would explain this
observation.
SD 19 — The data trends for SD 19 were much stronger than those
at SD 7 and the final observations based on both the raw data
and the statistical analyses vary only slightly from the summary
depicted by Figure 27 and Tables 17-20. The single apparent
change is for numbers of oligochaetes, which were in fact
enhanced by the discharge during both seasons, except directly
in front of the outfall, where there was a nontypical sand
substrate. This material was probably from the outfall and
in itself constitutes a localized discharge impact that reduced
the infaunal population at that site. This effect was contrary
to the overall areal trend only for oligochaetes. All other
populations tested showed definite evidence of depletion near
the outfall.
Summary — The general trends observed with respect to waste-
water impacts on the benthic infauna near the four outfalls
indicated enhancement of oligochaetes (aquatic earthworms) at
all sites. The greatest enrichment of the worm population
occurred at the CSO sites, probably due to the "predigested"
(in human guts) nature of the particulate carbon of the
discharge, a mixture typically rich in carbohydrates, pro-
teinaceous compounds, fatty acids and other substances more
readily degraded by microorganisms, which in turn are the
principal food source of the oligochaetes. By contrast, the
SDs emit principally humic substances, partially degraded
plant material that is more resistant to microbial degredation,
and tends to persist as particulates with relatively long
residence times. Decremental impacts, attributed to discharge
toxicity, substrate alterations and/or smothering by parti-
culates, were determined for chironomids (aquatic insect larvae)
at CSO 023, CSO 044 and SD 19; the chironomid trends for SD 7
were somewhat obscured by aberrant patches of substrate, but
appeared to indicate near-outfall enhancement during both
seasons. At CSO 023, CSO 044 and SD 19, the populations of
chironomids remained essentially constant near the outfalls
from February to September. During the same period, the
numbers of these organisms further away increased substantially,
97
-------
MARINE STUDIES
Discharge Monitoring
Discharge Loading Estimates for Single Events —
The CSO discharges generated by two storm events were sampled
quantitatively and qualitatively at the Denny Way Regulator
in order to determine typical particulate inputs to marine
waters by a large combined sewer outfall. Table 21 is a
summary of the storms monitored.
TABLE 21. SUMMARY OF RAINSTORMS MONITORED FOR
QUANTITY AND QUALITY OF DISCHARGES
INTO PUGET SOUND FROM THE DENNY WAY
REGULATOR OUTFALL, MARCH AND OCTOBER,
1978
Total Overflow Discharge Discharge
Rainfall Duration Volume Fraction^'
Date (cm) (hr) (m3) (%)
3/7
10/23-10/24
3.20
0.53
6.3
3.7
20730
12010
29.1
102
(a) The volume fraction of the total basin-incident rainfall that
was discharged by the outfall during the period noted.
Further, a statistical summary of the resultant contaminant
mass discharges is offered in Table 22. Relative to the
comparable data compiled for the freshwater study sites, the
total storm loading of each pollutant discharged by the
Denny Way outfall was much greater, due to its larger dis-
charge volumes. Only the mean concentrations of suspended
solids, total Hg and total P at Denny Way were similar to
those of the other combined facility, CSO 023. The mean
total and particulate Cu, Pb and Zn concentrations at Denny
Way were higher than those of either of the freshwater stations,
reflecting its large component of industrial and commercial
inputs (34% by land area). The mean concentration of particulate
chlorinated hydrocarbons at Denny Way was only a small fraction
of the concentrations measured at the two Lake Washington sites,
implicating residential use of pesticides as the principal
source.
Rainfall, Flow and Loading Summary —
As mentioned previously, the total measured rainfall for the
Denny Way station was 98.8 cm for the one year study period.
This was just slightly more than a 30-year annual mean
published for the same area by Phillips (1968). The rainfall
pattern and the resultant overflow response by the Denny Way
99
-------
system is shown in Figure 28. For a total of 59 periods of
measurable rainfall (separated by one or more dry days),
there were 36 controlled overflows at Denny Way. Taking
into account the fact that multiple overflows occurred
during some of the long storms, the fraction of storms
resulting in overflows was 51%; the average total rainfall
for those storms that failed to generate overflows was .25 cm.
The total overflow volume for the 12-month monitoring period
was 6.60 x 105m3 (175 MG). In combination with the mean
discharge concentrations given in Table 22, this value was
used to calculate total annual loads for the various parameters
of interest. The results are presented here in Table 23.
With the exception of chlorinated hydrocarbons, the loading
for all parameters at Denny Way was much greater than at CSO
023 and SD 7, due to its greater total volume of overflow.
TABLE 23. ESTIMATED MASS OF SELECTED CONSTITUENTS
IN DISCHARGES FROM THE DENNY WAY REGULATOR
OUTFALL, MARCH 3, 1978—FEBRUARY 28, 1979
Suspended solids
Cu
Hg
Pb
Zn
Al
Organic C
Total P
O&G
ClHC(b}
Total
Mass (kg)
85100
50.8
.396
254
188
1730
13700
812
10600, .
ND(C)
Particulate
Mass
85100
36
236
103
1640
7630
395
NA
(kg)
.6
.280
(a)
. 660g
(a) Not applicable.
(b) Selected chlorinated hydrocarbons: a-BHC, lindane,
heptachlor, heptachlor E, aldrin, dieldrin, endrin and
DDT (DDD + DDE + o,p DDT + p,p DDT).
(c) Not determined.
Turbidity Distributions of Discharge Plumes
The distributions and movements of wastewater discharge
plumes were monitored near the Denny Way outfall during
storms on April 15-16, 1978 and February 24-25, 1979. The
relative rainfall, total overflow volumes and tidal variations
are given in Table 24.
101
-------
TABLE 24. RAINFALL, OVERFLOW VOLUMES AND TIDAL
VARIATIONS FOR OVERFLOW TURBIDITY
DISTRIBUTIONS MONITORED AROUND THE
DENNY WAY REGULATOR OUTFALL, APRIL
1978 AND FEBRUARY 1979
April, 1978
15
16
17
February, 1979
24 25
Total Rainfall
Overflow Volume
Overflow Times
Monitoring Times
4.98 cm
56900 m3
1728 - 1210
5.18 cm
48100 m3
1322 - 0122
0517-0746 1254-1530 2050-2243 0555-0755
Tide
Time/Height
(a)
0502/2.
0901/2.
1640/0.
0
5
6
0013/3
0618/1
1057/2
1742/0
.1
.8
.4
.7
0105/3.
0717/1.
1216/2.
1841/0.
2
6
4
8
0339/3
0902/1
1409/3
2112/-
.6
.9
.3
0.4
0417/3.
0952/1.
1510/3.
1158/-0
7
6
4
.4
(a) In meters, relative to MLLW
As can be determined from these data, the most significant
differences between the two storms monitored are the strength
and number of tidal excursions associated with the overflows
prior to sampling, and the fact that the second 1978 sampling
was done post-storm, whereas that in 1979 was carried out
during the storm. All of the other important statistics are
similar for the two storms including the portions of the
tidal cycle sampled (low slack and ebb).
Interpretation of the Denny Way turbidity data was much more
difficult than were those collected previously in Lake
Washington. This was due to massive background interference
from the Duwamish River to the south (see Figure 2 for location
perspective). During substantial rainstorms the river flow
increases appreciably; as such, it is highly visible as a
muddy surface layer, 1.5-3.0 m thick, and as much 1000 m wide
(extending from the northeastern shore to the middle of
Elliott Bay) and moving along the shoreline through the
designated sampling area and northward around West Point.
The boundaries of this feature were verified during the
present project with light transmission measurements made
along transects extending more than one mile offshore into
Elliott Bay.
The collective influence of the river and the tidal strength
and direction can be seen in Figures 29 and 30 (refer to
freshwater section for a discussion of turbidity monitoring
103
-------
ui g 3 33
?! I s I
353
;:: in CM to o>
tex:
;"*:-::;!5:
0)
•H
4J
CJ
O
O
CN
Q)
M
^J
Cn
105
-------
N "088
0)
•H
-P
fl
O
O
O
ro
0)
107
-------
171m N
Figure 31. Transverse sections of contours of percent light transmission
around the overflow outfall of the Denny Way Regulator—
2050-2243 hrs.f 2/24/79. The perspective is that of a diver
looking south along the shoreline with Elliott Bay to his right,
The arrows indicate areas of particulate settling.
109
-------
Considering the extremely high discharge organic content at
Denny Way, and the 0% to 3% background measurements obtained
in conjunction with the biological studies, it seems apparent
that the three sediment samples collected in April showed
organic enrichment by discharged particulates. The poorly
sorted sample from 207 m S, 91 m W of the outfall seems to
have been influenced by the turbulence generated by tidal
action around Pier 71. This observation is further corroborated
by data. Together, these data constitute evidence that piers
can influence the settling characteristics of discharge parti-
culates; a phenomenon also observed near CSO 023 in Lake
Washington, as mentioned previously. The fact that the CSO
and SD particulates analyzed for the freshwater stations were
more poorly sorted (i.e., a larger range of particulate sizes
were intermixed) than those emanating from the Denny Way
facility would seem to indicate a greater tendency for
settling (a sorting process) within the latter system.
Three of the sites sampled near the Denny Way Regulator were
sampled twice for purposes of seasonal comparison. The
April, 1978 set was collected after a week of almost negligible
precipitation, whereas the February, 1979 sampling was done
toward the end of a storm that totalled 5.18 cm of rain,
and caused an overflow of 48100 m^ (12.7 MG). The differences
in the nearshore surface sediments under these two sets of
conditions is quite graphic. As shown by Figure 32, the
sediments collected following the large storm overflow had
an appreciably lower organic content due to dilution of the
particulate sanitary wastes with large quantities of inorganic
particulates carried into the system by storm runoff. These
more dynamic flow conditions also resulted in a much more
poorly sorted sediment surface layer.
Benthic Biota
The number, volume and rate of flow of overflows were appreciably
greater during the time prior to the first sampling (during
April, 1978) as compared to the time prior to the second
(during August, 1978) (Figure 28). Most overflows occurred
when the conduit was at least partially submerged. Therefore,
the discharge usually underwent immediate dilution. The
"wash out" zone was evident, however, indicating that scouring
was taking place.
Mollusc Bioassays--
Condition Index — The condition index (CI) of bivalves is a
measure of the "plumpness" of the soft tissue and can be used
to indicate the health of a population of these organisms. The
results of the in situ bivalve bioassay done for this project
showed that reproduction had the most significant effect on
condition index (Figure 33). Mussels in all test pots contained
large numbers of eggs in April. A marked decline in CI after
111
-------
that month was related to a characteristic decline in "plumpness"
of the test organisms associated with liberation of eggs.
There was no discernable increase or decrease in the CI of
oysters as a function of distance from the CSO; this suggests
that the CSO had no significant effect on the health of these
organisms. The experiments were terminated after approximately
four months due to the loss of all pots (through storms or
theft) except the one at 800 m N of the outfall.
The present results differ from those of a previous study at
Denny Way (Armstrong et al. , 1978). Armstrong et al. did see
a decrease in CI with decreasing distance from the CSO. The
organisms apparently did not reproduce during their study
period, however, and this may account for the differing results.
The organisms available for use in the present study had a
lower initial CI than those used by Armstrong et al. It may
be that, if the organisms had had a higher initial CI, a drop
in CI could have been detected.
Uptake of Metals — The heavy metal content of the meat of
mussels changed over the 10 weeks of exposure (Table 25). The
net increase in the concentrations of Cu and Pb was greatest
in the mussels closest to the CSO and decreased with distance
from the CSO. Zn concentrations decreased in mussel tissues at
all sites. The values for Cu and Pb were within the range of
values reported by Manly and George (1977) for the freshwater
mussel Anodonta anatina in polluted urban areas in Great
Britain. The Cu concentration found in mussels near the
Denny Way CSO was somewhat lower than the concentration in
the same species in polluted areas in southern California
(Young and Alexander, 1977). The Pb content was higher near
the CSO than in southern California, however. Similar to
the present study, Young and Alexander found that Zn appeared
not to be taken up by the mussels.
Lead concentrations in oyster meats appeared to be related
to proximity to the CSO (Table 26). There are no data
readily available from other geographic areas with which to
make comparisons.
Effects of Discharge on Algal Morphology —
Samples taken in April showed that the ratio of Enteromorpha
blade width to unit length decreased with increasing distance
from the CSO (Figure 34). This relationship was essentially
reversed in August. The difference between the results may
be related to the decline in number of overflows in summer.
There is no clear explanation evident either in the literature
or based on the growth strategy of this species. However, the
results may only indicate the relative influence of freshwater
on the organisms. Further study is needed to pinpoint the
causal factors involved.
113
-------
2.0-1
CSO
3.33
2.50-
1.67^
0.83-
60m N
6.0
LENGTH (mm)
10.0
12.0
60m S
150mS
(b) AUGUST
4.17
8.33 12.50
LENGTH (mm)
16.66 20.83
Figure 34. Size data for Enteromorpha collected
near the Denny Way Regulator outfall
during April and August, 1978.
115
-------
The total number of individuals found in the samples from
9 m depth differed appreciably among the sites (Figure 35).
The transect 5 samples had the most individuals and this
was primary due to the high abundance (1312 organisms) of
the pollution-indicating worm Capitella capitata. Infauna
was in low abundance at sites on transects 2, 6 and 7. The
latter results may be partially explained by the buildup of
black sediments at these sites. Transect 2, although
located relatively far from the CSO, showed low abundances
and this may be due to an eddy of materials at this site, which
is immediately seaward of a small cove. The results of
Armstrong et al. (1978) coincide with those of the present study
with the exception that they found fewer individuals of C.
capitata on transect 5.
The results from August were generally similar to those from
April. However, very few individuals (12) of C._ capitata were
found and the species exhibited smaller spacial variability.
Furthermore, there was a marked increase in the abundance of
the clam Axinopsida serricata in August. This species was
found to increase in number with increasing distance from the
CSO in April during the present study, and also during the
study by Armstrong et al. (1978) a year earlier. These results
suggest that the reduced flows from the CSO during summer result
in a change in the infaunal composition toward more natural
conditions.
The number of individuals found in the samples from 13 m also
varied widely among the sites. In April, the fewest individuals
were found at the site on transect 5. The number of individuals
was also low at sites on transects 1, 2 and 3. The results in
August showed that the number of individuals was low on
transect 4 and was again low on transect 5. Because the sites
on either side of transects 4 and 5 held more individuals, it
may be that this parameter is decreased at 13 m by materials
from the CSO. These results coincide well with those of Arm-
strong et al. (1978).
The number of taxa found at both 9 m and 13 m differed among
sites, and showed a general decline near the CSO. An exception
to this is the 9 m samples taken on transect 5 in April. The
higher number of taxa may be related to the larger number of
individuals that were collected at that time. The greatest
number of taxa were consistantly found in the samples from
the control transect (1). These data differed between sampling
periods, although the trends were similar. It is difficult to
explain the differences between sampling periods. The flux in
number of taxa between samplings was most pronounced on
transects 5 and 6 at 9 m, and transect 4 at 13 m. It may be
that a general northward spread of sewage-related materials
at depth results in the effect being most evident on transect
4. Again, these results are in concordance with those of
117
-------
Armstrong et al. (1978), except that they did not encounter
an increased number of taxa at 9 m on transect 5.
Species-richness curves (Hurlbert, 1971) were used to provide
an indication of taxocene diversity in the subtidal infauna.
Mollusc diversity was lowest at sites on transects 3, 4 and
5 at 9 m, and at the sites on transects 3, 5 and 6 at 13 m in
April (Figure 36). The samples from the control transect (1)
were most diverse. Diversity was low at sites on transect 5
in August also. Of note is the fact that relatively few mollusc
specimens were encountered on transect 5. It appears that
mollusc diversity and number of individuals are detrimentally
affected by the CSO discharges. The curves for most sites
were generally lower in April than in August, which suggests that
differences in flow from the CSO may be responsible for the
lowered diversity. Mollusc diversity has been shown to be
severely affected near sewage outfalls (Green, 1975).
Polychaete diversity also appeared to be affected by the CSO
discharge (Figure 37). This effect was evident primarily at
the 9 m sites during both April and August. The largest flux
between the seasons for this parameter was seen on transect
5 at 9 m and on transect 4 at 13 m. The sites on transect 1
(control) were consistently highest in polychaete species
diversity. Infaunal polychaete species diversity followed a
very similar trend in the April study by Armstrong et al. (1978).
In the present study there was a difference in the distribution
of curves between April and August. Polychaete diversity was
appreciably lower on transect 5 at 9 m in April and this may
be due to changes in flows from the CSO.
A cluster analysis of the subtidal infauna samples from
April revealed that sites tended to group by depth, and that
the site closest to the CSO (i.e., 9 m, transect 5) was sub-
stantially different from the rest of the sites (Figure 38a).
The sample on transect 5 at 13 m was also unique among the
13 m samples and clustered with the 9 m samples. The samples
were divided into five subgroups (i.e., A-E, Figure 38) and
the letters designating these subgroups were plotted on a
map of the site locations (Figure 39a). A unique infaunal
assemblage existed in April at 9 m on transects 5 and 6. The
similarity in infauna at sites 9/2, 9/3 and 13/5 may have been
related to the similar substrata at those sites. Black
sediment was found at each location (Table 27). The sediments
at the other sites at 13 m were light brown in color.
An eddying of materials immediately seaward of the cove adjacent
to transects 2 and 3 may be responsible for the disposition of
organic particulate matter at 9/2 and 9/3. The results of the
cluster analysis by Armstrong et al. (1978) were similar to
the results obtained in this study. However, in their study,
the site at 9 m on transect 2 clustered with the site on
119
-------
40
30
20
10-
f h-
300 1553
150 200 250 300
(d) AUGUST, 13m DEPTH
50
100 150 200
NO. OF INDIVIDUALS
250 300
Figure 37. Species-richness curves for polychaetes
collected near the Denny Way Regulator
outfall during April and August, 1978.
The numbers indicate sampling transects
121
-------
(a) APRIL
TRANSECT
— - 1 (CONTROL)
TRANSECT
(b) AUGUST
1 (CONTROL)
'cso
Figure 39. Positions of subgroups of sites from cluster
analysis of samples of subtidal infauna collected
near the Denny Way Regulator outfall during
April and August, 1978.
123
-------
(a) 9m DEPTH
100%
ARTHROPODS
100%
ARTHROPODS
100%
MOLLUSCS
Figure 40.
100%
MOLLUSCS
Proportions of the total numbers of annelids,
arthropods and molluscs collected near the
Denny Way Regulator outfall during April and
August, 1978. The numbers designate transects;
those in parentheses are for August. The
circles indicate 95% confidence intervals.
125
-------
100%
BURROWING DEPOSIT FEEDERS
100%
CARNIVORES
100%
CARNIVORES
100%
BURROWING DEPOSIT FEEDERS
(a) 9m DEPTH
100%
SURFACE DEPOSIT FEEDERS
(b) 13m DEPTH
Figure 41.
100%
SURFACE DEPOSIT FEEDERS
Proportions of total numbers of polychaetes
within each of three feeding-type categories,
for samples collected near the Denny Way Regulator
outfall during April and August, 1978. The num-
bers designate transects; those in parentheses
are for August. The circles indicate 95% con-
fidence intervals.
127
-------
TABLE 28. CHARACTERISTICS AND ANNELIDA DENSITIES OF
INTERTIDAL SOFT SEDIMENTS SAMPLED NEAR THE
DENNY WAY REGULATOR, APRIL AND AUGUST, 1978
Volatile Annelid
Organics Abundance
Date
April 24
Aug . 1 7
Site
a
b
c
d
e
f
g
h
i
j
a
b
c
d
e
f
g
h
i
j
Texture Color
silty-sand black
sparse cobble brown
silty-sand black
sparse cobble brown
n n
cobble-sand "
sm cobble, clay "
sm-med cobble, "
sand
n n
n n
silty-sand black
n n
it n
•• n
M n
•• n
M n
•i n
n n
Odor
H2S
11
"
"
11
11
11
none
"
11
H2S
n
n
n
n
n
n
n
n
n
(%) (No. I
0.86
0.83
7.80
0.99
1.35
1.03
0.87
0.66
0.64
0.70
1.49
0.71
2.32
0.72
1.51
1.02
0.59
0.73
0.52
0.68
ndividuals
42
51
121
122
41
53
242
3
0
1
6
8
70
9
10
9
43
42
14
23
(a) Capitella capitata.
diameter cores.
All counts are totals for four 31.3 mm
A plot of samples along canonical variables 1 and 2 did,
however, indicate that the periphyton communities at the two
sites nearest the CSO differed from the rest of the samples
(Figure 43). Of the taxa that entered the discriminant function,
Gomphonema was in high abundance and Synedra was in low
abundance, at the 16 m S site. Amphora was also in relatively
high abundance at this site. The 20 m N site had a very high
abundance of the filamentous diatom Melosira nummuloides. The
results of this analysis coincide with the conclusions of
Archibald (1972) who found that diatom diversity did not
always follow a pollution gradient, and that autecology of
dominant species proved to be the most important criterion
for assessing water quality.
129
-------
scouring sediments. Results presented elsewhere in this report
indicate that turbidity, measured as percent light transmittance,
is moderately high within approximately 300 m of the CSO.
Zimmerman and Livingston (1976) concluded that increased
turbidity (by Kraft Pulp Mill effluent) was the most important
factor in modifying macrophyte community structure at their
study sites in Florida.
A plot of the boulder wall sites along canonical variables
one and two revealed that the effect of the CSO may extend
to the site 380 m north (Figure 45). This site was more similar
1.0-
-.04-
-1.0-
2 -1.8H
•
150mS
120m N
30m S
(a) APRIL
600m N
380m N
30m N
800m N
-3.2
-.8
CANONICAL VARIABLE 1
1.5
2.7
1.8-
-.5-
2.3-
800m N
150mS
• 380m N
• 30m S
(b) AUGUST
600m N
30m N
120m N
-4.9
-.5 4.0
CANONICAL VARIABLE 1
8.5
13.0
Figure 45. Position on canonical variables of samples
of boulder wall taxa collected near the
Denny Way Regulator outfall, 1978.
131
-------
SECTION 6
DISCUSSION
EFFECTS OF DISCHARGES ON LITTORAL FOOD CHAINS
Freshwater Environment
The original intent of the freshwater biota studies was to
describe the numerical abundance, biomass, and types of
benthic organisms in control areas and areas receiving
sewer discharges in Lake Washington. The discussion would be
incomplete, however, without relating these findings to the
possible effects on fish at those sites.
It should be realized by the planner and fishery manager
that in a system as large and diverse as Lake Washington it
would be an impossibility to thoroughly assess the specific
effects on fish resulting from changes in the benthic community
A specific effect on one species of fish may'indirectly affect
several other species throughout the food chain. The possible
impacts to the system are virtually limitless. The purpose
of the following discussion is to consider the present findings
for the benthic communities relative to possible effects on
the feeding opportunities of fish in those areas.
Indigenous Fish Species and Principal Food Types —
A total of 39 species of fish are known to occur in the Lake
Washington drainage. These include a number of sportfish
(pink salmon, chum salmon, coho salmon, sockeye salmon and
kokanee, chinook salmon, coastal cutthroat trout, rainbow
trout and steelhead, brook trout, Dolly Varden, lake trout,
smallmouth bass, largemouth bass and black crappie) as well
as important fish-prey species (such as speckled dace, redside
shiner, threespine stickleback and six species of sculpins,
with the prickly sculpin being the most abundant in Lake
Washington) (Wydoski, 1972). Of the species common to Lake
Washington proper, at least 14 have been found to depend on
benthic organisms for all or part of their food, including
largemouth bass (Stein, 1970), cutthroat and rainbow trout
(McAfee, 1966; Wydoski and Whitney, 1979), black crappie
(Wydoski and Whitney, 1979), smallmouth bass (Emig, 1966),
and a prominent prey species, the prickly sculpin (Rickard,
1979) .
133
-------
bay, the observed alteration may represent a significant impact
on the total productivity of the bay's detrital-based benthos.
The dispersion of participate toxicants, and the alteration of
substrate consistancy by a large marine CSO such as that at the
su^LWayfR??Uif°r may als° m°re ai«ctly affect thl feeding
success of fish*. In general, the rocky and gravelly-sand
=™ ^" ^S Seattle area support a larger number of fish
species than do silt-mud habitats. Salmon, herring, anchovies
co?tf^ 'n3^-13^3' 9reenlin^' gunnels, sand lances,
cottids, pipefish and wolffish occur in rocky-gravel habitats
mudC°n^a^' =?ulpins and lampreys prefer area! of silt and
mud. Smelt, flounders and stickleback are found on both types
of substrate (B.S. Miller, University of Washington, personal
communication). Near the Denny Way CSO, the in?e??idal sSb-
is silt-mud0 Phv^6reah fcl?at °?. the contiguous subtidal region
is silt mud. Physico-chemical disturbances by the CSO in the
form of increased turbidity and siltation, and of increased
concentrations of particulate toxicants, may result in corres-
?hf h^h ?" " thS faunal =°">P°3ition and abundances of
the benthic fauna. These changes in turn can influence the
fn^=?9^UCCeS?1°f flsh Such as salmonids, which selectively
ingest the small epifaunal crustaceans that inhabit rocky-
gravel substrates (J.Q. Word, University of Washington personal
communication). The alteration of a benthic subst?a?e by fine
such as th^n °m ^ *"* S°S W°Uld tend to favor fi^ Cecils
inCa va%ietyDofehaMtats? n°"-Selecti- —- that does^ well
r,=^e<-arer,lndlcati°ns that shifts in the benthic assemblages
near the Denny Way CSO may be seasonal as was indicated for
the communities studied in Lake Washington. In the sprina
toward the end of the wet season, which is characterized bv
numerous overflows, a deposit feeding infaunal community pre-
s^on nSar the outfa11' However, the dominant summer ~
season organism is a small epifaunal crustacean, Nebalia
It is anticipated that this shift in type of prey
may result in a concurrent shift in the
species.
Due to the comparatively natural configuration and substrate
composition of the surrounding shoreline, and to the ecoloaical
importance of this semi-natural habitat, the Denny Way CSO
Fish occur in considerable numbers in the bay, and there is a
R^vpr T^ arTal *alm°nid run in the ^-flowing Duwamish
River. Juvenile salmonids, either from natural or hatcherv
stocks, migrate downstream in spring and summer months through
the Duwamish estuary and Elliott Bay. These fish spend much
of their time in nearshore habitats in these areas
138
the baseline total tor an organisms scuiipxeu. a^
lations for trophic level conversion assume that the fish
are 100% successful in locating and consuming all of the prey
organisms, and in this sense the estimates are obviously too
high.
136
depend on benthic food sources in Lake Washington. The
following is a discussion summarizing our findings concerning
the benthic communities at our study sites and a description
of the possible effects on the feeding opportunities of the
fish in those areas.
At Control Sites 3 and 4 there were generally moderate to
strong negative correlations between the numbers and biomass
of organisms and depth, whether considering types of organisms
separately or combined. This means that there were generally
higher numbers of organisms in shallow water. The only
exception was the Pisidium spp. or freshwater mussel which
either showed no relationship with depth or was found in
slightly higher numbers in deep water. These data would
indicate that at control sites the possibility for benthic
feeders to encounter food items would be greater in shallow
water, and equal at constant depth. This was not always the
case at the SD and CSO sites.
At CSO 023 there were generally higher predicted total numbers
of organisms (predominantly oligochaetes) and greater biomass
near the outfall. If fish feed in this area, it would mean
a greater opportunity to encounter food items near the source
of the discharge. For fish selecting a single type of organism,
however, this would not always be true at CSO 023. Chironomids
were generally found in low numbers near the outfall but at a
larger average size. This means that fish selecting only
134
-------
chironomids would encounter them less frequently near the
outfall but may receive a "normal" food volume due to the
larger average size of the prey items. Unlike chironomids,
the predicted number and biomass of oligochaetes in February
was greater near the source of the discharges. In September
there were no relationships between numbers and biomass of
oligochaetes and distance from the outfall.
This suggests that if fish feed near the outfall in February,
they would encounter more food items. In September, however,
fish that feed throughout the area would encounter the same
number of oligochaetes regardless of the location. It is
possible that between February and September higher^feeding
activity, rather than pollution effects on food availability,
was responsible for the reduced relative numbers and biomass
of oligochaetes near the outfall. Since copepods and mussels
were always found in fewer numbers near the outfall, it
seems likely that the discharge is a contributing factor.
Nematodes, on the other hand, were always found in higher
numbers near the outfall at CSO 023. This indicates
that either fish do not feed on nematodes near the outfall
or that the discharge increases the number of nematodes
near the outfall and therefore, the opportunity to encounter
food items. This raises the possibility that the distri-
butions of some benthic species around an outfall may reflect
indirect, rather than direct, effects of the discharges, i.e.,
that distributions of the benthic biota are dictated by the
feeding patterns of fish, which are in turn influenced by
the discharges.
Qualitative relationships between the discharges and feeding
opportunities for fish are obviously quite speculative for
CSO 023, as well as for any of the other three outfalls studied,
but generally follow an application of the above logic to the
relationships depicted by Figure 27. In a few cases, such as
for CSO 023 in both February and September, the numbers and/or
biomass of the benthic biota were dominated by a single type
of organism; this effectively removes one variable in the
complex cause-effect relationships. Subsequent speculation
would then focus on the compilation of a list of fish predator
species for that type of organism, and the relative depth/
temperature preferences and toxicity tolerances for each
(if available).
Toxicity responses are particularly important in those instances
when the predator may be attracted to the outfall by enhance-
ment of the numbers and/or biomass of the prey organisms, such
as observed for oligochaetes in February, or nematodes in
September at CSO 023. The impacts of ingested toxicants on the
health of the predator fish and the toxic and smothering effects
of the discharges entering breeding areas are additional impor-
tant - and complex - topics needing further consideration.
135
-------
on receiving water circulation. The relative effects of
loading magnitude and nearshore circulation on the local
biota were not quantified by the present analyses?
EXTENT OP NEARSHORE WATERS AFFECTED BY DISCHARGE PARTICULATES
Freshwater Environment
Uk. »,,hin,,to» ™y b. con,lder.i
as h -•
"
140
-------
The preceding calculation represents a very simplified
approach to an exceedingly complex question, and should not
be interpreted or expressed as a close estimate. Its principal
value is to provide an order-of-magnitude estimate for one
example of discharge impact on fish in Lake Washington.
Further, if used as an extreme for those weight/distance
relationships shown in Figure 27, together with knowledge of
the dimensions of impacted areas, it can be used to estimate
the magnitude of potential impacts for other taxa and stations.
For example, at CSO 044 in September the data indicated a
depletion of chironomid biomass around the outfall, but the
sub-background weights were limited to a circle with an 8 m
radius. This impacted area was appreciably smaller than
that determined for oligochaetes at CSO 023, as was the
change in biomass per unit distance. A comparison of these
two relationships, as depicted by Figure 27, indicates that
the net decrease in prey and potential predator (fish)
biomass at CSO 044 was perhaps 1% of the increase calculated
for oligochaetes at CSO 023. This would mean a loss on the
order of 0.5 kg fresh wt. of chironomids, and .05 kg fresh wt.
of fish.
Marine Environment
The coastland of Elliott Bay constitutes a substantial fraction
of Seattle's total marine shoreline (Figure 1). The present
study investigated an estimated 17% of this feature, approxi-
mately 15% of which was found to be impacted by discharges from
the Denny Way Regulator CSO. Of the shallow (0-13 m depth)
nearshore area, including both hard and soft substrata, about
8% was found to be biologically altered by the CSO discharges.
Much of the nearshore region in Elliott Bay is dominated by
man-made structures such as piers and sea walls. Since the
study area for the present project is relatively devoid of
such structures, it may represent the most natural habitat
available to the bay's biological communities, including
salmonid species. From this perspective, semi-natural
biological assemblages in the bay constitute a diminished
resource and the additional physiological stresses contributed
by polluted discharges represent a serious threat to their
survival and well-being.
The benthic food web in Elliott Bay is detrital-based, and
depends upon input of carbon from phytoplankton, benthic
algae and allochthonous material. The results of the present
analyses indicate that the primary productivity of benthic
macrophytes (in terms of standing stock measurements) is
significantly altered near the Denny Way CSO. Because the
boulder wall near the CSO is an important habitat for
macrophyte production and is not prevalent elsewhere in the
137
-------
-
and more dynamic nearshore circulation. Elliott Bay and the
a; s-zaji toi
S^s-i- ssi:,
have a mean annual discharge of over 7.6 x 10? m* (LS MG?
(Brown and Caldwell, 1979). The nearshore circulation patterns
include strong nearshore tidal currents with localized eddies
(Tomlinson et al., 1976), and the often silt-laden pfumTof
37 mMgTx loTL^1C\ Snt&rS BlliOtt Ba* at a »»an rate of
J/ m-> (3.7 x 1QJ gallons) per second.
Relative to the dispersion of discharged particulates, these
two characteristics are somewhat counterbalancing. Even though
the annual mass discharge from some of the marine outfalls iq
much greater than for a typical outfall in Lake Washington
the stronger nearshore circulation more effectively disperses
the marine discharge plumes. As additional consequences T
smaller fraction of the total particulate input settles near
the greater dispersion efficiency. The net results that the
area of measurable biological impact around the Denny Way C SO
was found to be only about ten times that at CSO 023 in Lake
Washington, even though the measured mass ratio of their annual
particulate discharges was 26:1. tneir annual
q^?fef t0 the n?arshore wate^s of Lake Washington, then,
Seattle's marine littoral zones are likely less impacted bv
Rfvernestua?vrmavUbateS- ThS heavily-i^strializ^uwaSsh
River estuary may be an exception to this supposition due to
potentially more toxic wastes. A few areas (including the
Denny Way CSO) may be expected to have biological impacts
greater than those determined for outfalls in Lake" Washington
due to unusually large discharge inputs. ^nington
DISTRIBUTION AND EFFECTS OF DISCHARGE-BORNE VIRUSES
KS^0t,tA ^ the Present study and in past investigations as
well (Tomlinson et al., 1976) there are numerous bathing
beaches and houseboat communities (where swimmers are also
frequently seen) in the Seattle area that are in close prox-
imity to combined sewer outfalls. In many instances the
thar^ctatf %are " ^e Path °f Polling longshore flow
that dictates the nearshore dissemination of discharge
plumes for the neighboring outfalls. Further, the tendency
142
-------
probably presents the greatest threat to the Elliott Bay biota;
however, the salmonids and other fish and their supportive food
webs are also threatened by other inputs. Due in part to the
influence of more than 20 other CSOs and unknown numbers of
storm drains, the water quality of the bay has for years been
characterized as fair to moderate. The levels of coliform
bacteria are typically higher within the bay than outside it
(Metro data), and significantly elevated levels of metals and
PCBs have been found in the sediments of the highly indus-
trialized Duwamish River estuary (Tomlinson et al., 1976). It
is in view of this more widespread deterioration of the bay's
natural environment that evaluation and abatement of discharge
impacts on naturally productive areas seems immediately
imperative.
EXTRAPOLATION OF PRESENT FINDINGS TO OTHER OUTFALL SITES
The results obtained in the present study may prove useful as
support information for management decisions concerning CSO
and SD construction or modifications. However, the extra-
polation of this information to other CSO or SD systems should
be approached with caution. The facilities investigated
herein are representative of residential drainage basins
having a moderate to high number of overflows during a typical
weather year of about 100 cm of rainfall. CSO 023 and SD
7 may have atypical discharge levels of chlorinated hydro-
carbons for residential areas; this possibility should be
investigated, bearing in mind that the fraction of the observed
biological impacts directly attributable to these toxicants
is unknown.
CSO and SD effluent from commercial and industrial areas (using
the results presented above for Denny Way studies as an example)
might be expected to have higher levels of heavy metals and
organic toxicants, and lower concentrations of viruses and
chlorinated hydrocarbons - the relative impacts of these com-
parisons are undefined except for the higher disease potential
at the residential CSO site. Generally speaking, storm drainage
is likely to be lower in total particulates and particulate
phosphorus than is CSO effluent. As opposed to that found in
storm drainage, the organic matter discharged by CSOs contains
a large fraction of materials processed by human digestion, and
may therefore be more immediately available as an energy source
to support and enrich the biological communities of the re-
ceiving waters or sediments.
One final note of caution is that the biological impacts at-
tributable to any outfall are functions not only of concen-
tration and species of discharged pollutants, but also of
their total mass loading to a given area. Greater loading
may result in either a greater area or a greater intensity
of influence, or both. The area of impact is also dependent
139
-------
to bP rtin ? ? ? the SUimer m°nths- ^rther work needs
to be done in tracing plumes toward the beaches and in
determining the relevant viral dilutions
144
-------
annual discharge would be on the order of 1.3 x 108 g of solids.
This is equivalent to 4.5 kg per meter of shoreline per year.
For an estimated mean width of 250 m for the adjacent littoral
zone, the relative deposition rate of a uniform blanket of the
discharged solids would be 18 g/m2/yr assuming 100% settling.
The corresponding rate for deposition of SD solids (based on
data for SD 7, Table 8) would be 22 g/m2/yr, for 49 SDs in the
same area.
Were the discharged particulates in fact retained and evenly
distributed throughout the littoral zone, the benthos within
each square meter would thus annually be subjected to sub-
stantial inputs of solid pollutants and the attendant impacts.
However, the distributions of discharge particulates have been
found to be very uneven, as summarized previously. The data
in Table 12 indicate that the background-subtracted input to
the sediments immediately surrounding each outfall is on the
order of 0.5 g/m2/day, or 1.9 x 102 g/m2/yr. Within an
average-sized area of near-outfall debris (perhaps 0.8 hectares),
this rate of deposition would account for more than a third of
the total inputs calculated previously for the entire littoral
zone. Stated otherwise, perhaps 35% of the particulates
discharged by the CSOs and SDs initially settles within about
50 m of the outfalls, a total area representing approximately
10% of the littoral zone.
Accordingly, if advective losses from the lake are also considered,
it seems probable that appreciably less than half of the total
annual mass of discharged particulates remains to blanket the
remainder of the littoral sediments through processes of diffuse
spreading. Much of the solid material that initially settles
around the outfalls moves offshore soon thereafter in near-
bottom plumes that stream into the profundal; resuspension
and longshore advection may ultimately redistribute part of
the remainder.
Based on these observations and the correlating information
presented in Figures 24-26 and Table 15, the prevalent sink
for the discharge particulates in the lake seems to be the
offshore depths. The CSO- and SD-derived solids that are
transported by longshore advection have only a limited tendency
to settle in the littoral zone as indicated by the data in
Table 15 for Control Site 3. For Lake Washington then,
discharge impacts on benthic communities, as measurable by
present techniques, are probably confined to the near-outfall
areas (representing perhaps 10% of the littoral zone), and to
the contiguous portions of the profundal.
Marine Environment
As compared to the physical regime controlling the dispersion
of particulates discharged into Lake Washington, there are
141
-------
Birch, P. B. 1976.
Washington, Seattle.""""' ^' "' aissert^tion, Univ. of
,.,t
Washington, Seattle. 96 p. mesis, Univ. of
Bnnkhurst, R.O., K.E. Chua, and E. Battosingh. 1969
0[ Lak'
PP.
An Introduction to
Academic Press, New York.
146
-------
of some of these systems, such as CSO 023, to overflow with
only moderate rainfall could result in the contamination of
bathing areas by spring or summer showers that are quickly
followed by warm weather and heavy use of the swimming areas. The
potential for viral infection of swimmers under such circum-
stances is apparent, but difficult to quantify, as discussed
below.
For several reasons, our estimates of the numbers of viruses
in combined sewer discharge are probably very low: First,
the virus filtering technique involves adjusting the pH of the
sample water from 3.5 to 11.5, and this is detrimental to some
types of viruses. Second, the BGM cells used in our assays
are not appropriate for the detection of many of the serotypes
of viruses present in sewage. They are, however, excellent for
the growth of polioviruses and Coxsackieviruses which are
prevalent in sewage from this area. Third, the viruses from
some 40 of our sewage samples, as identified by a laboratory
at the University of Washington, were found to consist of only
two types - polioviruses and Coxsackie B viruses (included in
those samples were the two raw combined sewer overflow water
samples processed for the present study; all of the samples
were processed using procedures identical to those used in
this study). The polioviruses and Coxsackie B viruses
account for less than 9% of the serotypes of viruses known
to be excreted in the feces of humans, including the adeno-
viruses, echoviruses, Coxsackie A viruses, reoviruses,
rotaviruses, and hepatitis A viruses. In the Seattle Virus
Watch program of 1965-1969 the adenoviruses were the types
most frequently isolated from human fecal specimens (Cooney
et al., 1972).
It has been shown by several investigators (Clark and Chang,
1959; Joyce and Weiser, 1967; and Metcalf and Stiles, 1967)
that viruses can survive for long periods of time in water.
The most critical influencing factors appear to be the
temperature and organic content of the water (Mitchell and
Jannasch, 1969; Prier and Riley, 1967). Survival is inversely
proportional to temperature and directly proportional to the
organic content. The temperature along Lake Washington
beaches rarely exceeds 20°C in the summertime and is generally
about 18°C. It is estimated that for the interoviruses
(polio-, Coxsackie-, and echoviruses) at about 15°C there
occurs only a 2 log reduction in virus titers after 10 days
(Metcalf and Stiles, 1967). Clearly, after taking into con-
sideration the absolute number of viruses discharged into
the receiving water, the survivability of viruses in water,
the organic material discharged by an overflow, the potential
drift of the plume toward the beach, and the very low dose
required for viral infections (Beard, 1965), one must
conclude that combined sewer overflows may present a health
hazard to people using Lake Washington beaches after overflows,
143
-------
«-
. .hit.. ,„, viroloqy,
Fox, J.P., c.E. Hall and M.K. Cooney.
r^'iosf— !-r " a
Gerba, C.P., E.M. Smith and J.L. Melnick
treatment
Vol2p r reame
, Vol. 2. Pergamon Press. Pp. 675-701
Green, C.S. 1975. A comparison of diversity indices in-
'
""
Sci. 14; 397-412.
Holland, P.V. and H J Alt^r i Q7c r.
RT'St. 2'
W.L. Drew, ed. Pp. 189-212.
148
-------
REFERENCES
Anger, K. 1975. On the influence of sewage pollution on
inshore benthic communities in the south of Kiel Bay.
Helgolander Wiss-Meeresunters 27: 408-434.
Anon. 1972. Water Quality Criteria. Environmental Protection
Agency, Wash. D.C. 594 pages.
APHA. 1975. Standard Methods for the Examination of Water and
Wastewater. 14th Edition. Amer. Pub. Health Assoc., Amer.
Water Works Assoc., Water Poll. Cont. Fed., Wash., D.C.
1193 pp.
Archibald, R.E.M. 1972. Diversity in some South African diatom
associations and its relation to water quality. Water
Res. 6>: 1229-1238.
Armstrong, J.W., R.M. Thorn, K.K. Chew, B. Arpke, R. Bohn,
J. Clock, R. Hieronymus, E. Hurlburt, K. Johnson, B. Mayer,
B. Stevens, S. Tettlebach, and P. Waterstrat. 1978. The
Impact of the Denny Way Combined Sewer Overflow on the
adjacent Flora and Fauna in Elliott Bay, Puget Sound,
Washington. A report prepared in cooperation with the
Municipality of Metropolitan Seattle. 102 pp.
Austin, R.W. 1973. Problems in measuring turbidity as a water
quality parameter. Paper presented at Environmental
Protection Agency Seminar on Methodology for Monitoring
the Marine Environment. October 16-18, 1973. Seattle,
Wash.
Barnes, R.S. 1976. A Trace Metal Survey of Selected Waters,
Sediments and Biota of the Lake Washington drainage. M.S.
Thesis, Univ. Washington, Seattle.
. 1979. The Interpretation and Use of Recent
Sedimentary Geochronologies in Calculating Trace Metal
Mass Balances in Lucustrine Systems. Ph. D. dissertation,
Univ. of Washington, Seattle.
Beard, J.W. 1965. Host virus interaction in the initiation of
infection. In: Transmission of Viruses by the Water Route,
G. Berg, ed. John Wiley and Sons, New York.
145
-------
Patten, B. , W. McConnaha, and L. Callahan 107*
as1 XL? .rsr ,-a-r r
Phillip, E.l 1968. lto.hington
con,,. Of A,riCuituc«, „„,„. stat.
Mana9ement f°r the Metropolitan
e
M.S. Thesi- Seattle?
Scheffe, H. 1959. The Analysis o£ yari.^0 John wiley &
Pb'aAd Ib^Tn"^^0? RateS and Mean Residence
Pb and Pb-210 i
Lake Washington, Puqet onn
na'C°aStal Regi°n- Lab- Radiation Ecol? Dept
. Univ. Washington, Seattle. P
Schell w.R E.E. collias, A. Nevissi, and E.E. Ebbesmever
qnof; Trace Contaminants from Duwamish River DreSge
Spoils Deposited Off Fourmile Rock in Elliott Bav 1Q7K
A report for the Municipality of Metropolitan Seattle 105
150
-------
Dalseg, R.D. and C.P. Leiser. 1970. Characteristics of Storm
Sewers and Combined Sewer Overflows Discharging into Lake
Washington — Preliminary Report. A report for the
Municipality of Metropolitan Seattle. 15 pp.
Dayton, P.K. 1971. Competition, disturbance, and community
organization: the provision and subsequent utilization of
space in a rocky intertidal community, Ecological Monogr.
41.: 351-389.
Dismukes, W., A.L. Bisno, S. Katz and R. Johnson. An outbreak
of gastroenteritis and infectious hepatitis attributed to
raw clams. Am. J. Epid. 89: 555-561.
Dixon, W.J. 1973. BMP - Biomedical Computer Programs. Univ.
Calif. Press. 3rd ed. 751 pp.
Dougherty, W. and R. Altman. 1962. Viral hepatitis in New
Jersey, 1960-1961. Am. J. Med. 32; 704-716.
Elliot, J.M. 1971. Some methods for the statistical analysis
of samples of benthic invertebrates. Sci. Publ. Freshwater
Biology Assn. 25: 1-144.
Emig, J.W. 1966. Largemouth Bass. In: Inland Fisheries
Management. A. Calhoun, ed. Calif. Dept. Fish and Game.
Pp. 332-353.
1966. Smallmouth Bass. In: Inland Fisheries
Management. A. Calhoun, ed. Calif. Dept. Fish and Game.
Pp. 354-366.
English, T.S. 1964. A theoretical model for estimating the
abundance of planktonic fish eggs. Couns. Perm. Int.
Explor. Mer. Rapp. et. Proc. Verb. Reun. 155: 174-182.
Environmental Quality Analysts, Inc. 1974. Study of Wastewater
Discharge Areas at the Richmond Beach, Carkeek Park, West
Point, Alki Point, Submarine Outfalls. A report for the
Municipality of Metropolitan Seattle. 12 pp.
1975. Winter Studies of Wastewater Discharge
Areas at the Carkeek Park, West Point Submarine Outfalls.
A report for the Municipality of Metropolitan Seattle.
11 PP-
Farrah, S., C. Wallis, P.B. Shaffer and J.L. Melnick. 1976.
Reconcentration of polio virus from sewage. Appl. & Env.
Microbiol. ^2: 653-658.
Farris, G.D., R. G. Swartz and N.R. Wells. 1974. Environmental
Management for the Metropolitan Area, Part II. Appendix C.
147
-------
Ph. D. dissertation, Univ. Washington, Seattle. 237 pp.'
''
of 0
f Balefin
121 Municipality of Metropolitan Seattle.
Wallis, J.R., N.C. Matalas and J.R Slack 1974 T,,=«-
Water Resour. Res. 10(2) : 211-219 Just a moment!
Walters, L j Jr., c E. Herdendorf, L.J. Charlesworth Jr.,
n c skes! M '
Wellings, P.M., A.L. Lewis and C.W. Mountain 1976 n«m
3-
Selectio» and evaluation of methods for
conditio-
Wetzel, R.G. 1975. Limnology. w.B. Saunders, Phila. 743
pp
152
-------
Kelly, S. and W.W. Sanderson. 1960. Density of enterovirus
in sewage. J. Wat. Poll. Control Fed. 32: 1269.
Lund, E. 1973. The effect of pretreatment on virus contents
of sewage samples. Water Res. !_• 873-879.
McAfee, W.R. 1966. Eagle Lake Rainbow Trout. In: Inland
Fisheries Management. A. Calhoun, ed. Calif. Dept. Fish
and Game. Pp. 221-225.
McAfee, W.R. 1966. Lahontan Cutthroat Trout. In: Inland
Fisheries Management. A. Calhoun, ed. Calif. Dept. Fish
and Game. Pp. 225-231.
Manly, R. and W.O. George. 1977. The occurrence of some heavy
metals in populations of the freshwater mussel Anodonta
anatina (L.) from the River Thames. Environ. Pollut.
14_: 139-154.
Menge, B. A. 1976. Organization of the New England rocky inter-
tidal community: role of predation, competition, and envi-
ronmental heterogeneity. Ecological Monogr. 46; 355-393.
Metcalf, T. G. and W.C. Stiles. 1967. Pp. 439-447 in
Transmission of Viruses by the Water Route. G. Berg, ed.
Interscience, New York.
Metro/Seattle. 1976. CATAD NPDES Overflow Output, Card Code
8, 1975-1976.
Mitchell, R. and H.W. Jannasch. 1969. Processes controlling
virus inactivation in seawater. Env. Sci. and Technol.
3.: 941-943.
Morgans, J.F.C. 1956. Notes on the analysis of shallow-water
soft substrate. J. An. Ecol. 25: 367-387.
Neefe, J.R. and J. Stokes, Jr. 1945. An epidemic of infectious
hepatitis apparently due to a waterborne agent. JAME 128:
1063-1075.
Nie, N.H. 1975. Statistical Package for the Social Sciences,
Second Edition. McGraw-Hill Book Company, Inc., New York.
675 p.
Osman, R.W. 1977. The establishment and development of a
marine epifaunal community. Ecol. Monogr. 47; 37-63.
Paine, R.T. 1966. Food web complexity and species diversity.
Amer. Naturalist 100: 65-75.
149
-------
APPENDIX
TABLE A-l.
DRY WEIGHT CONCENTRATIONS OF VARIOUS
PARAMETERS DETERMINED FOR SURFACE(a)
SEDIMENTS COLLECTED AT VARIOUS
NEARSHORE LOCATIONS(b) JN LAKE
WASHINGTON DURING JANUARY, 1978
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
1
1
1
1
1
1
2
2
2
2
3
3
3
3
3
4
4
4
4
4
5
5
5
5
5
cr
8
8
8
9
9
10
10
12
12
12
12
12
13
13
13
13
13
780105
730105
780105
780105
780105
780105
780105
730105
780105
780105
771216
771216
771216
771216
771216
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
771222
780106
78O106
780105
780105
780105
780105
780105
780105
780105
780105
780105
780105
1 A 65
2 A
3 A
4 A
1 B
2 B
1 A 55
2 A
1 B
2 B
1 A 56
2 A
1 B
2 B
3 B
1 A 58
2 A
1 B
2 B
3 B
1 A 64
2 A
3 A
1 B
2 B
3 B
1 A 61
2 A
1 B
1 A 55
1 B
1 A 56
1 B
1 A 64
2 A
3 A
1 B
2 B
1 A 66
2 A
3 A
4 A
1 B
12
33
17
13
14
11
11
18
12
10
5
12
17
16
12
7
9
19
30
19
10
0.500
0. 100
0.060
0. 100
0.060
0.080
0.050
0. 050
0. 055
0.090
0.050
0.060
0.500
0.200
0.500
0.200
0. 180
0.050
0.060
0.700
0.050
0. 060
0.063
0.072
0.060
0.060
0.500
O. O70
77
64
55
32
30
43
45
110
51
42
60
61
100
54
19
42
29
84
66
10
A.C!
93
160
67
80
54
51
59
65
70
44
~~
55
65
73
69
42
54
57
84
150
140
0.15
0. 19
0.28
0.67
O. 17
0.22
0.96
0. 18
0.38
0.43
0.61
0. 39
0.53
0.27
0.55
0.37
1 . 20
1.20
1 . 30
1 Jt *-»
1
120
190
140
1
160
70
240
310
30
1
30
250
50
70
1
240
1
1
1
220
510
70
50
130
100
1
1
250
1
170
500
560
4.95 57O
640
330
7.43 30O
31 0
3. 87 290
39O
5.01 330
280
250
26 . 64 35O
350
18.21 33O
240
10.87
270
6.71 360
450
14.04 370
280
280
(continued)
10.00 190
154
-------
Shepard, M.F. and R.G. Dykeman. 1977. A Study of the Aquatic
Biota and Some Physical Parameters of Lake Washington in the
Vicinity of the Shuffleton Power Plant, Renton, Washington,
1975-1976. Unpub. ms., Wash. Coop. Fish. Res. Unit, Univ.
Washington, Seattle. 408 pp.
Smith, M., C.P. Gerba and J.L. Melnich. 1978. Role of sediment
in the persistence of enteroviruses in the estuarine
environment. Appl. Environ. Microbiol. 35: 685-689.
Snee, R.D. 1974. Graphical display of two-way contingency
tables. Amer. Statistician 28: 9-12.
Sokol, R.R. and C.D. Michener. 1958. A statistical method
for evaluating systematic relationships. Univ. Kansas
Sci. Bull. 3£: 1409-1438.
Spridakis, D.E. and R.S. Barnes. 1976. The Effects of Waste
Water Diversion On Heavy Metal Levels in the Sediments of a
Large Urban Lake. Compl. Rpt. to OWRT (Proj. No. A-070-
WASH). Dept. Civil Engineering, Univ. Washington, Seattle.
1978. Contemporary and Historical Trace
Metal Loadings to the Sediments of Four Lakes of the Lake
Washington Drainage. Compl. Rpt. to OWRT (Proj,, No. A-083-
WASH). Dept. Civil Engineering, Univ. Washington, Seattle.
Standard Methods for the Examination of Water and Wastewater.
1976.Detection of Viruses in Water and Waste Water.
14th ed. M.A. Fransen, ed. APHA. Washington, D.C.
Pp. 968-975.
Stein, J.N. 1970. A Study of the Largemouth Bass Population
in Lake Washington. M.S. Thesis, Univ. Washington, Seattle.
69 pp.
Stevens, D.P., K.S. Warren and A.A.F. Machmood. 1977. Algorithms
in the diagnosis and management of enteric diseases. XVIII.
Acute viral hepatitis. J. Inf. Diseases 135;126.
Sverdrup, H.U., M.W. Johnson and R.H. Fleming. 1942. The Oceans.
Prentice-Hall. Pp. 970-971.
Swartz, R.G., A.A. Heyward, S.F. Munger, J. C. Condon. 1978.
Virus Removal and Inactivation in Pilot Scale Waste Water
Treatment Plants at the Seattle Metro West Point Treatment
Plant. Report for the Municipality of Metropolitan
Seattle. 117 pp.
Tewari, A. 1972. The effect of sewage pollution on Enteromopha
prolifera var. tubulosa growing under natural habitat.
Botanica Marina 15: 167.
151
-------
TABLE A-2.
OUTFALL 023, STORM DRAIN 7 AND CONTROL SITE 3
CSO 023, Depth=0-0.5 cm
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
CSO 023,
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
3.35 1.54
2.87 1.85
.101 .072
39.2 19.74
168 79.2
129 71.9
5.63 .401
Depth=7-8 cm
X __S _
1.45 .246
.843 .686
.043 .021
40.2 48.3
91.5 72.8
50.5 55.2
5.56 .575
ppm Cu
Zn ppm Pb % Al
.738 * .688* .459*
(23) (23) (22)
.348 .646**
(23) (22)
.384*
(22)
% C % P ppm Cu
**
942 .096 -.060
(22) (21) (20)
.180 -.052
(21) (20)
-.152
(20)
.188
(23)
.238
(23)
.181
.(23)
*
.363
(22)
ppm Zn
**
.664
(21)
**
.674
(21)
.276
(21)
.098
(20)
*
.454
(23)
*
.415
(23)
**
.565
(23)
**
.711
(22)
*
.472
(22)
ppm Pb
**
.798
(21)
**
.821
(21)
.013
(21)
.024
(20)
**
.714
(21)
.318
(23)
.109
(23)
**
.531
(23)
it
.360
(22)
.063
(22)
*
.407
(23)
% Al
.482*
(21)
*
.436
(21)
.244
(21)
-.318
(20)
.335
(21)
.304
(21)
(continued)
156
-------
Woodey, J.C. 1972. Distribution, Feeding and Growth of
Juvenile Sockeye Salmon in Lake Washington. Ph.D. Thesis,
Univ. Washington, Seattle. 207 p.
Wydoski, R.S. 1979. Checklist of Fishes Occurring in the Lake
Washington Drainage. Interim Report 34, Coniferous Forest
Biome, Univ. Washington, Seattle (In press).
Wydoski, R.S. and R.R. Whitney. 1979. Inland Fishes of
Washington. Wash. Coop. Fish. Res. Unit, Univ. Washington,
Seattle. (In press)
Young, D.R. and G.V. Alexander. 1977. Metals in Mussels from
harbors and outfall areas. In: Coastal Water Res. Project
Annual Report. El Segundo, California. Pp. 159-165.
Zar, J.H. 1974. Biostatistical Analysis. Prentice-Hall, Inc.,
Englewood Cliffs, N.J. 620 pp.
Zimmerman, M.S. and R.J. Livingston. 1976. Effects of kraft-
mill effluents on benthic macrophyte assemblages in a
shallow-bay system (Apalachee Bay, North Florida, U.S.A.).
Mar. Biol. 34: 297-312.
153
-------
C 3, Depth=0-0.5 cm
TABLE A-2 (continued)
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
X
2.34
1.26
.080
19.6
131
86.6
6.13
C 3, Depth=7-8
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
.
X
1.27
.397
.038
7.19
51.2
20.7
5.60
•• — -^— .— ^— .
S
X—
1.30
.800
.034
10.4
31.8
47.0
.399
cm
S
.074
.396
.005
2.76
23.7
6.09
.218
1 ..
% C % p ppm cu
** ** ++
.951 .902 .982
C16) (16) (16)
.896* .951**
(16) (16)
**
.891
(16)
% c % P ppm Cu
**
.831 -.463 .406
(13) (13) (13)
-.397 .502*
(13) (13)
-.020
(13)
"~"^
ppm Zn
**
.684
(16)
**
.747
(16)
**
.645
(16)
**
.728
(16)
ppm Zn
-.197
(13)
.020
(13)
-.329
(13)
.293
(13)
••"
ppm Pb
**
.894
(16)
**
.927
(16)
**
.865
(16)
**
.885
(16)
**
.720
(16)
ppm Pb
*
.477
(13)
.368
(13)
.028
(13)
.407
(13)
-.028
(13)
% Al
-.028
(16)
.046
(16)
-.066
(16)
-.006
(16)
.135
(16)
.083
(16)
% Al
.262
(13)
.068
(13)
-.574*
(13)
-.093
(13)
.418
(13)
.052
(13)
•
(a) Refer ,to Figures 24-26 for core locations
Correlation significant at a= .05.
** Correlation significant at a= .01.
158
-------
TABLE A-l (continued)
STN
TYPE NO.
C
c
SD
SO
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
SD
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
CSO
13
13
6
6
6
7
7
7
9
9
11
11
11
13
13
14
14
16
16
17
17
17
1 703
1703
19
19
20
20
20
20
20
20
20
23
23
23
23
23
30
30
30
30
32
32
36
36
36
44
44
45
45
45
46
46
46
49
49
DATE
780105
780105
771228
771228
771228
771228
771228
771228
780110
780110
780109
780109
780109
780109
780109
780106
780106
780 1 1 0
780110
780109
780109
780109
780109
780109
780109
780109
780109
780109
780109
780109
780109
780111
780111
771227
771227
771227
771227
771227
780111
780111
780111
780111
780111
780111
780110
780110
780110
780110
780110
780110
780110
780110
780110
780110
780110
771222
771222
(C)(d) AL CU HO
R S (PPM) (PPM) (PPM)
2
3
1
2
1
1
2
1
1
1
1
2
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
2
3
1
jL.'
1
1
1
2
3
1
2
1
2
3
1
1
1
1
2
1
1
1
1
2
1
1
2
1
1
1
B
B
A
A
B
A
A
B
A
B
A
A
B
A
B
A
B
A
B
A
A
B
A
B
A
B
A
A
A
B
B
A
B
A
A
A
B
B
A
A
A
B
A
B
A
A
B
A
B
A
A
B
A
A
B
A
B
62
57
54
12
60
62
63
.8
70
68
58
24
68
56
45
48
60
56
53
58
38
50
68
40
29
33
28
51
23
17
26
10
7
11
19
29
7
9
29
35
1 30
38
17
21
30
34
27
60
2 1 60
120
54
34
38
33
44
10
10
140
51
26
48
8
64
47
0 . 500
0. 100
0.060
0.200
0. 100
0.200
0. 100
0. 100
0.200
0. 100
0. 100
0. 100
0. 100
0 . 300
0 . 200
0.200
0. 100
0. 100
0 . 200
0.400
0.200
0. 100
0 . 300
0. 100
0. 100
0. 100
0.500
0. 400
0.400
1 . 1 00
0. 160
0. 150
0. 100
0.200
0 . 300
0.050
0 . 500
0. 100
0.400
0.200
0.200
0.200
0.055
0.050
0. 100
0.090
0.500
PB
(PPM)
370
230
260
220
380
420
3 1 0
46
72
54
13
110
71
22
45
76
63
330
360
80
82
56
30
250
52
170
220
170
390
360
45
77
43
81
280
710
250
300
10
220
250
ZN TOC
( PPM ) 7.
134
1 20
130
150
150
150
110
140
31
47
68
82
140
1 50
95
1 60
1 50
410
1 80
80
120
81
70
1 00
92
1 500
720
390
160
170
240
120
96
400
560
300
120
220
1 50
180
160
0.84
0.79
1 . 1 0
1 . 70
1 . 40
1. 10
0.76
0 . 74
0.36
0.64
1 . 60
1 . 20
0. 11
0. 11
0.60
0.47
2.80
3.30
0.33
0.38
7 . 80
2 . 00
0.31
0.65
0.71
0.40
1 . 50
1.20
0.71
0.52
0.20
0.37
0.70
0.46
4.90
3.90
0 . 20
0. 29
1.10
1 . 80
O&G CH
(PPM) (PPB)
380
1400
790
2000
3500
1 800
1800
200
210
50
1
220
240
130
80
510
600
2600
3600
1 000
840
290
210
220
410
1 700
1700
730
980
2500
1 500
350
260
110
1 90
1 500
1 1 00
2800
3600
1
50
720
2100
132.60
118.47
4 1 . 46
18.74
5.57
18. 30
7.21
23.73
455.65
25.65
32.75
28.80
106.75
113.67
32.26
18.01
35. 73
1 03 . 78
4.76
42.73
TP
(PPM)
1 60
360
400
380
370
330
320
310
290
280
210
230
500
500
1 50
400
380
330
550
610
6 1 0
390
420
590
580
440
370
410
3900
4900
1 500
560
620
500
540
820
420
360
1500
600
490
530
1200
300
1 200
820
(a) Surface 3 cm, collected by SCUBA divers.
(b) Reference Figures 2 and 8.
(c) Replicate.
(d) Sample.
155
-------
carnivore - a flesh-eating animal.
of a subtaxon of the
techni
-------
TABLE A-2 (continued)
SD 1, Depth=0-0.5 cm
§ _ % C % P ppm Cu ppm Zn ppm Pb % Al
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
2-.08
1.09
.062
46.7
60.9
89.2
6.07
**
.818 .956
(17)
.804
.033
25.7
27.0
52.8
.328
** *
.859 .465
(17) (17)
** *
.827 .440
(17) (17)
.360
(17)
**
.844
(17)
**
.893
(17)
**
.828
(17)
.246
(17)
**
.850
(17)
**
.920
(17)
**
.729
(17)
**
.596
(17)
**
.826
(17)
.072
(17)
.094
(17)
.273
(17)
-.130
(17)
.028
(17)
-.010
(17)
SD 1, Depth=7-8 cm
Wet Wt/
Dry Wt
% C
% P
ppm Cu
ppm Zn
ppm Pb
% Al
X
1.46
0.99
.039
41.7
35.3
46.2
5.97
S,, % C
**
.181 .596
(17)
.781
.008
23.7
26.1
33.3
.286
% P ppm Cu
**
.670 .266
(17) (17)
.403 .134
(17) (17)
.236
(17)
ppm Zn
.239
(17)
.049
(17)
**
.715
(17)
-.069
(17)
ppm Pb
.002
(17)
-.134
(17)
.440
(17)
.295
(17)
**
.785
(17)
% Al
*
.554
(17)
.356
(17)
**
.800
(17)
-.121
(17)
**
.726
(17)
.402
(17)
(continued)
157
— mi cuumcu.
in one short moist season.
extrapolation - Estimating a function at a point which is
larger than (or smaller than) all the points at which
the value of the function is known.
feeding type - A grouping of animals based on a common
mode of feeding, such as filtration and ingestion of
suspended particulates from the surrounding water.
flux - The amount of some quantity flowing through a given
area per unit time.
•f- Vi
geometric mean - The n root of the product of n given
quantities, a calculated average for which the
influence of extreme values is less than for the
arithmetic mean of the same set of values.
grab sample - A sediment aliquot collected by a device
lowered from the water surface and triggered to take
a bite of the underlying substrate with mechanical
jaws.
161
-------
grain size distribution - A quantitative analysis of the
spectrum of particle sizes comprising a sediment "
meters'
infauna - Aquatic animals which live in the sediment
underlying a body of water. s»eaiment
inorganic - Pertaining to or composed of chemical compounds
that do not contain carbon as the principal element
(excepting carbonates, cyanides, and cyanates) , that is,
matter other than plant or animal.
intertidal zone - The zone between the high-tide and low-
tide marks .
isopleth - A line of equal or constant value of a given
quantity with respect to either space or time!
lens - A thin layer of surface water of relatively limited
dimensions and having properties distinct from those
of the water body beneath.
light transmission - The process in which light travels
without bein9 absorbed
betweene^L°^°^ per^nin9 to the biogeographic zone
between the high- and low-water marks.
macroalgae - Large, conspicuous varieties of algae.
median -The quantity or value of that item which is so
positioned in a series, when arranged in order of
n^f1Cai ?i?antit^ or val"e, that there are an equal
Sa^itude S greater ^gnitude and lesser
metric ton - A unit of mass equal to 1000 kilograms or to
approximately 2204.6 pounds.
the !^ef ?rga"ism of one of ^e divisions of phyla
i^ St klngdom Containing clams, mussels, oysters,
ils, slus oc
slugs, octopuses, and squid.
morphology - A branch of biology that deals with the form and
structure of animals and plants. Also, the external
structure of rocks in relation to the development^
erosional forms or topographic features. °pment o±
162
-------
nematode - A member organism of a group of segmented worms
which have been variously recognized as an order,
class, and phylum.
oligochaete - A member organism of a class of the phylum
Annelida, which includes worms that exhibit both
external and internal segmentation.
organic - Of chemical compounds based on carbon chains or
rings and also containing hydrogen with or without
oxygen, nitrogen, or other elements.
particle size distribution - See grain size distribution.
patchiness - A term used to describe the spacial
variability in terms of numbers or biomass of
organisms comprising an aquatic bottom community.
pelecypod - A member organism of a large class of the
phylum Mollusca containing the clams, oysters, and
other bivalves.
periphyton - Sessile biotal components of a freshwater
ecosystem.
pollutant loading - The time-integrated mass of a pollutant.
polychaete - A member organism of the largest class of the
phylum Annelida, distinguished by paired, lateral,
fleshy appendages on most segments.
profundal - The region occurring below the open water zone
and extending to the bottom in lakes deep enough to
develop temperature stratification.
r-selected - Organisms that are adapted for having a high
growth rate, in terms of numbers or biomass.
recruitment - The settling and attachment of immature
aquatic organisms.
regression analysis - Given two dependent random variables,
regression functions measure the mean expectation of
one relative to the other.
serotype - A serological type of intimately related micro-
organism, distinguished on the basis of its antigenic
composition.
163
-------
skewness -A measure of assymetry in a statistical density
function describing numbers of counts vs. a
characteristic expressed on a continuous scale.
sorting - The^process by which sedimentary particles
similar in size, shape, or specific gravity are
selected and separated from associated but dissimilar
particles by the agent of transportation.
species richness - A numerical expression of the number of
species in a collection.
storm drain - A drain which collects and conducts storm
runoff water from rain-incident surfaces and inter-
stitial soil seepage into a combined sewer, or a
receiving water.
stratification - The arrangement of a body of water, such as
* t* ' ° tw° or more horizontal layers of
differing characteristics, especially densities.
subtidal zone - The region of the bottom of a tidally
influenced body of water below the Mean Low Low
water mark.
taxa - Classified groups of biological organisms.
transmissometer - An instrument for measuring the extinction
coefficient (light transmission) of water.
164
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-111
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
FATE AND EFFECTS OF PARTICULATES DISCHARGED BY
COMBINED SEWERS AND STORM DRAINS
5. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
?.AUTHOR(s)
D. TomTinson, B. N. Bebee, A. A. Heyward,
S. G. Munger, R. G. Swartz, S. Lazoff, D. E. Spryidakis,
8. PERFORMING ORGANIZATION REPORT NO.
9.
NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Municipality of Metropolitan Seattle
821 Second Avenue
Seattle, Washington 98104
35 B1C
11. CONTRACT/GRANT NO.
R805602010
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin.,0h
Office of Research and Development
US Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 10/77 t.n 7/7Q
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: John N. English telephone(513/684-7613)
16. ABSTRACT
This report provides the details of an evaluation of the distribution and
biological impacts of particulate materials in combined sewer and storm
drain discharges in the Seattle, Washington region, and presents the
extent of the urban runoff problem in terms of statistics and observed
and anticipated impacts on water quality in Lake Washington and Puget Sound.
The potential public health risk related to enteric viruses associated with
such particulates is also addressed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
*Combined sewers
Urban runoff
*Sediments
*Particulates
Water Quality
Virus
*Storm sewers
Benthos
Toxicity
Metals
Hydrocarbons
Lake Washington
Puget Sound
Nutrients
Oil and grease
Light transmission
13B
18. DISTRIBUTION STATEMENT
release to public
19. SECURITY CLASS (ThisReport)
unclassified
21. NO. OF PAGES
183
20. SECURITY CLASS (Thispage)
unclassified
22. PRICE
EPA Form 2220-1 (9-73)
165
U.S. GOVERNMENT PRINTING OFFICE: 1980--657- 165/0120
-------
-------
to the sites closer to the CSO than it was to the sites
further away. All taxa encountered entered into the dis-
criminant function, with differences in the cover of barnacles
being the most significant. As reported by Armstrong et al.
(1978), the abundance of Fucus increased, and the abundance
of Sargasam decreased, with increasing distance from the
CSO~. The cover of the filamentous red alga Polysiphonia sp.
appeared to decrease nearer the CSO. Another delicate red
alga, Delesseria decipiens appeared to increase in abundance
nearer the CSO.
The data on low littoral communities from August differed
considerably from those collected in April. Fewer species
were found and the cover was generally less in August. Thorn
(1978) has shown that the number of species of benthic algae
in central Puget Sound is greatest in spring. Coinciding
with the onset of increasing desiccation pressure from high
air temperature and very low daytime low tides is a decline
in number of species. The most notable result from the
August samples was the low cover of algae at the sites
nearest the CSO (Figure 44). Ephemeral algae, abundant in
spring, had probably reproduced and died by that time. Open
space was not colonized due to stress from desiccation.
Cover was higher at the sites beyond the two closest to the
CSO due to the presence of the desiccation-resistant, perennial
algae Fucus, Gigartina papillata, and Iridaea cordata.
S_. Mutic, abundant in April, was no longer present near the
CSO. Barnacles, diatoms and Ulva were the most abundant
organisms there at the time of this study. Flows were
substantially reduced in summer, and this may be the reason
why diversity at the sites 120 m north and 150 m south appears
to have been unaffected (Figure 44).
The plot of sites along canonical variables one arid two for
August may be related to the cover of barnacles (Figure 45).
The relative positions of sites 380 m N, 30 m S arid 30 m N
in the two-dimensional space of the figure indicates similarity
in terms of extent of rock coverage. No pollution gradient is
readily apparent from these relationships.
132
-------
Boulder Wall Taxa — The majority of cover at +0 m (MLLW) along
the rocky portions of the study area was comprised of macroalgae
The sites 30 m north and south of the CSO were primarily cobble
in sand. The remainder of the sites were the stable boulders
forming the wall.
Species-richness curves for macroalgae at in April were
lowest at the sites located 150 m south and 120 m north of the
CSO (Figure 44). The sites closest to the CSO held more
species than the former two sites but less than the remaining
sites further from the CSO. Recent ecological work concerning
disturbance versus species diversity may provide a partial
explanation for this condition. It has been shown that dis-
turbance of rocky intertidal areas may raise species diversity
by opening primary space for colonization by fast growing
and reproducing (r-selected) organisms (Paine, 1966; Dayton,
1971; Menge, 1976; Osman, 1977). High flows from the CSO in
the period prior to the April survey may be regarded as a
disturbance. The species inhabiting the low intertidal
zone, especially during spring, are primarily ephemerals (r-
selected). Therefore, relatively high disturbance of cobble
sites near the CSO may have caused the increased diversity
there. The lowered diversity at the stable boulder wall sites
at 120 m north and 150 m south of the CSO may indicate the
effect of sewage-related, water-borne materials other than
251
10-
5-
800m N
380m N
00m
100
200
300
20
15-
10-
5-
150mS
600m N
120m N
800m N
(b) AUGUST
400 0
POINT COVER
100
200
300
400
Figure 44.
Species-richness curves for macroalgae collected
near the Denny Way Regulator outfall during April
and August, 1978.
130
-------
201
Figure 42.
380m N
150 300 450
NUMBER OF INDIVIDUALS
600
Species-richness curves for periphyton
samples collected near the Denny Way
Regulator outfall during April 1978.
150m S
•
140m N
^ 3.8-
oc
!.i«;
O
g-28.2-
" 44.2-
15
800m N 260m N
9
380m N
16m S
\ \ \ 1 \ \ 1 1 1 i I I I
D.9 -86.9 -22.8 41.2
9
20m N
105.3 IS'
CANONICAL VARIABLE 1
Figure 43. Position on canonical variables of periphyton
samples collected near the Denny Way Regulator
outfall during April 1978.
128
-------
Polychaete feeding strategies are a reflection of characteristics
of the sediment and overlying water (Jumars and Fauchald,
1976). Armstrong et al. (1978) showed that burrowing,
deposit-feeding polychaetes were proportionately most abundant
near the CSO. The results of the present study coincide
with those findings (Figure 41). At the control sites on
transect 1, the three major feeding types were in approximately
equal abundance. This result was true for both the April
and August samples. The differences in proportions of
feeding types among sites were generally greatest during
April. Significant seasonal differences were seen at sites
on transects 4, 5, 6 and 7, which again indicates the influence
of the CSO. Samples at sites on transects 6 and 7 in August
contained large numbers of the surface deposit-feeding
polychaete Prionospio steenstrupi, which accounted for their
position on Figure 41a. At 13 m, there were no significant
differences (confidence limits not plotted) between samples
from sites near the CSO and those further from the CSO
within the same month (Figure 41b). The samples from April
tended to have proportionally more burrowing deposit feeders.
This result was true for control sites (i.e., transect 1)
and sites near the CSO, which suggests that natural seasonal
changes are the prominant source of variation at 13 m.
Intertidal Soft-Sediment Community — A zone of black sediment
with a distinct H2S odor exists near the mouth of the CSO. This
zone appears to be intensely scoured by overflows and this is
reflected in the relatively low organic content of the sediments
and the high abundance of oligochaetes (Table 28). The poly-
chaete Capitella capitata was the most abundant infaunal species
in the intertidal zone. Its abundance appears to have been
related to the CSO, and this is best seen by comparing sites e,
f and g in the CSO cove with sites h, i and j located at the
same tidal height near transects 2 and 3 (Figure 5). This
difference was evident in April but not in August. Site C
located in the wash-out zone still held a large number of
individuals of C. capitata in August, although the overall
abundance of this species was reduced significantly. Intense
desiccation stress and decreased overflows may be the reason
for the reduced abundance in August.
Periphyton — Data on periphyton communities are civailable for
April only. Blocks set out in July were covered with a dense
mat of the green bladed alga Enteromorpha linza in August,
thus excluding any microalgal assemblage developmemt.
There was no clear evidence of a reduction in periphyton
diversity except on the block closest to the CSO (Figure 42).
These results differ from those of Armstrong et al. (1978)
who did see an indication of declined diversity further from
the CSO.
126
-------
transect 1. Furthermore, in the present study, site 9/1
clustered with site 9/4. This suggests that the influence of
the overflow in April, 1978 may have extended further to the
north.
The cluster analysis of the data from August produced strikingly
different results from those obtained in April. A depth
effect was less evident in August (Figure 39b). The assemblage
at 13 m on transect 4 was unique as was that at 9 m on transect
2. The effect of the CSO was most pronounced again at
transect 5 at 9 m with an indication of some modification on
transects 6 and 7 to the south. The poor correlation
between the results from April and August suggest that flow
levels from the CSO does affect the subtidal shallow infaunal
assemblage.
A graphic display of the proportions of individuals, among
the three major infauna phyla found at each site provides a
partial indication of the influence of the CSO (Figure 40).
Infauna composition at site 9/5 was significantly different
(p <0.05) from all other sites during both April and August.
These two samples were also significantly different from one
another. A very high proportion of the individuals in April
were polychaetes (primarily Capitella capitata). Conversely,
a large number of crustecea (primarily Nebalia pugettensis)
was found at this site in August. The samples on transect 1
(control) were significantly different from one another, but
they did exhibit an infaunal composition that was relatively
higher in mollusc individuals and lower in numbers of polychaete
individuals. The sites closer to the CSO had increased
ratios of annelids (polychaetes) to molluscs.
Seasonal changes were evident in the data from 9 m. Samples
from the same site were significantly different between
samplings except at transect 2. The greatest difference
(i.e., as measured by distance between points on the graph)
was generally between sites on transects 4, 5, and 6. If
seasonal changes observed at the control transect (1) can be
regarded as natural, it can be concluded that the effluent
from the CSO tends to increase between-season differences in
the infauna at 9 m.
The graphical analysis of samples from 13 m in April revealed
that there were no significant differences between sites
close to the CSO and the control sites (Figure 40b, confidence
limits not plotted). However, the August sample from transect
4 was disproportionately high in polychaetes and may suggest
a possible influence of the CSO at this depth. No significant
differences between seasons were evident at any 13 m site
other than that on transect 4.
124
-------
50
40
30
20
(a) APRIL
10-
i 50
8
5
40
30
20
10-
13/7 13/6 13/3 13/2 13/1 9/7 9/6 9/3 13/5 92 91 9/4 9/5
(b) AUGUST
13/113/3 9/1 13/6 .4/2 9/4 13/5 13/7 9* 13/4 9)2 9* 9 r^—
Figure 38. Dendrogram of'subtidal infauna samples
™iie??e* n®ar the Denny WaY Regulator
^f^Ld^^?_A?ril and Au^st; 1978.
122
-------
100 150
NO. OF INDIVIDUALS
250 275 368
Figure 36. Species-richness curves for molluscs
collected near the Denny Way Regulator
outfall during April and August, 1978
The numbers indicate sampling transects.
120
-------
600-
9AUGUST
(a) INDIVIDUALS AT 9m DEPTH
(b) INDIVIDUALS at 13m DEPTH
(d) TAXA AT 13m DEPTH
Figure 35. Density of infaunal individuals and taxa
found in the subtidal samples collected
near the Denny Way Regulator outfall
during April and August, 1978
118
-------
Community Distribution Analyses -
area was located in sandy sediments with
no
27)
transect 6 anH 7 =e at sites on
transect. s, b and 7 in August.
TABLE 27. CHARACTERISTICS OF SUBTIDAL SEDIMENTS SAMPLED (a)
NEAR THE DENNY WAY REGULATOR, APRIL AND AUGUST, 1978
Date
April 13-14
August 22
Depth
Transect (m) Texture Color
1 9
13
2 9
13
3 9
13
4 9
13
50
9
13
6 /-,
9
13
7 9
13
1 9
13
2 9
13
3 9
13
4 9
13
5Q
y
13
6n
9
13
7 9
13
sandy It. brown
it ||
fine silt black
brown
It. brown/black
n |,
11 M
black
11 n
It. brown/black
» n
„ (I
sandy brown
11 n
fine silt green/black
»
n „
" n
n „
black
It. brown/black
n „
n „
Volatile
Organics
Odor ( % }
none
mod. H2S
slight H2S
n
ii
11
mod. H2S
"
slight H2S
II
none
none
slight H2S
II
II
II
11
strong H2S
slight H2S
II
II
\ ** /
1.52
0.85
2.02
2.25
2.39
2.26
11.73
2.38
3.26
3.26
2.63
7.80
2.77
2.24
1.11
2.26
2.06
1.84
" *
2.72
1.90
2.60
2.57
5.54
2.22
1.39
3.18
2.42
(a) Duplicate samples collected at all sites.
116
-------
TABLE 25.
METALS IN
AT FOUR SXTES NEAR
THE DENNY WAY REGULATOR
Station Parameter
120 m N
380 m N
800 m N
Cu
Pb
Zn
Cu
Pb
Zn
Cu
Pb
Zn
6.5
5.2
287.5
6.5
5.2
287.5
6.5
5.2
287.5
21.0
34.0
205.0
20.0
22.0
274.0
14.0
18.0
286.0
+14.5
+28.8
-85.5
+13.5
+16.8
-13.5
+ 7.5
+12.8
- 1.5
TABLE 26. TISSUE BURDEN OF HEAVY METALS IN
OYSTERS KEPT IN POTS AT FOUR SITES NEAR
THE DENNY WAY REGULATOR OUTFALL, 1978
Feb 7
Station Parameter X
Apr 24
34 m N
120 m N
380 m N
800 m N
Cu
Pb
Zn
Cu
Pb
Zn
Cu
Pb
Zn
Cu
Pb
Zn
100.8
2.2
2960.*0
100.8
2.2
2960.0
100.8
2.2
2960.0
100.8
2.2
2960.0
31.45
0.65
705.99
31.45
0.65
705.99
31.45
0.65
705.99
31.45
0.65
705.99
176.7
29.3
2903.3
240.0
8.1
3750.0
150.0
6.7
2906.7
260.0
6.5
3710.0
17.64
2.03
619.85
20.82
1.38
241.32
47.26
3.18
854.91
23.09
2.13
127.67
+ 75.9
+ 27.1
- 56.7
+139.2
+ 5.9
+790.0
+ 49.2
+ 4.5
- 53.3
+159.2
+ 4.3
+750.0
(a) mg/kg dry weight.
(b) Standard error.
114
-------
100
80
X
UJ
Q
Z
z
o
60-
40-
20-
MUSSELS
^ X
OYSTERS
CSO COVE AT 34m N
120m N
FEB MAR APR MAY JUN FEB MAR APR MAY JUN
100
80
X
LU
Q
Z
1 60H
Q
Z
O
u
40-
20
380m N
(b)
800m N
Figure 33.
FEB MAR APR MAY JUN FEB MAR APR MAY JUN
Variation in the condition indices of
mussels and oysters kept in pots at four
sites near the Denny Way Regulator outfall
(a) pot lost; (b) reproductive condition
112
-------
91.9
25
DISCHARGE
D SURFACE SEDIMENTS
(1 cm depth)
SITE SAMPLED TWICE
Figure 32. Summary of particle size distribution
analyses for the Denny Way Regulator.
110
-------
theory) . Figure 29 depicts light transmission distributions
in the surface waters following one complete tidal flood and ebb
e^Le^nf10"'-?^633 FigUre 3° shows the influence of an
ebb tide only, wlth an appreciably greater strength of flow
^oVvaTu^of^iahtT' thS discha^e P1™*' characterized by
low values of light transmission, was found only to the north
of the outfall, and to depths of about 6 m. The heavy
particulate load of the river was confined to the freshwater
ih£ ^^^ ^ SUrfa°e and about 4'5 m dePth- " is app^nt
that the discharge plume effectively excluded the river water
from the nearshore area to the north of the outfall^in the
direction of the prevailing tide. Light transmission distri-
butions plotted for data collected more than one half of a
id^ KY
-------
IH
CO
O
•P
C
O
U
CO
0)
-H
-p
u
•o
C 14
3 O
O 4J
a C 0)
W O tf
Ci -rJ
,
a CQ fd
•H £
fd w >i
•H G c
M td C!
-------
0)
X!
•H 4J
00
C vo
I
O O r*
O rH
0) in
O '-P I
I
tQ TJ M
S -P
•H O fO
-P M M
U rtf 3
a) tr>
O. C (U
CQ O «
Cj »j™J
(U W >i
a, m ra
•H is
•H g
fO CO >,
•H C C
M r3 CJ
(U ^-l
•H
104
-------
0) I
JZ -H
r^ -H TJ
0) cr» CO
W rH CU >i
o x: ^-H
& O CO fQ
CO M rH -H
I-P
O O co
rH M 0)
••w MH
O
H
13 4J ^| -H
G 3 -P TJ
rd o
0) M
r-H >-l 45 O
r-H O EH *4H
rtJ 4->
M~l rd T3
fl rH t (U
•H 3 0> M
nj tJ>r^ o
M a) a» 4J
« rH-H
m c
C) >i O
fd >i g
>i^ M
M
CO
i 3 6 o
E| C M >-i -H
Ei C jQ o -P
"! (U 0) -P 3
M Q fa CO ,Q
CD
CN
0)
tn
•H
dO SQNVSnOHl) 3WniOA M01d«3AO 1V1O1
102
-------
TABLE 22.
STATISTICAL SUMMARY OF ESTIMATED
POLLUTANT LOADS AND CONCENTRATIONS IN
STORM DISCHARGES MONITORED AT THE DENNY
WAY REGULATOR, MARCH AND OCTOBER, 1978
— __
MASS RANGE:
Suspended Solids
Total Cu
Total Hg
Total Pb
Total Zn
Total Al
Total Organic C
Total P
Total 0 & G
Particulate C1HC
MEAN PARTICULATE MASS-
Suspended Solids
Cu
V-» L*
TT — -
Hg
Pb
•t ±J
Zn
£j 1 1
Al
xl_L
Organic C
p
i
O & G
C1HC
•
MEAN DISCHARGE CONCENTRATION-
Suspended Solids
Total Cu
Total Hg
Total Pb
Total Zn
Total Al
Total Organic C
Total P
Total 0 & G
Particulate C1HC
— • .
(a) x + la
(b) Not applicable.
(c) Not determined.
— — _
kg
1960 - 2120
.954 - 1.56
.0073 - .0135
5.45 - 6.78
4.40 - 4.63
54.4
432
17.1 - 21.7
204 - 310
1.25 - 120 ma
Percent (a^ of
Total Mass
100.0
72.1 + 0.4
70.6 + 6.0
92.7 + 4.1
54.6 + 20.0
94.9
55.6
48.7 + 13.4
ND(C)
J.N A-/
(mg/i) Range
129 92.4 -
-077 .074 -
.0006 .0006 -
.385 .306 -
•285 .187 -
2.62 ND
20.8 ND
1.23 .985 -
16.0 14.6 -
.001 ppb .000 -
=:
No of
i-N V-J • v-J J_
Storms
(2)
\ *" /
(2)
\ ** /
(2)
\ ** /
(2)
\ ** /
(2)
\ ** f
(1)
\ .*_ /
(1)
\ -1- /
(2)
\ ** /
(2)
\ *• /
(2)
— •I
No. of
S to Tmc:
*— ' U» V— / JL 11LO
(2)
(2)
\ *• /
(2)
(2)
(2)
(1)
(1)
(2)
1 ' i I. •
of la
181
.080
.0007
.486
.437
1.52
17.4
.013
1
•^"•^-•-^™-^™™»^»«,
100
-------
indicating a reduction in community stress concomitant with
the seasonal decrease in discharge.
These observations for chironomids imply that the area of
visible debris accumulation around an outfall exerts a detri-
mental influence the year around on some organisms. Although
it does not strictly constitute a dead zone, the numbers of
individuals of its sensitive resident species are depressed
and invariate relative to ambient conditions. The net seasonal
change in numbers from reproduction, mortality and recruitment
is minimal. Outside of this zone of maximum deposition lies
a region that may be seasonally influenced by the discharges;
there, the potential for a rapid return of the affected species
to ambient levels is much greater. Strictly speaking, with
additional data the linear and log-linear relationships of
Figure 27 might be revised to include a slope break to define
the two zones of influence.
Copepods (Harpacticoida, i.e. benthic microcrustaceans) also
seemed to be negatively impacted by discharges from CSO 023
CSO 044 and SD 19. The general tendency at these stations
was for the copepod populations to decrease between the wet
and dry seasons, perhaps on response to natural cycles.
However, in each instance the seasonal decreases observed
within the discharge debris were appreciably greater than
those outside of it (SD 19 experienced copecod population
increases away from the outfall). The net effect of these
changes between February and September was to intensify the
positive correlation of numbers vs. distance from each outfall
A clear interpretation of the more specific nature of the
discharge impacts relative to these concurrent phenomena was
not apparent. As was observed for the chironomid populations,
the trends for seasonal changes of numbers of copepods were
unusual at SD 7, perhaps due to losses from near-outfall
scouring.
The tendencies noted for the communities of nematocles (round-
worms) near the outfalls were extremely variable, ranging from
significantly positive at SD 19 in February to significantly
negative at CSO 023 in September. The variable response of
nematodes to discharges may be due to a wide range of species
tolerances to the particulate toxicants. In general, the
numbers of pelecypods (freshwater mussels) varied little and
unsystematically with distance from the outfall, from one
station to the next.
98
-------
As opposed to the data reviewed for CSO 023, the copepods and
nematodes showed no significant response to discharge from
CSO 044 during February, and the nematodes had a reversal of
correlations during September. Noting that only positive
relationships between nematode counts and distance from the
outfall were seen at CSO 044, SD 7 and SD 19, the negative
response determined for both seasons at CSO 023 constitutes
an anomaly with no obvious explanation.
It is of interest to note that count and weight measurements
for the total population of benthic infauna at CSC) 044 show no
net correlation with distance from the outfall. For both
seasons tested, the positive and negative relationships noted
for the various taxonomic groupings were collectively nullifying
in algebraic summation. It is apparent that the information to
be derived from measurements of the total population is
potentially misleading.
SD 7 — Correlations with both distance from the outfall and
depth were also calculated for numbers of infaunal organisms
and biomass determined for the two storm drains, SD 7 and SD 19.
For reasons mentioned previously, i.e., localized substrate
differences, near-outfall scouring effects, and the relatively
poor juxtaposition of the discharge plumes with the sampling
arrays, this information (Tables 17-20 and Figure 27) must be
interpreted with care. The substrate and scouring problems had
the most obvious impact on the statistical results. These
effects reduced the number of organisms found at the sampling
site closest to the terminus of each outfall. These aberrations
in turn influenced the slopes of the counts vs distance and
weight vs distance relationships by intensifying the positive
correlations and weakening the negative ones. Accordingly,
further interpretations of these relationships were supplemented
by an inspection of the raw data. Compared to the coefficients
of determination calculated for the two control sites, the
values derived for both of the storm drains indicate an
"unnatural" moderation of the strength of the counts vs depth
and weight vs depth relationships. Whereas the controls had
a statistically strong increase in numbers and biomass with
decreasing depth, these relationships were less definite at
the storm drains; as detailed below, this is likely due to the
physical coincidence of decreasing depth with increasing
proximity to the outfall and the influence of its associated
discharge particulates.
Collectively, the available evidence for the distance relation-
ships (including an inspection of the raw data to identify and
compensate for the aforementioned scouring impacts) indicates
the following, which differs in a few instances with the
picture presented by Figure 27 and the statistics given in
Tables 17-20: the numbers of chironomids and oligochaetes were
moderately enhanced by the discharge (negatively correlated
96
-------
DISTANCE
NON-SIGNIFICANT CORRELATIONS
'2'° COUNTS: Ch
WEIGHT: O (Co. N, P NOT TESTED)
20 DISTANCE (m)
-20-L. -2.0
NON-SIGNIFICANT CORRELATIONS
COUNTS: O
WEIGHT: Ch, O (Co, N, P NOT TESTED
COUNT WEIGHT Img)
WEIGHT (mg)
NON-SIGNIFICANT CORRELATIONS
COUNTS: NONE
10-L 1.0 WEIGHT: O (Co, N, P NOT TESTED)
NON-SIGNIFICANT CORRELATIONS
COUNTS: O. P
10-i- 1.0 WEIGHT: O, T (Co, N, P NOT TESTED)
Figure 27. (continued)
94
-------
TABLE 20.
Independent
Variable
Taxonomic
Group
Sampling Area
C3 C4 CSO 023 CSO 044 SD 7 SD 19
Depth
Distance
Chironomids (-).15 (-).0002
Oligochaetes (-).08 (-).02
Copepods
Nematodes
Pisidium
Total (-).21 (-).15
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
(-).£2 (-).Ol
(-)-06 .04
.02
(-).Ol
02 .17 (-).02
(-).004
Depth+Distance Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
.11
.02
.06
.18
Minus signs indicate negative correlation.
Underlined values represent nonsignificant (a= .05) correlations.
.02
.18
.01
29
04
09
92
-------
TABLE 18. COEFFICIENTS OF DETERMINATION (r2) FOR LINEAR
AND MULTIPLE REGRESSIONS USING TOTAL ORGANISM
COUNTS IN SEPTEMBER AS THE DEPENDENT VARIABLE
Independent
Variable
Depth
Distance
Depth+Distance
Taxonomic
Group C3
Chironomids (-) .10
Oligochaetes (-).71
Copepods (-) .19
Nematodes (-) .46
Pisidium .01
Total (-).48
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
Chironomids
Oligochaetes
Copepods
Nematodes
Pisidium
Total
Sampling Area
C4 CSO 023 CSO 044 SD 7
(-•) -01 (-).08
(-).25 (-).07
(-).Ol .04
(-).52 (-).IO
.16 NA(a)
<-).39 (-).13
-10 .27 (-).15
(-).0003 (-).39 (-).02
•25 .26 (-).OOl
(-).32 .12 .02
.18 (-) .07 (-) .15
(-).05 .01 (-).17
.16
.07
.13
.17
ND
.17
SD 19
(-).Ol
.02
.02
.001
.11
.01
.28
.04
.21
.07
NACa)
.22
.43
.04
.22
.09
11
.25
(a) No association.
(b) Not determined.
Minus signs indicate negative correlation.
Underlined values represent nonsignificant (a= .05) correlations
90
-------
To compare the relative station-to-station strengths of the
depth and distance regressions, coefficients of determination
were calculated (C.D. = r2, where r is the correlation
coefficient). These values, presented here in Tables 17-20,
represent the proportion of the total variation in organism
counts or weights that is explained by each fitted regression.
The correlations are all significant at the 95% confidence level
unless otherwise noted.
Further, plots of the distance regression data were prepared to
show the net change in numbers of organisms and biomass per
core (means of 6 and 4 cores per site, respectively) as functions
of distance from the CSO and SD outfalls (Figure 27). Separate
curves were plotted for all significant relationships of the
various taxonomic groups for the February and September sampling
periods. Because no samples were collected directly at the ends
of the storm drain outfalls, those values were estimated by
extrapolation.
As judged by the discharge plume turbidity assessments and
sediment contaminant distributions, the two CSO stations are
more appropriately assessed than are the SDs by the hypothesis
that discharge effects are a function of distance from the
outfall. Although there are indications that lighter parti-
culates and their associated contaminants may settle in con-
centrated deposits remote from an outfall, the outfall-
surrounding orientation of the CSO sampling arrays is believed
to have been reasonably appropriate to the hypothesis. However,
because of the on- or near-shore configurations of the SDs,
which prohibited anything but semi-circular, offshore sampling
arrays, a larger fraction of the discharge particulates bypassed
those sampling grids by longshore advection. In addition,
there was evidence of substrata differences and physical
scouring effects at the near-outfall sampling stations at both
SDs. These conditions strongly affected the results of the
distance regressions, as described below. With regard to the
regressions themselves, the typically low coefficients of
determination (r^) for all stations also reflect the natural
patchiness of the benthic communities and the interference of
discharge plumes from neighboring outfalls.
Collectively considered, the data summaries offered in Tables
17-20 and Figure 27 serve to define the nature and intensity
of the effects of wastewater discharge on the near-outfall
biota. Taken station by station, these impacts were determined
to be as follows:
CSO 023 — In order to assess the relationship between the
biotic parameters and distance from either CSO outfall, zero
distance correlation was assumed for the control sites, implying
that any significant correlation determined for a CSO at
88
-------
designed for six cores per 0.5 m2 grid placement, this number
representing the best balance of acceptable statistics with
minimal cost and effort. Therefore, six cores were collected
from each of 12 locations at the control sites, 13 around the
SDs and 20 around the CSOs. The sampling locations at the
control sites represented varying depths; those around the
outfalls represented variations in both depth and distance
from the discharge point.
In most cases, the statistical confidence derived from using
the 38 mm core tubes was not substantially greater than for
the 29 mm tubes. In almost half of the comparisons, the
confidence was even greater using the smaller tubes. Since
the total amount of sediment (and the relative analytical
effort) associated with the 29 mm tubes is only 58% of that
affiliated with the 38 mm tubes, the smaller cores were
selected for use in the intensive studies.
Abundance and Biomass Comparisons — The principal objective
of this portion of the project was the determination of the
extent to which CSO and SD discharges might be affecting
benthic organisms in the lake. This was perceived to be a
difficult undertaking due to potential concurrent influences
of natural factors, including depth, water temperature, currents
light and type of benthic substrate. Of these variables, depth
is the most easily measured, and perhaps the most consistent
with regard to documented, strong influences on freshwater
benthic communities; it was, therefore, used to assess
differences between areas in terms of relative number of
organisms and biomass, i.e., it was hypothesized that
differences in the biota-depth relationships at the outfall
sites relative to those determined for the control sites would
imply the concurrent influence of other factors, including
deposits of CSO and SD particulates. Because depth was found
to be relatively constant within the two CSO study areas, only
the SD data were examined for depth effects.
It was also hypothesized that any influence of the discharges
would demonstrate a linear relationship with distance from an
outfall. These relationships were quantified using simple
linear regression at the CSOs and multiple regression (with
distance and depth as variables) at the SDs. Linear regression
techniques were also used to determine the biota-depth
correlations for the control sites.
These analyses were done both for the total infaunal community,
and for groups of organisms that were either especially
numerous in our samples or that are known to be important in
the lake's food chains. Included in these categories were
chironomid larvae, oligochaetes, nematodes, copepods and
pelecypods. A comprehensive summary of the taxa sampled
follows:
86
-------
layer were comparable to background levels. Together with the
relatively uniform sediment distribution shown by Figure 23,
this information implies that the sediments at C 3, one of the
"cleanest" areas in the lake, have been contaminanted in
recent years by particulates carried thence by longshore
advection. This interpretation of the data corresponds to
observations made previously relative to the nearshore turbidity
assessments. The closest "upstream" outfall is a combined
sewer structure located 730 m N of the center of C 3. The
dispersion hierarchy determined at C 3 by ranking carbon:metal
correlations was Cu> Pb> Zn (r=.951, .927 and .747, respectively);
this order was the same as that determined above for CSO 023.
There are no shoreline areas in Lake Washington that are more
isolated from-CSOs or SDs than is control site C 3. It follows
from this and from observations of their offshore movement that
it is probable that measurable levels of particulate pollutants
from CSO and SD discharges cover most of the lake bed.
Typical enrichment factors for total metals contributions
from all sources to the top 0.5 cm layer of the nearshore
bottom sediments may be derived from data in Table 15, by
dividing mean metals concentrations in this surface layer by
those for historical (pre-1900) sediments. The resultant
values (for Pb, Zn and Cu respectively) are: CSO 023 - llx,
3x, 2x; SD 7 - 7x, none, 3x; C 3 - 7x, 2x, none. Maximum/
near-outfall enrichment can be much greater. As shown by Spy-
ridakis and Barnes (1976), a large portion of the particulate
Pb in the lake is of aeolian origin.
Viruses — Although previous investigators (Smith et al., 1978)
have found viruses in sediments using methods identical to those
used for the present study, none were detected in surface
sediments from either CSO 023 or SD 7. Altogether, 36
separate samples were analyzed, including two sets of six
samples taken at the sediment trap locations (Figure 7) within
24 hr of storm overflows, and a third set of six each
collected one week after an overflow. The total volumes of
the relevant overflows ranged from 415 to 479 n\3 at CSO 023
and 125 to 2230 m3 at SD 7. The range of recovery efficiences
of the internal standards was 8-20%. One set of six samples
was also collected at the control site, and likewise yielded
no evidence of viruses.
Benthic Biota —
Evaluation of Biological Sampling Techniques — For purposes
of sampling the freshwater benthos, a statistical evaluation
of potential techniques was performed, with the number of
cores to be collected per placement (by divers) of the 0.5 m2
wire sampling grid, and the core tube diameter (i.e., sample
volume) as variables. The analysis was performed on three of
the principal taxonomic groups - chironomids, oligochaetes and
copepods. Table 16 shows the upper confidence limits expressed
84
-------
collections were generally 2-5x those of the 0-0.5 cm sediment
surface layer. This observation implies selective removal to
deeper areas, of the finer participates, with which the highest
metal concentrations are typically associated (Guy and
Chakrabarti, 1976). Sediment cores collected along across-
lake transects contiguous to CSO 023 and SD 7 and to a major
secondary treatment plant outfall abandoned in 1968 clearly
show an increase of Pb, Zn and Cu concentrations with lake
depth (Barnes, 1979). As is evident from a comparison of the
CSO 023 and SD 7 sediment trap data with values for profundal
(deep offshore) surface sediments, however, the latter materials
have been appreciably diluted by relatively uncontaminated
secondary sources, including diatoms and river detritus. Barnes
and Spyridakis (1976) have estimated that contemporary fluvial
inputs (including CSO and SD discharges) are responsible for
26%, 90% and 53%, respectively, of the Pb, Cu and Zn entering
the lake.
This offshore movement of discharge particulates is also
evident in the distributions of sediment metals near the two
outfalls. The isopleths for Pb and Zn around CSO 023 (Figure 24)
show the effects of near-bottom turbidity plumes moving downslope
to the southeast (refer to Figure 3 for areal bathymetry).
For Cu, the concentration isopleths imply the additional
influence of advective transport toward the northwest, where
the particulates settled due to flow disruption by condominium
support pilings. Although the prevalent direction of flow
of subsurface currents at the CSO 023 site was not clearly
defined by the light transmission studies, the northwesterly
advection implied by the sediment distributions is logical
as the most direct route to the flow outlet for Lake Washington
(Figure 2). The considerable differences in sediment distri-
butions for Pb and Zn vs. Cu indicate that the Cu-bearing parti-
culates were smaller and/or lighter, and thus more easily in-
fluenced by water motion. Carbon was found to be more widely
dispersed than the metals, in keeping with the nature of the
large, light carbon-bearing particulates. Correlation coef-
ficients calculated for Cu, Pb and Zn vs. C in the surface
layer of the bottom sediments were .646, .415 and .,238,
respectively, implying that the relatively light Cu and C
particulates have the most similar transport characteristics,
and further confirming the Cu > Pb > Zn dispersion hierarchy
suggested by Figure 24.
The selective transport tendencies of Cu, Pb and Zri at CSO
023 are also delineated by the concentration ratios given in
Table 15. The Zn:Cu ratio increases as the particulates
move from the outfall to the traps, and again, from the
traps to the sediments; this is because of the higher losses
of Cu through advective transport. The decrease of the
ZnrPb ratio represents an external input to this system -
82
-------
CENTER OF SEDIMENT TRAP ARRAY
LEAD (mg/kg)
8m W
23m E
53m E
84m E
•100-
-100'
•150
ZINC (mg/kg)
40
COPPER (mg/kg)
TOTAL CARBON (%)
Figure 26. Dry wsight distributions of lead, zinc, copper
and total carbon in the surface centimeter of
sediments collected at Control Site 3.
80
-------
COMBINED SEWER OUTFALL. 023
LEAD (mg/kg)
30m W
OUTFALL
LINE
30m E
84m E
ZINC (mg/kg)
COPPER (mg/kg)
TOTAL CARBON (%)
Figure 24. Dry weight distributions of lead, zinc, copper
and total carbon in the surface centimeter of
sediments collected near Combined Sewer Outfall 023.
78
-------
For the single month of mutual sampling (September), the sedi-
ment organic contents for CSO 023 and SD 7 were low (< 2%) and
approximately equivalent. With the exception of one site, the
median particle diameters (silt to very find sand), the sorting
coefficients and the skewness for the two areas were also
similar. The anomalous sample was from 21 m N, 30 m E of the
CSO 023 outfall, where the sediments were very poorly sorted
and the size spectrum was skewed more toward very small particles
than was that of any of the other 14 samples; the source of this
material was not apparent. Distributions of sediment surface
contaminants presented elsewhere in this report give expectations
for similar grain size characteristics at 21 m N, 30 m E and 21
m S, 30 m E, whereas the present data demonstrate significant
differences.
The discharge particulates from CSO 023 and SD 7 had extremely
disparate median diameters, being fine to medium sand for the
CSO and clay to fine silt for the SD; for both stations the
median diameters of the outfall debris were in the fine sand
range and larger than for the surrounding sediments. From this,
it would seem that the sediments near the opening of the SD 7
outfall were washed free of the finer particulates by the dis-
charge turbulence and were not predominantly of wastewater
origin. Sediment surface distributions of metals tend to con-
firm this supposition.
As denoted by Figure 23, two of the sediment sites at CSO 023
were sampled during both March and September, thus providing a
seasonal comparison of parameters. Sediments from both sites
were found to have a lower organic content, a slightly smaller
median particle diameter, comparable sorting and to have dis-
tributions favoring finer particulates in September than in
March. Generally speaking, these are the changes that one might
expect to occur during a dry period with few overflows, when
organics are dissipated and the natural breakdown of particulates
is not masked by fresh wastewater inputs. These data are too
few to be conclusive, but the trends are logical.
Considering the relative site-to-site magnitudes for each of the
three parameters representing particle-size distribution (M2,
S0 and Sk), one can generalize the pertinent findings of the
analyses as follows:
The particle-size distributions of sediments lying immediately
in front of discharges may be altered by the emission turbulence.
The most prominent effect is the shift in the skewness of the
distributions toward larger particles (as for CSO 023 and SD 7),
i.e., the finer particles are washed away. If the discharge
particulates themselves have a relatively high median particle
diameter (as for CSO 023), a significant portion may settle
close to the outfall, altering the median size accordingly.
Simultaneously, the settling of the larger particles and
76
-------
TABLE 13. SUMMARY OF VIRUS ANALYSES OF RECEIVING
WATERS NEAR COMBINED SEWER OUTFALL 023
AND STORM DRAIN 7 DURING AND AFTER STORM
DISCHARGE
Discharge Recovery
Volume Number of Efficiency
Station/Condition
CSO 023/Discharging
24 hr Post-Discharge
SD 7/Discharging
24 hr Post-Discharge
Date
2/6/79
2/7/79
2/10/79
3/6/79
(ra3)
155
155
335
1880
Virus es/m^
6600
0
0
0
(%)
3.7
3.1
3.2
2.4
virus analyses). The receiving water samples were collected
near the outfalls in patches of minimum light transmission in
order to optimize the chances of sampling the discharge plume.
As mentioned previously, however, plume separation does occur
due to particle size and density differences, so that the
measured virus concentrations may not represent maxima.
The 24 hr post-overflow sample was in each instance taken
from the same location as those collected during the overflow.
No viruses were detected at CSO 023, indicating that the
plume had dispersed, an observation confirmed by trans-
missometer readings.
Sediment Measurements —
Particle Size Distributions — Fifteen particulate samples were
analyzed for percent organic content and grain-size distribution
These samples included sediments from Nr NE, SE and S of the
CSO 023 outfall and NE and SE of the SD 7 outfall. Also, end-
of-pipe sediments and discharge particulates were analyzed for
both stations. Because large station-to-station sediment
variations were observed by divers even in areas relatively
free of outfalls, no single location was designated for control
measurements; rather, trends were sought within each area that
might relate to an outfall location and the characteristics of
the discharge particulates.
The resultant data were summarized in terms of four parameters
for each sample: organic content, median particle diameter,
sorting coefficient and skewness. The latter three variables
collectively represent the shape of a particle size distribution
curve (Sverdrup et al., 1942, p. 970). Figure 23 is a summary
of this information.
74
------- |