Unite tates
Environmental
Protection Agency
Office of Water
AWPD (WH-553)
Washington, DC 20460
EPA 841-F-92-O01
August 1992
oEPA
Water Quality Progress Reports
Water Quality Program Highlights
Horse Creek, South Carolina
Lower Fox River, Wisconsin
Flint River, Michigan
Housatonic River, Connecticut
Upper Potomac Estuary
Washington, DC, Virginia,
and Maryland
Upper Trinity River, Texas .
Tillamook Bay, Oregon
Passaic River, New Jersey .
Koshkonong Creek, Wisconsin
Whitewood Creek, South Dakota
Duwamish River, Washington
• January 1985
June 1985
• . August 1985
January 1986
• . March 1986
July 1986
August 1986
• October 1986
February 1987
June 1987
April 1991
Water Quality
Progress Reports
Water Quality
Program Highlights
New York State’s Wasteload
Allocation Procedure April 1986
The Deleware River Cooperative
Monitoring Program July 1986
Arkansas’ Ecoregion Program . September 1986
EPA’s National Dioxin Study . November 1986
The Massachusetts Fish Toxics
Monitoring Program January 1987
Maine’s Biologically Based Water
Quality Standards April 1988
Minnesota’s Nonpoint Source
Assessment Program May 1988
Ohio EPA’s Use of Biological
Survey Information May 1990
Multimedia Toxics Study of the
Calcasieu River Estuary,
Louisiana March 1991
Eutrophication Management in
North Carolina May 1991
Florida’s Method for Assessing
Metals Contamination in
Estuarine Sediments
June 1991
-------
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
January 1985
v>EPA Water Quality Progress Report
Horse
Creek,
South
Carolina
Horse Creek is a tributary of
the Savannah River on the
South Carolina side. By
1970 it had succumbed to
point source discharges
from inadequate sewage treatment plants and the textile
factories which have long been the region's economic mainstay.
The stream and its tributaries were murky, discolored, and deadly;
sections were devoid of fish and insects. Langley Pond, a
manmade impoundment of Horse Creek, was virtually lifeless.
Water quality professionals acknowledged Horse Creek as one of
the most polluted streams in the southeast
Today there are no significant point source discharges to Horse
Creek. The stream runs clear and has become a community
resource instead of a drawback. Although problems that had their
start almost a century ago have not been completely eradicated,
progress has been dramatic and satisfying.
This report discusses the efforts to rehabilitate a once "dead"
stream, the monitoring efforts that defined the extent of the
damage and the rehabilitation, and the role of monitoring in Horse
Creek's future.
ASSESSING THE PROBLEM
As a major tributary of the Savannah River, Horse Creek was part
of a 1970 study by the Environmental Protection Agency (EPA)
which documented water quality on the Savannah's middle
stretch. At that time, the major discharge sources on Horse Creek
and its tributaries were two municipal sewage treatment plants
and the mills and plants of two textile companies. Of the
combined discharges and their effects on Horse Creek and the
Savannah River, the study found that:
• About 55,000 pounds per day of BODs (five-day biochemical
oxygen demand) were being discharged into Horse Creek,
significantly depleting the creek's oxygen resources. This is
roughly the daily equivalent of raw, untreated sewage from a
city of 325.000 people.
• Sediments in the creek contained 990 milligrams per kilo-
gram (mg/kg) of chromium, a metal that is toxic to fish and
other aquatic organisms and at the time was a component of
textile dyes.
• Fecal coliform densities in Horse Creek and its tributaries
ranged from 500,000 to over 16 million colonies per 100
milliliters (ml).
• The creek was biologically "dead," with no living macro-
invertebrate organisms for at least 1.6 miles upstream from its
confluence with the Savannah River.
Following this study, EPA recommended that:
• The textile operations eliminate all chromium discharges to
the creek.
• South Carolina, in cooperation with EPA, establish discharge
limits for Horse Creek industrial and municipal waste sources.
• A feasibility study be conducted for a waste collection system
and centralized waste treatment facility for the entire Horse
Creek basin.
A 1972-73 Governor's Study conducted by the State of South
Carolina corroborated these findings and reached similar
conclusions.
INSTITUTING CONTROLS
In 1973 the Aiken County Board of Commissioners contracted with
an engineering firm to study alternatives for wastewater treatment
within the basin and to determine the most effective treatment for
combined industrial and municipal use. The firm recommended a
regional treatment facility with a 20 million-gallons-per-day (mgd)
capacity, expandable to 40 mgd, and an extended aeration
activated sludge process which their studies had shown was most
effective in treating textile wastes. Construction of the new facility
began in 1977 under the jurisdiction of the newly established Aiken
County Public Service Authority, and was completed late in 1979.
This plant now serves the basin's municipalities and industries, dis-
charging treated effluent directly to the Savannah River.
MONITORING FOR RESULTS
In 1979, the South Carolina Department of Health and Environ-
mental Control (SC-DHEC) expanded its trend monitoring pro-
gram for Horse Creek in order to thoroughly document physical,
chemical, and bacteriological water quality conditions before,
during and after start-up of the new wastewater treatment plant
Secondary stations, sampling monthly from May through October,
were upgraded to primary stations, sampling monthly all year
around. Heavy metals sampling was continued on a quarterly basis,
and sediments were sampled once a year. All monitoring data was
entered into STORET, EPA's computerized data base system for
the storage and retrieval of water quality data.
WATER QUALITY IMPROVEMENTS
The Horse Creek wastewater treatment plant virtually eliminated
discharges to Horse Creek. As expected, following its start-up in
1979, water quality in the creek showed dramatic improvement as
documented by 1982 monitoring data compared to the 1970 EPA
findings.
Oxygen. In July 1970, dissolved oxygen (DO) levels in Horse
Creek measured less than 2.0 mg/l; Langley Pond contained no
measurable amounts of DO. (South Carolina water quality
standards require an average of 5.0 mg/l of DO to maintain a
healthy aquatic habitat) The 1982 data show oxygen concentra-
tions from 6.7 mg/l to 8.2 mg/l at the sampling stations on Horse
Creek. By September 1982, the BODs concentration in the creek
had dropped to less than 2.0 mg/l from nearly 30 mg/l in 1970.
pH. Horse Creek is classified for fish and wildlife propagation;
South Carolina water quality standards require streams with this
classification to maintain a pH between 6.0 and 8.5. In 1970 pH
values ranged from 9.3 to 11.6 due to alkaline discharges from the
textile processing plants. Following construction of the Horse
Creek facility, the stream pH levels dropped to a range of 6.0 to 7.8,
the normal range for a healthy stream system.
-------
Chromium. Four of the textile plants contributed most of the
chromium discharged into the creek; one plant, which used
large amounts of chromium in its dyeing processes, discharg-
ed 865 pounds per day. During the 1970 EPA study the
chromium concentration in the water at one sampling station
was 440 pg/I and in the sedIment, 990 mg/kg. In 1976 EPA
recommended a stream criterion of 100 pg/i to protect
freshwater aquatic life. By 1982 chromium concentrations in
the creek were below detection limits (<50 pg/i).
Bacteria. During the 1970 study, the two sewage treatment
plants were contributing fecal coliform densities to Horse Creek and
its tributaries ranging from 500.000 to over 16 million colonies
per 100 ml. Salmonella bacteria, a genera that Is pathogenic to
both humans and animals, were Isolated in the poorly treated
domestic waste of the old plants and in the creek. In 1982,
following the removal of these discharges from the creek to the
new regional facility, fecal collform densities in the creek dropped
to 200 to 400 colonies per 100 ml, well within the state’s limit of
1,000/100 mL No Salmonella were found.
BIOLOGICAL MONITORING PROGRAM
The 1970 EPA study found Horse Creek biologically dead and it
remained so until after the regional facilityellmlnated point source
discharges. In 1979 South Carolina added biological monitoring
to its stepped.up monitoring program in order to track the
anticipated recovery of aquatic life when water quality improve-
ments due to the new treatment plant were realized. Monitoring
was done at the existing stations and at a new station on Langley
Pond.
Once each summer macroinvertebrates were collected from
HesterDendy multiplate samplers after a four-week exposure
period. Periphyton was collected with periphytometers after a two-
week exposure period. Phytoplankton was collected from the
Langley Pond station once during the spring and fall and twice
during the summer of each year. and fish were collected from
Langley Pond for metals and organic chemical analyses in 1981
and 1982.
Dissolved oxygen, temperature, pH and conductivity measure-
ments were taken each time biological samples were collected and
when sampling devices were deployed.
THE RECOVERY OF THE AQUATIC COMMUNITY
In June 1979, rio macroinvertebrates were collected from the
Hester- Dendy multiplates. But in 1980, following the start-up of
the treatment plant, macroinvertebrate communities firmly estab-
lished themselves. Residual effects of the prolonged discharge of
industrial and domestic wastes were still apparent, however. At
Station A, the insect order Dlptera, or true flies, were the only
macroinvertebrates represented. Of these, 92 percent were
chironomids (midges), a group considered very tolerant of poor
water quality and generally the first macroinvertebrate community
to colonize a recovering stream. Farther downstream, at Station
B, 46 percent of the community were chironomids, but four
additional orders of insects and a fairly high number of mayflies
and caddisfiies were found. These orders are generally intolerant
of poor water quality. In 1981, mayflies and caddisfiles were
represented at Station A as well as Station B. in 1982 the rate of
increase slowed as the community showed signs of stabilizing.
Figure 1 shows total macroinvertebrate specimens and species
for these two sampling stations collected in 1979 through 1982.
Shannon-Weaver diversity indices (d) and equitability values, also
shown in Figure 1, provide generalized indications of the health of
biological communities. The d is a measure of both the diversity
of species and the distribution of individuals among the species;
a range of 3 to 4 generally reflects waters unaffected by oxygen-
demanding waste. Although the creek has ad range of 3 to 4 the
measure may not adequately reflect water quality, since it is toxic
substances (e.g., chromium) that have been the culprit more than
oxygen demanding waste.
j I I I I
200 600 ‘ 1000 1400 1800
400 800 1200 1600 2000
FIgure 1 Macroinvertebrate Population 1980 - 1982
Equitability is that portion of d that involves the “evenness” of
distribution, and tends to be the more sensitive measure.
Equitability values of less than 0.5 (on a scale from 0.0 to 1.0)
indicate that specimens are not as evenly distributed among the
species as they might be, though a more equitable distribution
may occur as the system continues to recover.
Fish. With greatly improved DO concentrations and the virtual
elimination of discharges, there was good recruitment of fish into
Langley Pond from Its nonpolluted tributaries, canals, ditches, and
perhaps backwater areas of the pond.
Fish tissue analysis for metals and organics was done first in March
1981, and has been done annually since then. At first, only
nongame fish such as chubs and golden shiners could be found for
tissue analyses. By September 1981, more important game
species such as bluegills, crappie and largemouth b ’ss were being
collected, as well as a greater variety of nongame spi cies. Number
of specimens also increased.
Since sampling began, no fish have been found with contaminant
levels in tissue exceeding the Food and Drug Administration’s
(FDA) recommended safe tolerance limits. Although high levels
of polychiorinated biphenyls (PCBs) and metals (chromium,
copper, manganese, lead and zinc) were detected in the sedi-
ments in August 1979 and June 1981, accumulation has not
reached problem levels in fish, perhaps because the fish are
relative newcomers. Continued testing will help determine the
longer term effects, if any, on the fish population.
This fish tissue monitoring data will also be incorporated into a
statewide data base for future use in setting more inclusive limits
for safe and acceptable levels of metals and organics in fish tissue.
The sediment in Langley Pond is still a concern because it
contains significant levels of chromium as well as other metals.
South Carolina is currently conducting a study to determine the
extent of contamination of Langley Pond sediments and will use
the data gathered to determine the sediments’ potential impact
on the water chemistiy and biological communities in the pond.
Material for this report was furnished by William Cosgrove,
Environmental Services Division, U.S. EPA Region IV, and
Russ Sherer and Harry Gaymon, Division of Water Quality
Assessment and Enforcement, SC-DHEC.
report is produced by EPA to docuronnt progress achieved in
trtproving water quality. Contritutions of information for similar
reports are jnvjte’l. Please oxitact b.F. Drabkc ski, EPA, I4EED,
WH—553, 401 M Str -t S.W., Washington, D.C. 20460, (202)382—7056.
10 2° 30 40 50
5 15 25 35 45
I I i I I
1980 SpecIes
1982
1980 SpecImens
1981
: :: 1982
Station A
Station B [ J
Shannon-Weaver
Diversity Equitability
IC )
Station A
980 2.40 0.04
1981 3.38 0.48
982 3.22 0.35
Station B
980 3.20 0.54
1981 2.95 0.75
1982 3.39 0.42
-------
(Jnited States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
Water Quality Progress Report
r Lf L
Lower
Fox
River,
Wisconsin
The 40 mile stretch of the
Lower Fox River from Lake
Winnebago to Green Bay has
one of the heaviest concentra-
tions of point source dis-
chargers for its length in the country. Six municipal wastewater
treatment facilities and 15 pulp and paper mills are located along
this reach. Since the 1800s, the combination of nearby sources of
pulpwood and available hydropower has attracted a number of
paper mills. Their development has taken a heavy toll on the river,
in terms of degrading water quality, which has gone largely un-
checked until recently. Municipal loadings have also been signifi-
cant, accounting for about 12 percent of the total loadings. In 1973
the average daily five-day biochemical oxygen demand (BOD )
loadings from industrial and municipal point sources exceeded
218,000 pounds per day, the equivalent of untreated waste from a
population of 1.3 million.
Today the Lower Fox River is able to support game fish, and the
sludge beds and mats of paper wastes that were common sights
have disappeared. Controls implemented by the paper mills and the
publicly owned treatment works (POTW) have been responsible for
the clean-up, and the river is close to meeting the Wisconsin stan-
dard of 5 mg/I DO for recreation and aquatic life protection.
AMBIENT MONITORING
In conjunction with the (IS. Environmental Protection Agency’s
(EPA) timetable for municipal and industrial compliance with the
Clean Water Act, a computer model of the Lower Fox was
developed for Wisconsin DNR in 1969 to predict the impact of
future best practicable treatment (BPT) and secondary treatment on
the water quality of the river. However, little water quality data ex-
isted to verify the model. In 1972, an automatic monitoring system
was put into operation which consists of five monitoring stations
located at hydroelectric plants and paper mills. Still in operation,
the stations monitor DO, temperature, pH, and conductivity,
transmitting hourly readings to a computer. Besides providing DO
profiles of the river, the data is used to detect spills and nonpoint
source runoff events.
MONITORING PARAMETERS FOR MODEL INPUTS
Intensive water quality surveys were also instituted by WDNR in
1972 and one or more conducted virtually every year through 1980.
These synoptic surveys measured water quality in as many places
on the river over as short a period as possible to form an overall pic-
ture of the river at a specific time. Field measurements including
DO, temperature, pH, conductivity, secchi depth, and light extinc-
tion were taken at about one-mile intervals. Laboratory samples
were taken at five to ten locations, and analyzed for BOD (five-day
and long-term), nitrogen series, phosphorus, solids series, and
chlorophyll a.
Model input also included all effluent flows and quantities of various
pollutants for the day of each survey and the previous five days, and
atmospheric conditions, river flow, and other variables. Sediment
composition data was collected at several stations and extrapolated
to sediment oxygen demand levels. Phytoplankton was collected to
determine growth and respiration characteristics of various algae
groups.
After each survey was completed, a complete DO profile of the river
was drawn. These profiles were compared with the model output
and used to calibrate the model to various specific conditions; that
is, to determine the coefficients that enable the model to duplicate
actual conditions when it is given appropriate input data.
WASTELOAD ALLOCATIONS
Based on the river’s assimilative capacity to maintain a DO of 5
mg/I, modeling output indicated that this minimum standard would
not always be met under technology-based effluent limits.
Therefore, very early in the 1970s WDNR had labeled the Lower Fox
River a water quality-limited segment; that is, requiring more strin-
gent controls to meet the Wisconsin standard and other Clean
Water Act requirements (e.g., as specified under Sections 208,303,
and 305).
Wasteload allocations (WLAs) were therefore needed for all
municipal and industrial dischargers. Following the governor’s
designation of the area under Section 208 of the Clean Water Act as
one having substantial water quality problems, the Fox Valley Water
Quality Planning Agency (FVWQPA) was formed in 1976. WDNR
asked the FVWQPA to assume responsibility for developing WL.As,
with guidance from a task force of EPA, WDNR, municipal, and in-
dustrial officials. The dischargers also agreed to the use of the
model for developing wasteload allocations.
To generate wasteload allocations, QUAL Ill, a mathematical com-
puter model, was developed based on variables that had been
determined critical to DO levels in the Fox River. QUAL Ill is a site-
specific refinement of the generally available QUAL II model.
Parameters include river flow and temperature, headwater
biochemical oxygen demand, nutrient concentrations, algae conS
centrations, sunlight intensity, chlorophyll a, and the BOD loading
from each discharger. The QUAL Ill model is unique in incor-
porating so many parameters, and is thus capable of finely.tuning
the wasteload allocations for maximum protection of the river
without undue restrictions on dischargers. Since dischargers in
close proximity to
Green Bay
one another affected
the river collectively
Green Bay as if they were one
large discharger,
_ _ causing a down-
stream DO sag, the
dischargers were
grouped into
clusters. There are
three clusters on the
Lower Fox, each by
- - coincidence made
up of five pulp and
paper industrial dis-
Grouped chargers and two
municipal sewage
dischargers.
Q(JAL Ill was not applied to Cluster Ill dischargers located just below
Green Bay, because the model does not incorporate all variables
governing the dynamic river-bay estuarial system on that portion of
Rapide Croche Dam
CLUSTER II
Figure 1. Point Source Dischargers
Into Clusters
EPA
June 1985
LLUSI LX
• = Point Source
Dischargers
-------
the river. The wasteload allocations for this group of dischargers are
currently under development and are expected to be implemented
In 1986.
DETERMINING BASEUNE LOADS
To allocate BOD, discharge wasteloads, It was necessaiy to
calculate baseline loads for each discharger. These baseline loads
established each discharger’s equitable portion of the cluster’s total
maximum daily load (TMDL), a figure which is based on the river’s
assimilative capacity. Baseline loads for the pulp and paper mills
were calculated based on each mill’s highest production for seven
consecutive day s in 1973. The loads were expressed in pounds of
BOD, per day. For POTWs, daily maximum BUD, loads were
calculated based on 1976 and 1977 flow rates and on the assump-
tion that monthly average BOD, for industries is comparable to
monthly average loadings for municipalities with secondary treat-
ment. Approximate flow rates were selected in order to calculate
BOD, concentrations in parts per million (ppm).
Table I shows Cluster l’s baseline 80!), loadings In poundsper day
and the dlschargers’ loads as a percentage of the cluster total. To
develop a wasteload allocation that would be equitable to existing
disch rgers and still allow for residential, commercial, and in-
dustrial growth for 20 years In the Fox Valley, a reserve capacity
was established for municipal POTWs. No reserve capacity was
allocated to the pulp and paper mills.
Table I Cluster I Baseline Loads (including Reserve Capacity)
George Whiting Paper Co
365
14
Neenah-Menasha PO1W
6,336
23.8
Kimberly-CLark-Joint
2,328
88
Bergstrom Paper Co
8.788
33.0
Kimberly.CLark Lakeview
4,097
15.4
Wisconsin Tissue
3,198
12.0
Fourth PLant POTW
3,485
5.6
TOTAL.
26597
100.0%
With dischargers grouped into clusters and each cluster baseline
load determined, reductions of each cluster’s baseline load were
made and analyzed using the QUAL Ill model until a minimum DO
standard of 5 mg/I was maintained at all times and all places along
the river, It was determined that reductions were not necessary
throughout the year but only during times when a combination of
factors, notably summer low flows and high river temperatures,
reduced the river’s assimilative capacity below baseline levels.
Wasteload allocation matrices were developed for each cluster,
specifying allowable pounds of BOD, per day given varying flow
rates and temperatures and representing the assimilative capacity
of the river to maintain the minimum DO standard of 5 mg/I. These
data were further broken down into four summer and early fall
periods to further reflect the fluctuating assimilative capacity of the
river throughout the low flow season.
WATER QUAUTY IMPROVEMENT
By 1981, daily average BUD, loadings had dropped to less than
40,000 pounds per day total from more than 218,000 pounds per
day total in 1973. The average DO saturation level near Green Bay
rose from an average of 77 percent in 1976 before BPT controls, to
90 percent In 1977 where it has remained. (DO saturation levels are
temperature dependent; that is, a 100 percent saturation level in the
summer represents a DO level somewhere around 7 mg/I, while a
100 percent saturation level in the winter represents a DO level of
about 10-11 mg/I,) In 1973, DO concentrations at Raplde Croche,
historically the critical sag point in the river, never reached 5 mg/I
from mid-July to late October. in 1983, the DO concentrations at
this same point never dipped below 5 mg/I. Improvements in DO
levels ‘ were attributed mainly to the BPT controls implemented by
the paper mills.
Nutrient levels also improved, attributable mainly to the secondary
treatment installed by the POTWs. Monitoring data from one sta-
tion showed that average soluble phosphorus levels decreased 50
percent over the period 1977 through early 1983, from 0.05 mg/I to
0.025 mg/I; total phosphorous levels decreased 41 percent, from
0,22 mg/I to 0.13 mg/I; and organic nitrogen decreased 16 percent,
from 1.55 mg/I to 1.3 mg/I.
BIOMONITORING
In 1975, the Institute of Paper Chemistry conducted a biological
survey to compare benthic invertebrate communities upstream and
downstream of major discharges. The study was repeated in 1979 to,
assess changes as a result of installed controls. During each study,
two sampling stations were located above point source discharges
and nine stations were located below major discharges. An artificial
substrate sampler was used to collect benthic invertebrates at each
station after a six-week colonization period. There were three col-
onization periods from May through September during each sampl-
ing year.
During the 1975 survey, samples from most stations downstream
of the discharges were dominated by aquatic worms (NaididRe),
which are indicative of areas of degraded water quality with low DO
and high deposition of organic material. The highest worm den-
sities occurred at stations where there were dense bacterial slime
growths, suggesting that the worms used bacteria as a food source.
During this same period, the more sensitive caddisfly was virtually
absent from the middle and lower stretches of the river, although
caddlsfly communities were found at upstream stations. Cad-
disflies are associated with enriched but not degraded water where
there are no extended periods of low DO and where deposition of
organic materials Is low.
During the 1979 post-secondary survey, caddlsflies populated all
sections of the river and worms were virtually absent. Slime growths
were also nonexistent.
ASSESSING PROGRESS
In July 1983, the Toxic Substances Task Force on the Lower Fox
River System, made up of EPA, WDNR, and other representatives,
issued a report which identifies problems still to be addressed.
Among their findings:
• The clean-up of conventional pollutants has enabled the river to
support gamefish, thus presenting a possible human health risk
if toxic substances in the water and sediment bioaccumu-
late in the fish. —
• More than 100 chemicals have been identified in the Lower Fox
River system, including PCBs, chlorophenols, and ammonia.
• Thirty-seven toxic pollutants have been identified in the
discharges from pulp and paper mills and POTWs on the Lower
Fox.
The Lower Fox continues to be monitored by EPA, WDNR, and
various other groups, with emphasis on sediment and fish tissue
monitoring and analysis. Although toxic pollutants have been
found in the water and sediments, not enough is known yet about
their activity and fate to determine what, if any, course to pursue in
dealing with them. If present controls effectively limit the discharge
of toxic pollutants, and contaminated sediments are eventually
covered by clean sediments, the threat of their resuspension and
bioaccumulation in fish tissue can be minimized.
Material for this report was furnished by Jerome McKersle, Bureau of
Water Resources Managemen4 WDNR; Tun Doelger, Lake Michigan
Distnct HQ, WDNR; and Michael MacMullen, U.S. EPA Region V.
Biological survey data was taken from “Water Quality Improvements
in the Lower Fox Rwer, sansfr1, 1970-1980: M HLstor cai
Peispecthe,” by Bruce Markert, published fri the Proceedings of the
1981 Environmental Conference (AtLanta. Trade Association of the
Ai4 and Paper Industry).
This report u produced by EPA to document progress achieved In lmpmulng
water quality Contnbutbns of information for sunhlar reports ar e Invited. Flease
contact EF L)rabkowski. EPA. MDSD, WH.553. 401 M Street LW,
Washington, D C. 20460, (202)382-7056.
Name
BO
(bs/day)
% of Cluster
Total
-------
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
August 1985
Water Quality Progress Report
ww w
- - - -
- -
w w w w w W W
Flint
River,
Michigan
The Flint River basin drains
approximately 1350 square
miles and includes significant
agricultural as well as urban
areas. It is a major tributary of
the Saginaw River, which
empties into Lake Huron’s
Saginaw Bay. As part of the
Great Lakes watershed, it is
of national and international concern as well as local significance.
The entire Flint River system was designated for agricultural and
industrial water supply, navigation, partial body contact recreation,
warmwater fish, and public water supply in the late 1960s; yet stret-
ches of the river have been consistently unable to support these
uses until recently. This report documents monitoring surveys con
ducted to identify water quality problems and to demonstrate the
effectiveness of point source controls. These monitoring surveys
have also identified the continuing need for nonpoint source abate-
ment programs as well.
The river flows through the City of Flint before joining the
Shiawassee River about 50 miles downstream. The Flint
metropolitan area, which is the third largest in Michigan, is the ma-
jor source of domestic and industrial wastewater inputs to the river.
While a number of industrial dischargers have come and gone over
the years, the City of Flint publicly owned treatment works (POTW)
has been a constant source of nutrient loadings. The City of
Flushing POTW and Genesee County’s Ragnone POTW (7 and 19
miles downstream of the Flint POTW outfall, respectively) have
also contributed to pollutant loadings in the downstream reaches
of the river. Storm sewer discharges, combined sewer overflows,
county drains, and urban runoff add nonpoint source pollutant
loadings.
WATER QUALITY MONITORING
Biological and chemical monitoring surveys have been conducted
approximately every five years since 1969 by the Michigan Depart-
ment of Natural Resources (MDNR). The results have been used to
refine discharge permit limits and to identify needed im
provements in point source controls.
Fifteen to 18 stations were sampled during these intensive surveys
covering the 50-mile reach of the Flint River from about nine miles
above the City of Flint to approximately 40 miles downstream, just
above the confluence of the Flint River with the Shiawassee River.
The stations were consistent for all surveys and were sampled dur-
ing August of each survey year. Sampling techniques for chemical
parameters were standard and consistent for all surveys, though
laboratory methods improved with each survey. Qualitative
biomonitoring has been used to determine impacts on the Flint
River biological communities.
In 1969, degraded conditions were evident from the City of Flint
downstream, and the Flint P01W was found to be the major con-
tributor to a significant nutrient enrichment problem in the river.
As shown in Figure 1, MDNR survey results indicated that the in-
stream five-day biochemical oxygen demand (BOD 5 ) concentration
6.7
nearly doubled from 3.5 mg/I above the P01W to 6.7 mg/I below
the POTW outfall, ammonia-nitrogen increased eight times from
0.15 mg/I above to 1.4 mg/I below, and total phosphorus was four
times greater below the outfall at 2.1 mg/I than above at 0.49 mg/I.
For 10 miles downstream of the city, diurnal dissolved oxygen
(DO) fluctuations were extreme, often violating Michigan’s 4.0 mg/I
daily minimum standard. Macroinvertebrate community structure
indicated degraded stream quality.
_ j. 1 _
8OD NH 3 N Tot P
Figure 1. Water Quality Upstream and Downstream of the Flint P01W,
1969 (mg/I)
As a result, significant improvements were made to the Flint
P01W over the next five years, with funding from U.S. En-
vironmental Protection Agency (EPA) construction grants, as well
as State and local sources. Nonetheless, the 1974 survey showed
continued degradation downstream of the Flint POTW, as shown in
Table 1. Phosphorus control measures had significantly decreased
the Flint facility’s average effluent concentration of total and solu-
ble phosphorus by the 1974 survey, as shown in Table 2, but
loadings of BOD 5 and suspended solids had increased. This was
primarily the result of population growth and the many industrial
sources that were now directing their wastes to the treatment facili-
ty rather than to the stream. A similar change in effluent loadings
had occurred downstream at the Ragnone POTW, where total
phosphorus levels had dropped from 10.0 mg/I during the 1969
survey to 7.0 mg/I in 1974; and at the Flushing POTW, where total
phosphorus had dropped dramatically from 13 mg/I to 1.1 mg/I
over the same period. At Ragnone. BOD 5 and suspended solids,
however, increased significantly despite newly constructed secon-
dary treatment facilities.
Table 1. In-stream Survey Results 5 Miles Downstream from Flint P01W
1969-1983; August Survey Data (mg/I)
969 1974 1978 1983
DO
58
5.9
6.7
7.5
BOD 5
6.7
13.0
130
3.6
SS
27.0
38.0
33.7
N/A
Tot-P
2.1
1.8
0.46
0.25
SoI.P
1.5
L2
0.22
0.16
NH 3 .N
Sowces: 1969
1.4
-78 data from MDPIR.
1.8
1983 data co
2.8
urtesy of the City
0.29
of flint.
In 1978, monitoring data showed that BOD 5 remained a problem at
all three POIWs, and suspended solids concentrations were still
very high in the Flint and Ragnone POIWs’ effluents. Greater
operating efficiency at Flint and Ragnone was offset by higher
average annual flows at both plants. Total and soluble phosphorus
levels continued to improve at all three plants, however.
EPA
__El
BOD. NH N Tot P
-------
Table 2. Effluent Survey Results at Flint P TW; August Survey Data
(mg/I)
DO
BOD
ss
Tot-P
SoI-P
NH ,-N
Flint P0
TW Effluent
1969 1974
1978 1983
4.6
4.7
5.4
7.7
8]
17.0
23.7
‘
10.5
30.5
488
N/A
120
7.2
5.9
0.9
11.5
4.7
4.5
07
14.9
6.8
11.6
02
CBOD was measured but is not directly comparable to BOD .
Sowres: 1969-1978 data from MDNR. 1983 data courtesy of the Oty of flint.
BIOLOGICAL SURVEYS
Biological surveys over this same period reflected the findings for
the chemical parameters. In 1969, rooted aquatic plants and at-
tached filamentous algae essentially blanketed the stream bed,
adversely affecting habitat and DO levels for more than 30 stream
miles downstream of Flint. Macroinvertebrates that thrive in
organically enriched waters were abundant. Bottom-dwellIng
aquatic: organisms, sensitive to pollution, were very low in number
over this reach and never approached the species diversity found
upstream of the city, where considerable numbers of “clean water”
organisms were found.
Fish tissues, analyzed in 1973, indicated levels below the Food and
Drug Administration’s limits for polychiorinated biphenyls (PCBs),
dieldrin, DDD, DDT, and heavy metals.
Biomonitoring in 1974 indicated continued severe impacts on
stream organisms. Downstream from the Flint POTW almost to the
confluence of the Flint River with the Shiawassee, heavy nuisance
growths of aquatic plants persisted, and sewage odors and sludge
beds in shallow and backwater areas were still a problem.
Macroinvertebrate species diversity and numbers remained
depressed, with “clean water” organisms such as mayflies and cad.
disflies almost completely absent from the entire study reach. The
fish community was dominated by carp and minnows — pollution.
tolerant species — with only small numbers of panfish.
In 1978, aquatic plant growth was still at near maximum produc-
tion throughout the study reach, contributing to excessive diurnal
variations in pH and dissolved oxygen.
WATER QUALITY PROGRESS
By 1983, however, marked improvements in all the conventional
water quality parameters from the Flint POTW downstream were
evident. Advanced treatment capability was increased at the Flint
POTW that increased the discharge from 20 million gallons per day
(mgd) to 50 mgd resulting from expanded collection systems,
population growth, and industrial expansion. As shown in Table 1,
BOD 5 had decreased over 70 percent in-stream from 1978 to 1983.
Phosphorus levels and ammonia-nitrogen levels had decreased
significantly in the effluent as well, with corresponding decreases In
in-stream levels. Figure 2 shows in-stream phosphorus levels above
and below the City of Flint from 1973 through 1983.
Biological surveys verified that aquatic plant growths were decreas-
ing, especially within the river reach three miles downstream of the
Flint POTW. Dense plant growth was found only where channel
depths were less than one foot. Overall plant growth observed bet-
ween 1980 and 1983 appeared to be less extensive than in past
surveys; this was attributed to reduced point source discharges of
nutrients.
Benefits from the improved water quality have been realized in in-
creased recreational use of the river and an improved fish habitat.
The MDNR Fisheries Division began planting coho and chinook
salmon in the Flint metropolitan area in 1979, and sport fishing
has produced satisfactory returns. Walleye have moved into the
river from upstream impoundments as well as from Saginaw Bay in
Lake Huron. and established themselves in portions of the Flint
-J
C ’)
0
I
0.
C’)
0
I
0.
-J
0.25’
0.50-
WATER YEAR
Figure 2. Mean In-stream Phosphorus Levels Above and Below the
City of Flint
River. According to U.S. Fish and Wildlife Service survey data from
1980.81, white suckers, yellow perch, northern pike and crappie
are also present in considerable numbers during higher flow
periods. MDNR plans to conduct an intensive survey of the river’s
fish population within the next two years to determine production
levels.
REMAINING PROBLEMS IDENTIFIED
Although the Flint River has improved markedly since the 1960s,
monitoring has shown that urban and agi-icultural runoff are persis-
tent problems within the Flint metropolitan area and in the river’s
upstream reaches, respectively. A statewide strategy for the abate-
ment of rural nonpoint source pollution is in the planning stages,
based on a 1984 memorandum of understanding signed by
representatives of several federal and Michigan agencies. In addi-
tlon, Michigan strategies for dealing with urban and transportation-
related nonpoint sources will be developed.
Monitoring data has also shown chlorine and ammonia toxicity
problems, particularly downstream of the Ragnone POTW.
Seasonal disinfection and, potentially. dechlorination coupled with
improved treatment efficiency at this facility should alleviate tox-
icity concerns and improve DO levels downstream as well.
Ambient water quality monitoring data collected since the early
1970s at stations above and below the City of Flint has been
entered into EPA’s STORET system, a national computerized data
base for the storage and retrieval of water quality data. MDNR is
currently using this data to do a comprehensive long-term trend
analysis of conventional pollutants in the Flint River. Though no
organic pollutant problems are evident at this time. MDNR plans to
focus future monitoring and analysis efforts on organic pollutants
such as dioxins.
Future monitoring plans for the Flint River also include its par-
ticipation in an annual surveillance program for the entire Saginaw
basin that has been proposed by the International Joint Commis-
sion, an advisory body formed under the Boundary Waters Treaty
of 1909 to protect (IS. and Canadian boundary waters. The pro-
gram recommends both chemical and biological monitoring.
Material for this report was furnished by Manj Ellen Falion, Waler
Quality Swveillance Section, MDNR, and Michael MacMullen, U.S.
EPA Region V.
This report is produced by EPA to document progress achieved in improving
water quality. Contributions of information for similar reports are InvItecL
Please contact E. F. Drabkowski. EPA, MDSD, WH-553, 401 M StreetS. W.,
Washington, D.C. 20460, (202) 382- 7056.
125
1.00
0.75
1973 74 75 76 77 78 79 80 81 82
-------
United States
Environmental
Protection Agency
January 1986
EPA
Water Quality Progress Report
wW W
H ousatonic
River,
Connecticut
The Housatonic River drains
approximately 2000 square
miles in western Connec-
ticut, southwestern Massa-
chusetts, and eastern New
York State. As an important
regional resource, the river
has long supplied hydroelectric power from a series of dams and
provided extensive recreational opportunities along its length. Prin-
cipal recreational opportunities occur in a series of impound-
ments—Lake Lillinonah, Lake Zoar, and Lake Housatonic—on the
lower Housatonic River in Connecticut. Although designated as
Class B water bodies (fishable, wimmable). historically these
impoundments have suffered from severe eutrophication. Recre-
ation activities have often been impossible during the summer
when scum or dense mats of blue-green algae covered the water.
Today, water quality in the three Housatonic Lakes is significantly
improved as a result of Connecticut’s basin-wide strategy to reduce
nutrient loading to the river. Controls implemented at industries
and publicly owned treatment works (POTWs) in Connecticut and
Massachusetts have reduced phosphorus discharges, resulting in
a corresponding decrease in nuisance algal growth.
ASSESSING THE PROBLEM
Starting with the US. Environmental Protection Agency’s (EPA)
National Eutrophication Survey in 1972, a series of studies docu-
mented the highly eutrophic condition of the Housatonic Lakes
and also identified phosphorus as the growth-limiting nutrient for
algae.
To better understand the transport and fate of pollutants in the
Housatonic, the Connecticut Agricultural Experiment Station
developed a hydraulic model (RVRFLO) and calibrated it for that
portion of the river that included Lake Lillinonah. The phospho-
rus budget developed by the model showed that point sources were
an important source of loading in the watershed and that Lake Lilli-
nonah was the major source of phosphorus to the downstream
impoundments. Based on the model results, the State conducted
a study to assess the effects of phosphorus removal at one of the
major point sources (the Danbury P01W) which is described below
in addition, a decision was made to focus on reducing phospho-
rus loads to the most upstream lake.
The RVRFLO model also supplied important information concern-
ing the relationship between river flow and hydraulic residence time
in the impoundments. During the spring high-flow period when
nonpoint source input of phosphorus is relatively high, water moves
through the impoundments in less than a week. During the July
and August low-flow period, when point source phosphorus loads
dominate, water takes 3 to 12 weeks to move through the impound-
ments. Since algal blooms had been observed to occur when the
incubation period in quiescent water exceeded several weeks, it was
particularly important to control point source input during the sum-
mer recreation period, when conditions were most likely to produce
algal blooms.
INTERSTATE TRANSPORT OF POLLUTANTS
Previous water quality monitoring studies as well as the phospho-
rus budget indicated that a significant proportion of nutrients was
transported into Lake Lillinonah from outside the State. To deal
more effectively with this and other interstate pollution problems,
EPA’s Region I office formed the Working Group on the Interstate
Transport of Pollutants. Quarterly meetings of the Working Group
and the involvement of officials from Connecticut, Massachusetts,
and New York (as well as representatives from business and munic-
ipalit es) are resulting in a successful cooperat 1 e effort to control
point source discharges.
INSTITUTING CONTROLS
In its Housatonic Basin Plan, the Connecticut DEP recommended
a strategy of phosphorus controls for all significant point sources.
Table 1 lists the important point source dischargers (of phospho-
rus) that affect the Housatonic Lakes, along with the approximate
effluent loads before and after controls. In each case, the NPDES
permits were modified to require phosphorus removal; for the Pitts-
field plant, the revised permit required Connecticut to monitor
water quality in the Housatonic River and Lakes for four years fol-
lowing the beginning of controls.
TABLE 1. Phosphorus Dischargers in the Lake Lillinonah Ba.
sin Before and After the Implementation of Controls (lL iay)
Year
Discharge
Discharge
controls
before
after
were
controls
controls
Source
instituted
(total P)
(total P)
Danbury POTW
1977*
250
50
New Milford POTW
1982
12
1
Bethel POTW
1982
15
2
Kimberly-Clark, Inc.
1980
10
1.5
Nestle’s, Inc.
1983
30
5
Pittsfield POTW (MA)
1982*
283
95
TOTAL
600
155
The Danbury plant did not control for phosphorus in 1978,
and Pittsfield did not control in 1983 or 1984.
Although nonpoint sources are less significant than point sources
as a cause of nuisance algal blooms in the river system, they do
represent an important source of nutrients. As a result. Connec-
ticut has identified the Housatonic basin as a priority area for imple-
mentation of agricultural best management practices.
WATER QUALITY MONITORING: 1976-1977 -
To assess the water quality effects of point source phosphorus con-
trols, Connecticut DEP, in cooperation with FMC Corporation, con
ducted several intensive monitoring studies. The first study was
conducted during the summers of 1976 and 1977—before and dur-
ing the first year of seasonal phosphorus removal at the Danbury
P01W. Samples were collected biweekly from April through
October for both years at five monitoring stations below the Dan-
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
,u I eId
-------
bury plant, four monitoring stations on the Housatonic River, and
four monitoring stations on Lake Lillinonah. The samples were ana-
lyzed for approximately 20 water quality parameters, including crit-
ical eutrophication measures such as phosphorus and nitrogen
species, transparency, chlorophyll a, and phytoplankton compo-
sition.
Results from the 1976-77 Danbury study were not conclusive due
to operational problems at the Danbury treatment facility and
incomplete flushing of the impoundment; however, definite
improvements in water quality were observed. This was true par-
ticularly in the comparison of August and September data for the
two years. During this period in 1977, the lake was undergoing its
second flushing since the initiation of phosphorus treatment, and
cumulative system downtime improved from 33 to 16 percent.
August and September were also the months when algal concen-
trations had reached peak levels in previous years. Analysis of the
limnological data for these August and September months showed
the following results:
• Mean chlorophyll a concentration was ieduced from 35.2 /Lg/L
in 1976 to 25.4 jig/L in 1977
• Mean soluble reactive phosphorus concentrations were reduced
from 27.9 g/L in 1976 to 18.6 rg/L in 1977
• Mean secchi disc transparency increased from 1.2 meters in
1976 to 2.6 meters in 1977.
WATER QUALITY MONITORING: 1981-82
A second intensive monitoring survey was undertaken as a cooper-
ative effort between the EPAs Region I Office and the States of Con-
necticut and Massachusetts. This study, designed as a two-phase
program, examined both the transport and the effects of phospho-
rus that was discharged (in the summer) from the Pittsfield, Massa-
chusetts, POTW. This plant is located approximately 40 miles north
of the State line and 100 miles north of Lake Lillinonah.
In 1981, the Pittsfield plant did not remove phosphorus, and in 1982
the plant removed phosphorus during the summer recreation sea-
son. The intensive survey measured flow rate and concentrations
of total and dissolved phosphorus twice a month during the sum-
mer at four monitoring stations along the Housatonic River (from
Pittsfield to the headwaters of Lake Lillinonah) and at three stations
on Lake Lillinonah. The Lake stations were sampled for phospho-
rus, transparency, and chlorophyll a. Data from the two years were
then compared to evaluate the effects of this change in treatment.
The 1981 survey revealed that a significant fraction of the Pittsfield
phosphorus was transported by the river to Lake Lillinonah. In addi-
tion, sediment samples in the riverbed above the impoundments
contained relatively little phosphorus, indicating that nutrients
attenuated by the river during low flows were transported to the
lakes during high flow periods.
During the second year (1982), attempts to measure water quality
improvements attributable to controls at the Pittsfield treatment
plant were complicated by several factors. These included heavy
rains and a 100-year flood event in June (which dramatically
increased nonpoint loading over that of 1981); a two-week suspen-
sion of phosphorus removal at the Danbury plant due to hydraul-
ic problems caused by the flood; and the initiation of phosphorus
removal at the New Milford and Bethel POTWs. In spite of these
conditions, it was possible to attribute changes in phosphorus levels
in the river, above the three Connecticut POTWs and major non-
point sources, directly to the reduction in loading at Pittsfield.
Removal of approximately 190 ll day at Pittsfield resulted in a
reduction of 80 to 100 lt*Iay at the State line, and a reduction of
50 to 70 lbUay at the headwaters of Lake Lillinonah.
Water quality at three fixed stations in Lake Lillinonah was moni-
tored in 1976, 1981, 1982, and 1984. These data suggest trends
as well as the effects of point source phosphorus controls. Sum-
mary statistics for three parameters during July and August are
presented in Table 2. In the Table, the effects of phosphorus con-
trols initiated at the Danbury plant in 1977 are evident when 1976
data are compared to 1981 or 1982 data. Further improvements,
as a result of phosphorus controls at Pittsfield in 1982, are sug-
gested by comparing 1981 data with 1982 data. In 1984, Pittsfield
again did not remove phosphorus, and the data indicate a decline
in water quality.
TABLE 2. Lake Lillinonah Survey Results:
Mean Values for July and August
1976
1981
1982
1984*
Total P (mg/L)
0.104
0.063
0.050
0.078
Secchi depth (m)
1.39
1.59
1.83
1.42
Chlor. a (jLg/L)
91.416
16.207
21.167
13.93
*Dath are from August only.
BIOMONITORING STUDIES
In 1975, phytoplankton studies conducted by the EPA found the
dominant organisms to be the nitrogen-fixing blue-green algae
Anabaena sp. and Aphanizomenon sp. Both species are capable
of utilizing atmospheric nitrogen, and are typically found as mats
or scum floating on the surface. Additional biosurveys for plank-
ton were carried out as part of both the Danbury and Pittsfield phos-
phorus removal studies. In the 1976-77 Danbury study, the U.S.
Geological Survey monitored for algal growth potential (AGP) dur-
ing the study period. Stations below the Danbury plant showed a
significant reduction in AGP. AGP data collected at a station down-
stream from the Danbury plant showed 121 mg/L (dry weight of
algae) in 1976 compared with an average of 57 m IL in 1977.
A qualitative study of species composition during the Pittsfield sur-
vey in 1982 indicated a shift in species composition. What had been
dominated by Anabaena/Aphanizomenon, with Spirogyra and
Hydrodictyon as minor dominates, appears to have shifted into a
community co-dominated by Anacgstis cyanae and Lyngbya bir-
gei. While all are considered to cause nuisance conditions, the
former species are more often found in highly eutrophied water
bodies, and, unlike the latter species, form surface mats and scums.
FUTURE MONITORING
Results of the 1981/1982 monitoring study were inconclusive con-
cerning the trophic response of Lake Lillinonah. This was largely
the result of variability in river flow and in nonpoint source load-
ings due to flooding, and changes in point source loadings in addi-
tion to the change at Pittsfield. To resolve this question, Pittsfield’s
revised NPDES permit requires the Connecticut DEP to monitor
water quality in the Housatonic River, including Lake Lillinonah,
during the summers of 1985 through 1988. Again, the objective
will be to compare conditions without phosphorus removal (1981
and 1984) to conditions with phosphorus removal (1982 and
1985-1988). During these future surveys, Connecticut will also mon-
itor river flow records, precipitation records, and the operational
records and effluent monitoring data for other phosphorus-reg-
ulated dischargers. The intention is to account for all the major fac-
tors that influence trophic conditions in Lake Lillinonah.
Efforts to assess water quality in this basin will continue, and addi-
tional controls developed as needed.
Material for this report was furnished by Charles Fredetle, Principal
Sanitary Engineer for the Connecticut Department of Environmen-
tal Protection, and Eric Hall, U.S. EPA Region I.
This report is produced by EPA to document progress achieved in
improving water quality. Contributions of information for simila-r reports
are invited. Please contact E. F Drabkowski, EPA, MDSD, WH-553,
401 M Street S.W, Washington, D.C. 20460(202) 382.7056.
-------
March 1986
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
Water Quality Progress Report
Upper
Ik tomac
Estuary
Washington, DC
Virginia,
and Maryland
The Potomac River estuary forms
the border between the States of
Virginia and Maryland, with the
tidal freshwater portion of the estuary flowing for nearly 35 miles
(south from the Chain Bridge) through the Washington, DC,
metropolitan area. The Potomac serves as both an aesthetic and
recreational resource to a large number of Washington, DC, area
residents—over 3 million people. Sport fishing, boating, and to a
limited extent, swimming are all available. In 1981 approximately
25 marinas operated on the 40-mile stretch of the Potomac
between the District of Columbia and Maryland Point.
During much of the past century, the Potomac estuary has been
plagued by problems of high coliform counts, low dissolved oxy-
gen, and nuisance algal growth, The primary source of these
problems was urban development: increased loading of organics,
nutrients and sediment from wastewater treatment and stormwater
runoff. Unlike other major estuaries along the Atlantic coast, there
are very few direct industrial discharges in the Potomac basin.
Early concern by Federal and State officials led to the first Potomac
Enforcement Conference in 1957. A second conference, held in
1969, resulted in a formal agreement to control organic and
nutrient discharges from publicly owned treatment works(POTWs).
Since then, over I billion dollars has been invested, and water qual-
ity in the tidewater Potomac has improved significantly. During this
period, Federal, State, and local agencies supported various water
quality monitoring efforts along the river. The result has been both
the confirmation of water quality improvements, and an increased
appreciation for the complexities of estuarine dynamics. This
report documents the role of monitoring in the ongoing Potomac
river cleanup.
IMPLEMENTING CONTROLS
U.S. Public Health Service officials, writing on conditions in the
Potomac estuary in the late 1950’s, described the river as
“malodorous . .. with gas bubbles from sewage sludge over wide
expanses of the river.. . and coliform content estimated as equiva-
lent to dilution of I part raw sewage to as little as 10 parts clean
water . Samples collected near the District showed dissolved oxy-
gen (DO) levels with values of 0.1 to 0.6 mg/L and, during the low-
flow summer months, massive algal blooms and fish kills were
commonplace. In 1970, the District of Columbia, Virginia, and
Maryland signed a Memorandum of Understanding that commit-
ted each government to sharp reductions in biological oxygen
demand (BOD) and nutrient loadings. These limits were incorpo-
rated into the first NPDES permits (issued in 1974), and, with the
exception of nitrogen removal (dropped from most permits pend-
ing further study of the effects of phosphorus removal), the treat-
ment levels called for in 1970 are still in effect.
By 1981, construction of nearly all advanced treatment units had
been completed, and 10 smaller treatment plants in Virginia were
deactivated, with their combined flows (about 18 mgd) redistributed
to other upgraded facilities. The primary pollutant reduction (in
terms of mass loading) was achieved by the Blue Plains wastewater
treatment plant in Washington, DC. This facility is the largest sin-
gle point-source discharge to the river, accounting for up to 80 per.
cent of all municipal point-source flow. Blue Plains has an annual
average capacity of 309 mgd (with peak flows of 650 mgd) and
stringent discharge limits for BOD 5 (5.0 mg/L), total suspended
solids (7 mg/L), and phosphorus (0.22 mg/L). Table I shows the
reduction in total pollutant loading as a result of improved treat-
ment measures undertaken by the nine POTWs that discharge to
the upper Potomac estuary.
TABLE 1. Total Wastewater Flow and Loadings
to the Tidewater Potomac
Year
Average
flow
(mgd)
BOD 5
(lb’day)
Total
phosphorus
lb day)
Total
nitrogen
(lL ’day)
Suspended
solids
(lb’day)
1970
325
136,000
23,500
57,700
140,000
1977
361
76.500
7,800
60,400
80.000
1981
437
36,800
4,100
55,000
1983
455
18,700
2,200
59,700
17,600
1984
463
16,818
841
57,213
13,133
Source: Metropolitan Washington Council of Governments
WATER QUALITY MONITORING
Along the Potomac, there is a long history of water quality monitor.
ing that includes efforts by the EPA, the U.S. Geological Survey
(USGS), and various State and local agencies. For a long time,
however, monitoring systems were not coordinated and, as a result,
reported data were often not comparable. In 1977 the USGS began
a 5-year monitorinq study of water quality in the tidal Potomac.
Then, in 1982, with support from State and local agencies in
Maryland, Virginia, and the District of Columbia, the Metropolitan
Washington Council of Governments (MWCOG) organized the
Coordinated Potomac Regional Monitoring Program. Under this
program, water quality samples are routinely collected at 15 sta-
tions in the free-flowing portion of the river and at4l stations in the
tidewater. Stations are located at or near former EPA and USGS
sampling locations to maintain a comparable data base; and all sta-
tions are sampled on the same day and at the same low slack tide.
Major nutrients (nitrogen and phosphorus series), chlorophyll a,
suspended solids, and BOD parameters are consistently measured
by all participating agencies, with split-sample analyses used to
ensure consistency among laboratories.
As originally designed, the Coordinated Monitoring Program was
to consist of regular monthly sampling. However, beginning in
August 1983, in response to reports of heavy algae concentrations
in portions of the river, sampling frequency was increased from
monthly to weekly and the number of stations reduced in order to
concentrate efforts in the area of heaviest algal growth. These
monitoring efforts during the algal bloor. .s of 1983 and 1984 were
invaluable in helping to understand processes at work in the river.
EPA
Table 2 presents data on instream water quality during the period
-------
before and after treatment plant improvements. These data were
gathered at Wilson Bridge (mile 12 below Chain Bridge), site of the
traditional DO sag below the Blue Plains facility. The effects of algal
blooms, which occurred during 1983 and 1984 are not reflected in
this table since these blooms occurred downstream below mile 20.
TABLE 2. Average Summer Concentrations, Mile 12
(Wilson Bridge)
DaytIme
Dissolved
dIssolved
Total
inorganic
oxygen
C
hloro. a*
phosphorus
nitrogen
Year
(mg/L)
(jig/L)
(mg/L)
(mg/L)
1970
4.2
75
0.62
1.75
1981
7.2
50
0.15
2.25
1984
7.5
13
0.10
1.35
Source: MWCOG
ChIorophyll a concentrations, more than other parameters, may be
influenced by flow rate.
MODELING WATER QUALITY
By 1980, the control program had significantly improved water
quality in the Potomac even though some major investments were
still needed to meet the original effluent and water quality goals.
Local officials, as well as State and Federal agencies, questioned
whether completion of the original program was necessary, and
there were suggestions that less costly nonpoint-source controls
might be traded off against municipal treatment efforts. To address
these concerns, the EPA and MWCOG developed the Potomac
Eutrophication Model (PEM) to estimate the effect of point- and
nonpoint-source inputs on algal growth and dissolved oxygen. In
December 1982, the model was formally approved for planning
purposes and it has now become one of the primary tools used
to guide nutrient control decisions in e Potomac.
The development of the PEM was enhanced by the relatively long
and complete record of monitoring data. The model was calibrated
over a 4.year period in the late 1960’s and verified for 3 year per.
lod in the late 1970’s. The PEM was then applied to evaluate the
effectiveness of both planned effluent limitations and other con-
trol alternatives. Important conclusions of this modeling effort
were that (1) the instream goal of 25 g/L chlorophyll a could not
be consistently achieved in the estuary, even with full imple-
mentation of the control program, and (2) point-source phospho.
rus removal would be effective only in the upper 30.35 miles of
the estuary, with sediment phosphorus release more important
below this point. This was useful information in the effort to reex-
amine the control program, however, its application was thrown
into question with the appearance in 1983 of a major, unexpected
algal bloom on the Potomac River.
ANALYZING THE ALGAE BLOOM
With the exception .of 1977, no significant algae blooms had
occurred in the Potomac since 1971, and a primary nuisance
species, the blue-green algae Microcystis aen.iginosa, had seeming-
ly disappeared. Then, in the summer and fall of 1983, a major out-
break of M. aeruginosa took place in a 20.mite stretch of the
Potomac estuary. This bloom was similar to those that occurred
during the late 1960’s, with chlorophyll a concentrations as high
as 250 g/L to 300 g/L. However, in 1983, phosphorus and BOD
loads from municipal plant sources were at their lowest point ever.
M. aeruginosa blooms also occurred during the summers of 1984
and 1985, although these were nowhere near as widespread or per-
sistent as the 1983 bloom. The question was, what went wrong?
The EPA, with support from the States, convened an Expert Panel’
to investigate the cause of the bloom. Using hydrological and
nutrient data from the Regional Monitoring Program, the panel first
tried to “predict” the 1983 bloom using the PEM. While the model
did capture the early phase of the bloom (up to 100 zg/L chlor. a),
it was unable to reproduce the highest observed concentrations.
The model also underpredicted the observed total phosphorus con-
centration by a factor of 2 to 5. The task of the Expert Panel was
to uncover the source of this nutrient.
Monitoring data as well as model calculations suggested that the
source v as iiot the result of a noripoirit slug or a treatment plant
failure, but rather a s stained.event originating in the sediment.
Numerous hypotheses were advanced; however, the one that
seemed to best explain the observed data relied on a new theory
of “p11-mediated” sediment release. The Expert Panel hypothesized
that algal uptake of carbon dioxide raised the pH to a point that
caused phosphorus to be desorbed from the sediments and enter
the water column as an additional supply for algal growth. During
the months when chlorophyll concentrations were greatest, sam-
pling in the bloom area recorded large increases in pH (to nearly
10) and total phosphorus concentrations between 03 and 0.4 mg/L
Figure 1 shows a plot of chlorophyll a, p11, and total phosphorus
levels recorded at the center of the bloom area in 1984. As a result
of the Expert Panel’s analysis, the PEM was modified to include pH-
alkalinity chemistry and the possibility of controlling future blooms
by enhancing alkalinity in the water is now under study.
CHL-A H
( gIL)
200
150
100
50
0_
Source: MWCOG
Figure 1. Chlorophyll a, Total P. and pH Recorded
at Smith Point, 1984
Phos
(mg/L)
BIOLOGICAL MONITORING
Along with the algae bloom, 1983 may be remembered as the year
in which submerged aquatic vegetation (SAV) returned to the
Potomac after an absence of several decades. This reappearance
is viewed positively since many of these plants are only found in
clean water environments. SAVs also play an important role in
providing food, cover, and a natural habitat for a variety of aquatic
life. Additional surveys, conducted by the USGS in 1984, indicated
that SAVs increased in abundance and distribution over levels
recorded in 1983; in many cases, three to six different species were
present in close proximity. However, Hydrilla verticeliata has also
been noted, and the rapid, dense growth of this plant can disrupt
navigation and recreation.
During the summer of 1984, the District of Columbia’s Environ-
mental Control Division sponsored a fish survey along the tidewater
Potomac. It focused on determining characteristic species, diver-
sity, abundance, and species’ h4bitats at various locations. Since
there are no previous surveys with which to compare the results,
the 1984 sample is of limited value for assessing trends. Neverthe-
less, the diversity (36 species) and abundance of fish indicate water
quality that is adequate to support a healthy fishery. Future sur-
veys should provide an excellent way to track water quality trends.
In the meantime, at least one indicator of renewed fisheries poten-
tial in the Potomac is the return of professional fishing guides; at
the last count, six were available in the DC Metropolitan area.
Material for this report was furnished by Stuart Freud berg, Metropoli-
tan Washington Council of Governments; Paul Eastman, Director of
the Interstate Commission on the Potomac River Basin; Jim CoWer,
DC Department of Environmental Services; and Chafles App, U.S.
EPA Region Ill.
This report is produced by EPA to document progress achieved in
improving water quality. Contributions of information forsimilar reports
are invited. Please contact E. F Drabkowski, EPA, MDSD, WH-553,
401 M Street SW, Washington, D.C. 20460(202)382-7056.
-------
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
EPA Water Quality Pmgress Report
Upper Trinity
River, Texas
The upper Trinity River basin
drains approximately 8,000
square miles in the area
around Dallas and Fort Worth,
Texas. Before the early 1900’s,
when the first water supply
reservoirs were constructed,
the river ran through the prairie and was dry for most of the year—
resembling a true river only after a major rainfall. Today, in nearly
all respects the Trinity is an “urban” river. Flow in the river is
dependent n the discharge of wastewater effluent, on stormwater
runoff from extensive impervious surfaces (overall, about one third
of the basin is urban land), and on releases from manmade reser-
voirs in the upper watershed.
Because of its limited assimilative capacity, the Trinity has long
been subject to severe water quality problems. The State Health
Department, writing in 1925, described the Dallas-Fort Worth sec-
tion of the river as having an “inky surface putrescent with oxidiz-
ing processes to which the shadows of over.arching trees add
Stygian blackness and the suggestion of some mythological river
of c’ th.” By the early 1970’s, primary and some secondary sew-
ag reatment systems had been installed; but the river was sub-
ject to numerous small point sources, and many publicly owned
treatment works (POTWs) suffered operational problems due to the
phenomenal growth that was occurring in the Dallas Fort Worth
region.
Today, many of the smaller treatment plants have been phased out
and replaced by a network of larger “joint system” plants with
enhanced secondary treatment. As a result, water quality in the river
has improved significantly. and there is now a growing interest in
utilizing the flood plain along the Trinity. The river flows through
the heart of the Dallas-Fort Worth metropolitan area, and both cities
as well as other local governments have recently proposed recrea-
tional and open space plans to make extensive use of the Trinity
River corridor.
Along with the improved treatment, federal, State, and local agen-
cies have cooperated in the development of an innovative monitor-
ing program that uses continuous automated monitors to record
water quality values along the river. This monitoring system has
both confirmed the positive effects of investments in improved
treatment systems and provided a powerful diagnostic tool for
investigating remaining water quality problems in the region.
IMPLEMENTING CONTROLS
In the 1960’s, local governments around Dallas and Fort Worth
began a coordinated process to plan for the future wastewater treat-
ment needs of the region. Working through the North Central Texas
Council of Governments (NCTCOG), a designated regional water
quality planning agency, the Upper Trinity River Basin Comprehen-
sive Sewerage Plan (UTRBCSP) was developed; this regional plan
was one of the first in the nation and it was the first to be approved
by the US. Environmental Protection Agency (EPA) in 1971.
The Treatment Works Improvement Program that emerged from
the plan detailed a system of regional wastewater treatment plants
that would eliminate many of the smaller single-community plants.
The premise was that larger “joint system” plants, operated by high.
ly trained personnel, would produce a better quality effluent at fewer
locations. Additionally, smaller cities could realize economies of
scale by eliminating daily operation and maintenance costs in favor
of an annual rate for sewage treatment at a regional plant. The orig-
inal plan delineated plant locations, regional system operators, a
plan for financing, and managerial responsibilities.
Between 1970 and 1985,21 POTWs and 13 private treatment plants
were phased out, and the joint systems’ share of the total flow
increased from 79to 95 percent. During the same period, the total
permitted flow increased from 283 to 485 million gallons per day.
As a result of the consolidation and treatment system improve-
ments, oxygen-demanding wastes have been reduced dramatically.
For the seven joint-system plants that discharge directly to the river
or its tributaries, annual flow-weighted loadings of biochemical oxy.
gen demand (BOD) decreased from 44 mg/L in 1977 to 14 mg/L
in 1985. The POTWs, as a group, continue to produce an average
effluent that is cleaner than the federal standards for secondary
treatment (30 mg/L BOD), and many are now achieving an average
that is below the Texas secondary standard of 20 mg/L
In 1985, the Texas Water Commission completed wasteload allo-
cations affecting two important segments in the upper Trinity River
basin. Asa result, most of the major POIWs will implement addi-
tional treatment to meet more stringent effluent limitations. In par-
ticular, seasonal limits on ammonia-nitrogen are required for major
dischargers—with lower limits during the summer months when
oxygen-demanding compounds have the greatest impact.
WATER QUALITY MONITORING
Monitoring activity in the Trinity River has increased significantly
through the installation of a Continuous Automated Monitoring
System (CAMS) that was designed as part of the areawide planning
studies conducted by NCTCOG in 1974 under Section 208 of the
Clean Water Act. The system was installed with the cooperation of
the U.S. Geological Survey (USGS) through an interagency agree-
ment with NCTCOG. Following installation, the cities of Dallas and
Fort Worth, two regional water authorities, the State, and the USGS
assumed funding for the operation and maintenance of the moni-
tors through a cooperative agreement. With the assistance of com-
mittees representing all parties, NCTCOG performs the analysis of
CAMS data and publishes semiannual reports.
As of April 1986, five CAM stations have been installed in the river
basin. Each measures stream flow, dissolved oxygen, temperature,
pH, and specific conductance. The monitors operate continuous-
ly day and night, with values recorded on digital tapes every hour.
The data are retrieved and incorporated into a computerized data
base which is used to compute trends.
Sites for the automated monitors, shown in Figure 1, were carefully
chosen based on hydrology, relationship to USGS flow gauges, and
the location of major dischargers. CAM 1 is located on the West
Fork Trinity River and is upstream of all municipal sewage treat-
ment plants, but downstream of the Fort Worth urbanized area. The
segment receives runoff from urban and rural portions of the
July 1986
-------
watershed and a few industrial dischargers, but none contribute sig-
nificant oxygen-demanding pollutants that would affect the moni-
tor. This site was chosen to characterize river flow and water quality
before it is impacted by a major POTW. The remaining CAM sta
tions are spaced downriver between and below the major POTWs.
CAM 5 is located at the downstream point of the Dallas-Fort Worth
designated planning area, and is almost 100 river miles below the
CAM I station. During the summer of 1986, three more monitors
will be added to the CAMS network, two on the East Fork Trinity
River and one downstream of CAM 5. While treatment facilities on
the East Fork have recently been upgraded, they are now exper-
iencing some of the highest growth rates in the Dallas area—
so these monitors are expected to be valuable additions.
Source: NCTCOG
Figure 1. Upper Trinity River Basin Showing CAM Sites
and Major Dischargers
WAFER QUALITY PROGRESS
Historically, dissolved oxygen (DO) levels have played a major role
in water quality management on the effluent-dominated upper
1 nity River. Effluent limitations for the joint system plants dis-
diging to the Trinity River have been established for BOD as well
is suspended solids based on water quality modeling of anticipated
DO response in the river. For this reason, much of the CAM work
to date has focused on dissolved oxygen.
During the eight-year period of record (1 977.1 985), the five CAM
stations have generated approximately 245,900 valid hourly read-
ings of dissolved oxygen. For most years, there are between 6,000
and 8,700 valid readings per CAM station. This compares with a
record, prior to the establishment of CAMS, of only a few hundred
annual readings at a given location, with most grab sampling con-
ducted during normal business hours.
Funding under Section 205 (j) of the Clean Water Act has enabled
NCTCOG to analyze this extensive data base. Basic statistics and
frequency distributions of DO concentrations have been calculated
for each CAM station. Trend regressions have also been calculat-
ed for each station (except CAM 3 which has been in continuous
operation only since February 1984). The results indicate a general
upward trend for the downriver CAM sites that are impacted by
major wastewater discharges, while the trend line for CAM 1 re-
mained statistically constant. Figure 2 is a graph of the mean
monthly DO values at CAM 2, along with the calculated trend
regression over the eight-year period.
All three CAMS (2,4, and 5) show improvements from low annual
averages of 2 to 3 rng/L DO during 1977-78 up to average concen-
trations in the 5 to 6 mg/L range for the 1984-85 period. Equally
improved during the same period at the three monitors has been
the percent oxygen saturation and the level of mean conductance
values.
As indicated in the graph of CAM 2 data, dissolved oxygen values
can fluctuate widely. This variability depends on several conditions
including water temperature, river flow, BOD, and ammonia levels
in effluent discharges. Future work with the CAM data will he
directed at analyzing for the effects of temperature and other nat-
ural seasonal factors (including flow) in an effort to better represent
the trends in dissolved oxygen.
I
I
Source: NCTCOG
FIgure 2. Mean Monthly Values and Trend Regression for
Dissolved Oxygen at CAM 2, 1977 Through 1985
ANALYZING NONPOINT SO(JRCES AND FISH KILLS
Past analyses have focused on the relationship between ambient
DO levels in the river and wastewater effluent quality—usually at
periods of critical low flow. Increasingly, however, CAMS data are
being used to examine water quality during peak flow events asso-
ciated with rainfall. Many of the low dissolved oxygen events meas-
ured by CAMS during recent years have been associated not only
with the low flow summer conditions, but also with periods during
and immediately following heavy rainfall.
Fish kills in May 1984 and July 1985 followed peak flow events,
and in both cases, dissolved oxygen concentrations at CAM 5 indi-
cated substantial oxygen depletion at that time. The specific cause
of dissolved oxygen depressions during high flow events is not yet
known, since the effect could result from a variety of sources. In
addition to loading from urban and rural runoff, bypasses from
overloaded wastewater treatment plants can occur, and sewage lines
can be damaged during heavy storms. Resuspension of bottom
sediments has also been suggested as an important cause of the
dissolved oxygen sags during high flows.
During the spring and summer of 1986, EPA, the Texas Water Com-
mission, and an associated multi-agency committee are monitor-
ing instream impacts during significant rainfall events. During these
studies, the CAMS data have supplied crucial information needed
for tracking the short-term effects of storms on flow and dissolved
oxygen levels. The EPA Region VI office will sample for heavy
metals, pesticides, and other toxicants as well as conduct toxicity
tests (using biomonitoring techniques) of grab samples taken dur-
ing storm events. Additional studies will be directed at obtaining
more quantitative information on the relative impacts of point and
nonpoint source loadings.
Material [ or this report was furnished by Samuel Brush, !‘ICTCOCI;
Richani McVay, Texas Water Commission; and Cathy (lilmore,
U.S. EPA Region VI.
This report is produced by EPA to document progress achieved in
Improving water quality. Cont, butions ol information for similar reports
are invited. Please contact E. F Drabkowski, EPA, MDSD, WH-553,
401 M Street S.W. Washington. D.C. 20460 (202)382-7056.
14
Apr. 7$ Apr. 79 Apr. 80 Apr. 81 Apr. 82 Apr. 83
MoatS (May 1977 through Octob.r 1985)
Apr. 84 Apr. 85
• Continuous Automated Monitors
eatment Ptgnta
-------
(inited States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
Water Quality Progress Report
Tillamook
Bay,
Oregon
The Tillamook Bay drainage
basin in northwestern Ore-
gon includes five major river
systems and Tillamook Bay.
The bay, which covers nearly
11,000 acres, is the most
productive oyster and clam
growing water in Oregon, producing 80 percent of the oysters com-
mercially harvested in the State and generating between 2 and 3
million dollars per year in revenue. However, much of the lowland
areas adjacent to the bay and rivers are used intensively for dairy
operations, with nearly 20,000 dairy cattle on 118 farms lining the
lower portion of the watershed. The presence of concentrated
livestock wastes along with the region’s wet weather (about 100
inches of rainfall per year) have created severe runoff problems and
contaminated conditions in Tillamook Bay.
Over the last ten years, following moderate to large storms, the bay
has been closed to commercial shellfishing due to high concentra-
tions of fecal coliform bacteria. Affected along with the commer-
cial oyster industry has been recreational clam digging, fishing,
boating and other activities on the river and tributaries that attract
nearly a million tourists and sportsmen to the area each year.
In 1981, Tillamook Basin was selected as one of 21 project areas
to be funded under the Rural Clean Water Program (RCWP) which
is sponsored jointly by the U.S. Environmental Protection Agency
and the U.S. Department of Agriculture. This program is designed
to encourage the implementation of best management practices
(BMPs) to control nonpoint source pollution, and to evaluate the
effecti ’iiess of individual BMPs. Although implementation of
BMPs is not yet complete, the Titlamook Bay project has been able
to show significant water quality improvements in both the rivers
and the bay. In 1985 bay closures were invoked less frequently,
employment in Tillamook’s oyster industry was the highest since
1952.
ASSESSING THE PROBLEM
The fecal coliform(FC)standard for commercial shellfishing waters
is a log-mean of 14 cells per 100-mL sample. In late 1977, the U.S.
Food and Drug Administration surveyed shellfishing areas in Tilla-
mook Bay and found FC concentrations far in excess of the stand-
ard; the agency strongly recommended closure of the bay for oyster
harvesting until appropriate controls could be developed. In
response, the Oregon Department of Environmental Quality (DEQ)
initiated the Tillamook Bay Bacteria Study to comprehensively
assess the sources, extent, and dynamics of fecal contamination in
the watershed. There were three major aspects to the study.
Water Quality Monitoring. During 1979 through 1981, DEQ con-
ducted an intensive weather-related survey to determine fecal coli-
form densities and to identify the major sources of fecal conS
tamination in the basin. The survey included 79 river and stream
stations selected on the basis of adjacent land use, 14 bay sampling
stations selected for proximity to shellfish growing areas, and the
5 small sewage treatment plants that discharge either to the bay
or into the lower reaches of one of the major rivers. Because non-
point source loading is closely related to precipitation and soil con-
ditions, water quality data were collected during four different
weather periods: (1) heavy rain on saturated ground, (2) rain after
a period of dry weather, (3) summer low-flow during dry weather,
and (4) first “freshet” storm at summer’s end with sampling begin.
fling prior to soil saturation.
Data were analyzed by comparing the concentration at each sta-
tion for each weather event against the fecal coliform standard. By
examining monitoring results in light of upstream and surround.
ng land uses, DEQ established that the primary sources of con-
tamination were surface water runoff from dairy operations,
inadequate onsite septic systems, and malfunctioning sewage treat-
ment plants. Of the three, agricultural operations were found to be
the single largest source.
Analysis of Dispersion and Residence Time. The fecal coliform
standard for rivers and tributaries that flow into the Tillamook Bay
(waters that are designated for recreational use) is over 10 times
higher than is allowed for shellfishing. For this reason an important
study was conducted to determine the dispersal and purging of the
bacteria contained in rivers that entered the bay near sheilfishing
beds Data from dye studies indicated that Ibliowing moderate rainS
fall, the bay purges itself in 2 to 3 days; following periods of heavy
rains, especially during the winter, purging takes 3 to 7 days. These
relatively short residence times make water quality in the bay
responsive to changes in FC inputs.
Development of Harvesting Criteria Using the results of monitor.
ing and dispersion studies, DEQ and the State Health Department
developed a set of five criteria, specific to the Tillamook watershed,
any one of which can be used to close shellfish beds for 5 to 10 days
The criteria, implemented in 1982, are based on rainfall and river
flow conditions or on known sewage bypasses. In 1987, these
closure criteria will be re-evaluated based on continuing FC mon-
itoring results and trends.
INSTITUTING CONTROLS
To attack the problem of animal waste management, the Tillamook
Soil and Water Conservation District (SWCD), with funding under
Section 208 of the Clean Water Act, developed an extensive non-
point source pollution abatement plan in 1981. Although the plan
specified that waste management practices should be individual-
ized for each farm, the overall strategy relied on two basic princi-
ples (1) prevent rainwater and clean surface water from coming into
contact with manure and (2) when this is not possible, prevent con-
taminated surface water from reaching the streams or the bay.
One hundred nine dairy farms, covering 9800 acres, were desig-
nated as “critical” dairies, and these farms were eligible, under the
RCWP project, to receive upto 75 percent of the cost of implement-
ing approved BMPs up to $50,000. To achieve the goal of a
70-percent reduction in fecal coliform loading, all critical dairies
must undertake BMPs. As of July 1986, contracts to implement
BMPs had been signed by 103 of the 109 dairies, and the Tillamook
SWCD, which is administering the RCWP project, has received over
$4 million in cost-share funds. Farmers in the project area have
committed more than $2 million of their own money. The most
common BMPs were the installation of underground storage tanks
for manure, the addition of gutters to ns to control runoff, and
EPA
August 1986
-------
the construction of fences to keep cattle out of streams.
The two nonagricultural sources of coliform bacteria were also
addressed in the SWCD plan. A study of the five sewage treatment
plants located in the drainage basin found that while treatment
levels were adequate when the plants operated properly, malfunc-
tions did occur. DEQ worked with the staff of each plant to improve
monitoring as well as operations and maintenance procedures to
ensure that malfunctions occurred less frequentl in addition, each
plant agreed to install alarms to alert operators in case of equip-
ment failure. The second nonagricultural source of pollution was
failing onsite septic systems, and many of these have been elim-
inated by extending a municipal sewer line.
WATER QUALITY MONITORING
The bay and tributaries in Tillamook basin were sampled for many
years prior to the initiation of the RCWP project in 1981. As part
of the RCWP, this sampling was continued at 14 bay sites and
12 river stations. Bay sites were sampled quarterly for salinity, tem-
perature, and fecal coliform; tributary sites were sampled month-
ly for pH, temperature, stream flow, and fecal coliform. All data are
entered into EPAs STORET system.
Starting in January 1985, DEQ and the SWCD began a coordinated
2-year monitoring effort using monies available under Section
205 (j). This survey has been designed to both assess the effective-
ness of site-specific BMPs and to evaluate progress toward reduc-
ing fecal coliform concentrations in the bay and tributaries. The first
part will be accomplished by focusing on smaller tributary sites
where the effectiveness of individual BMPs can be evaluated by
monitoring comparable sites above and below dairies. The second
objective will be accomplished by monitoring all bay and river sta-
tions during weather conditions as close as possible to those mon-
itored in the 1979 to 1981 DEQ study. The resulting data will aid
in evaluating progress, as well as provide an overall assessment of
the effectiveness of BMPs.
WATER QUALITY IMPROVEMENT
Tlllamook Bay. Coliform concentrations in Tillamook Bay are the
result of complex interactions involving upstream manure inputs,
prior precipitation, stream flow, estuary tide stage, and bottom-
stirring action by winds. As a result, a large amount of monitoring
data and some relatively sophisticated analytical techniques were
needed to isolate water quality trends in the bay. An additional com-
plication was introduced since post -BMP monitoring focused more
on runoff events (when coliform concentrations would be expected
to be higher) than did pre-project monitoring. However, with assist-
aqce from the National Water Quality Evaluation Project (NWQEP),
Oregon has been able to document unequivocal success in improv-
ing the impaired water resource.
Table I shows the percentage reduction in FC concentration at the
eight bay sampling sites that are located within oyster and clam
beds. (Six other bay sites are within channels formed by the tribu-
taries draining into the bay, and as such represent incomplete mix-
ing zones with expected higher coliform levels.) The percentage
reductions are based on a linear regression model that uses salin-
ity measurements to account for meteorologic differences in pre-
and post-treatrn nt sampling.
DEQ estimates that mid-1983 was the point at which over half of
tl ’ total BMPs were installed or under construction; thus, as
remaining BMPs are completed and the number of observations
is increased, FC reductions will be quantified with greater certainty.
TABLE 1. Fecal Coliform Reductions in Tillamook Bay
(pretreatment: 1/1975 - 6/1983; post-treatment: 7/1983 - 7/1 985)
Bay sampling
site
Percent reduction
FC concentration
I
56%
6
41%
7
56%
8
59%
10
16%
11
43%*
13
60%*
14
64%*
Mean FC
reduction = 49.4%
Source: NWQEP 1985 Annual Report
aStatistically significant at 95% confidence level.
Table 2 shows reductions in FC concentration at sampling sites
along the five rivers. Using January 1982 as the dividing date be-
tween pre- and post-treatment, four of the five river sites show
statistically significant reductions. Again, as more data become
available, the evaluation can be improved, since interim data (data
collected during the time when BMPs were being installed) can be
deleted from the pre- and post-data sets to obtain a clear picture
of results.
TABLE 2. Fecal Coliform Reductions in Tillamook Rivers
(pretreatment: 1/1975- 12/1981; post-treatment: 1/1982 -6/1985)
River (Site)
Log mean
FC concentration
Percent
reduction
P i e
Post
Kilchis (K4)
87
61
30%
Miami (MM4)
276
60.7
78% *
Trask(TR8)
168
63.4
62%*
Tillamook(T4a)
387
162
58%*
Wilson (W13)
147
68.6
53%*
Source: NWQEP 1985 Annual Report
*S tistically significant at 95% confidence level.
FUTURE PROJECTS
Results of the intensive survey now underway in the Tillamook
basin will be used to develop a bay water quality model. The model
will more accurately describe the water quality effects following
specific precipitation events, and it will be used to refine and sup-
plement the bay closure criteria that are now used to determine
when shellfishing is hazardous.
Material for this report was furnished by John Jackson, Water Quality
Division, Oregon DEQ; The National Water Quality Evaluation
Th-oject, North Carolina State University; and Elbert Moore, U.S. EPA
Region X.
Rivers. Five major rivers empty into Tillamook Bay and each of
these has numerous tributaries often with only a few dairies located
in their drainages. Tributary water quality data are being analyzed,
focusing on both “before and after” changes in mean FC concen-
trations and “above and below” trends in several small tributaries
where only one or two dairies have implemented BMPs. Definitive
results on these studies await monitoring data from the intensive
survey to be conducted during 1985-86.
This report is produced by EPA to document progress achieved in
improving water quality. Contributions of information for similar reports
are invited. Please contact E E Drabkowski, EPA, MDSD, WH-553,
401 M Street S.W, Washington, D.C. 20460(202) 382-7056.
-------
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
EPA Water Quality Progress Report
w w W W
Passaic River,
New Jersey
The freshwater portion of the Passaic
River is 70 miles long and drains ap-
proximately 800 square miles in
northeastern New Jersey. In the
upper two thirds of the basin, the
river is a slow moving stream that
flows through residential and rural
land. Municipal treatment plants are
a major source of organic pollution,
and high nutrient loading in this subbasin. The lower third of the
river drains the densely populated and older urban area of Pater-
son, New Jersey. Along this section, there are approximately 35
industrial point sources and eight major publicly owned treatment
works (POTWs). The Passaic River also serves as a source of potable
water for many of the urban and suburban communities in
northeastern New Jersey, and, as such, water quality in the river is
a major pubic health concern.
By the 1960’s, because of rapid population growth and industrial
development in the basin, water quality in the Passaic River had
become severely degraded. Large diversions for water supply and
industrial use reduced the capacity of the river to assimilate
wastewater, and violations of dissolved oxygen and ammonia stan-
dards were common during the summer months. Domestic
sewage was identified as the primary source of oxygen demanding
waste, and, consequently, efforts to improve water quality and meet
the dissolved oxygen (DO) standard focused on upgrading treat-
ment levels at POTWs.
While several sections of the Passaic River still exceed the standards
for dissolved oxygen and ammonia toxicity during the summer
months, monitoring efforts have demonstrated significant im-
provements in the river’s water quality. Improvements have been
documented both by routine monitoring over the last ten years and
by intensive surveys completed before and after upgrading a P01W
from secondary to advanced wastewater treatment (AWT). This
report documents continuing efforts by the State of New Jersey to
restore water quality and to refine management options in the
Passaic River basin.
PREVIOUS MONITORING STUDIES
In 1973, a report prepared by the U.S. Geological Survey (USGS)
examined water quality and streamflow in the Passaic River Basin
above Little Falls, New Jersey. This portion of the watershed
includes about 80 percent of the total basin, and its streams are
designated primarily for water supply. A statistical analysis of water
quality data in STORET(EPAs water quality file) showed that water
quality in the basin had deteriorated steadily from 1945 to 1970.
An upward trend in the concentrations of dissolved solids, hard-
ness, chloride, and sulfate was evident during this period as was a
gradual decline in DO.
In 1976 and again in 1979, the New Jersey Department of
Environmental Protection (NJDEP) carried out comprehensive
water quality management studies in the Passaic River basin with
funding under Sections 303(e) d 208 of the Clean Water Act. To
investigate the water quality impacts of projected pollutant
loadings. NJDEP utilized a steady-state model to simulate levels
of DO, carbonaceous BOD(CBOD), and nitrogenous BOD(NBOD)
in the Passaic basin. This model was based on the original U.S.
Environmental Protection Agency (EPA) SNSIM model, which uses
an expanded form of the Streeter-Phelps equation. Based on the
projections of this modified model, the State established water
quality-based effluent limits for most dischargers and AWT
requirements for nearly all POTWs.
However, the SNSIM model used for the 208 and 303(e) studies
was inadequate in several respects. Because field data were not
available, the model did not adequately represent photosynthesis,
sediment oxygen demand (SOD), and nitrification processes. Also,
nonpoint source loadings were not addressed, and some important
hydraulic characteristics were not modeled. It was felt that to better
evaluate management options and to justify the stringent AWl
requirements, the State should collect new field data and employ
a more complex water qualit ’ model.
1983-84 PASSAIC RIVER STUDY
To support the application of a more comprehensive water quality
model, NJDEP conducted three intensive surveys (August 1983,
October 1983, and September 1984) of the entire Passaic River
watershed These surveys, carried out during low-flow conditions,
each lasted two or three days, and provided a comprehensive
“snapshot” of water quality in the river. Altogether, 44 sampling
stations were located to account for prominent hydrologic features
as well as point and nonpoint discharges. Every station was
sampled once on each day of the survey period, and at each station,
three samples were collected across the width of the river and then
combined to provide a single composite for the station.
The 1983-84 Passaic River Study, partially funded under Section
205(j) of the Clean Water Act, yielded data that have been used for
three major tasks. First, EPA’s QUAL II model was calibrated and
verified using the survey results (i.e., model parameters were
adjusted to fit site-specific conditions). Second, NJDEP predicted
the present and future water quality response under various
wasteloads and water diversions. The model was used to simulate
hydraulics and water quality for conventional pollutants, including
DO, CBOD, nitrogen species, dissolved phosphate, dissolved solids.
and chlorophyll a. This information was used to establish effluent
limitations, including seasonallimits, for individual dischargers.
Finally, water quality improvement due to previous point source
controls was evaluated.
MODELING RESULTS
Once it was calibrated and verified using the 1983-84 monitoring
data, QUAL II was used to gain a better understanding of the
physical and chemical processes at work in the river. For example,
this model enabled NJDEP to analyze the DO deficit (the dif-
ference between measured and potential DO concentrations) along
the length of the river. The model calculated the relative contribu-
tion of the three principal oxygen-demanding components (DO
sinks) CBOD, NBOD, and SOD, as well as the effect of the
individual oxygen producing components, reaeration and net
photosynthesis. The results for the deficit sink components are
shown in Figure 1.
Figure 1 shows that the range of values and the importance of in-
dividual deficit components vary greatly over the length of the river.
The two DO source components, reaeration and photosynthesis
(not shown in the figure) act to reduce the DO deficit, and these also
vary along the length of the river. Reaeration dominates in the
October 1986
-------
-J
E
C -,
5.
0
—60
SOD
NBODLI
Rivermile (FLOW —)
Figure 1. Components of the dissolved oxygen deficit.
upstream areas where its magnitude counterbalances high SOD.
Downstream, in backwater areas, algal photosynthesis provides the
greater portion of available DO.
There are several important ramifications of the deficit component
analysis. First, because the major determinants of DO in the
upstream reaches are sediment deposits and reaeration, this area
of the Passaic River is likely to show only marginal improvement
as a result of reduced pollutant loading from upgrading POTWs
beyond AWT In addition, the segments just downstream of two
waterfalls (mile 12 and 5)showa high SOD. Although these zones
are char terized by high reaeration and a sharp DO peak (or valley
on the deficit curve), this DO is quickly consumed. Thus, the signifi.
cant reaeration occurring at the waterfalls affects only a small
portion of the river.
WATER QUALITY PROJECTIONS
Currently there are 12 POTWs, with a combined design flow of ap-
proximately 50 million gallons oer day (mgd) in the mainstem or
near to the mainstem Passaic River. While several POTWs have
been renovated to provide advanced treatment, the majority of
plants hold permits for secondary treatment. Because of the high
costs involved in upgrading to AWE an important objective of the
Passaic River Study was to determine whether the comprehensive
AWT requirement would be effective for restoring water quality in
the river. The model was used to examine water quality outcomes
(under critical low flow) for three conditions: actual point source
discharges, existing permitted discharge levels, and recommended
discharge levels assuming AWT at all treatment plants.
Model results indicated that under summer Iow.fIow conditions,
existing effluent levels and the current permitted levels have a
similar impact—extensive violations of DO and unionized
ammonia standards along the river. In several sections, the
minimum average DO concentration could drop to zero. Cinder the
advanced treatment alternative, the portion of the river experienc.
ing DO violations would be significantly reduced, and the
minimum DO concentration would be upgraded to 2.8 mg/L and
4.2 mg/L in the upper and mid.Passaic River, respectively. In
addition, no ammonia violations were predicted under the AWT
scenario.
Of particular interest is the fact that some areas which continue to
show DO violations under the AWT scenario contain no significant
point source discharges. In these segments the DO violations can
be attributed to nonpoint sources (e.g., SOD or bank erosion) or
natural processes such as low reaeration rates. Consequently, ef-
forts to upgrade treatment levels beyond AWT cannot be justified
by the model due to nonpoint source impacts on the river.
WATER QUALITY IMPROVEMENTS
As a result of past and ongoing federal and State efforts to upgrade
treatment levels, measureable water quality improvement has been
observed in parts of the Passaic River system. A review of historical
monitoring data indicates that pollution levels peaked in the 1960’s
and since then there has been a gradual trend toward improved
water quality.
Direct evidence that upgrading treatment levels does improve
water quality was obtained during NJDEP’s 1983-84 Passaic River
Study. Bernards Township sewage treatment plant discharges
approximately 1.5 mgd of treated effluent to the Dead River, an
important headwater tributary of the Passaic River. This plant was
upgraded from secondary to advanced waste treatment in
1984—prior to the last NJDEP survey and after the 1983 surveys.
Table 1 shows a comparison of water quality before and after treat-
ment plant improvements were completed. Measurements were
taken at a station downstream of the outfall.
TABLE 1. Water Quality Improvement Downstream
of Bernards Township Treatment Plant (mg/L).
Survey
DO
BOD 5
NO 3 -N
TKN*
TP”
August 1983
3.2
10.3
0.75
9.94
2.53
(secondary WI)
September 1984
8.4
1.55
11.0
1.07
2.35
(advanced WT)
Source: NJDEP
*TKN Total Kjeldahl nitrogen, TP = Total phosphorus.
Except for nitrate (NO 3 ) and phosphorus, a dramatic improvement
in water quality of the Dead River is evident from the ‘before and
after” study. The concentration of nitrate increased due to nitrifica.
tion of the plant effluent, and although phosphorus limits for point
sources are under consideration by NJDEP, currently there are no
phosphorus removal requirements for the Passaic River.
A major component of nonpoint source pollution in the Passaic
is benthic deposits. These deposits, which are the result of both
natural conditions and previous wastewater discharges. can serve
as a dissolved oxygen sink for years before they are fully oxidized
or become buried in deep sediments. Nevertheless, the 1983.84
Passaic River Study did show evidence that a decrease in the oxy-
gen demand of benthic deposits had occurred. This reduction was
particularly significant at the station downstream of the Whippariy
River confluence which has historically shown high levels of SOD
(a measure of DO consumption by sediments). In 1973, SOD at
this station was measured as 10.1 to 12.8 g EXYm 2 kJay; in 1978,
SOD was 1.4 to 3.6: and in 1983, the measured SOD ranged from
below detection limits to 5.6(both in situ and laboratory measures
were employed). This improvement may be attributed, at least in
part, to upgrading treatment at several POTWs(located upstream)
and the removal of a large industrial discharge.
FUTURE STUDIES
The 1983-84 Passaic study was the first step in an ongoing project
to assess water quality in the river. Future monitoring studies will
include sampling for toxic constituents and biomonitoring for sedi-
ment and water column toxicity in the river. In addition, NJDEP
hopes to conduct additional studies of SOD rates in the river since
this parameter appears to be a dominant DO sink during low flow
periods.
Mat erial for this report was turn Lshed by Dr. Shing.Fu Hsueh, Bureau
of Water Quality Standards and Analysis, HJDEP; and Rocdlla
O’Connor. U.S. EPA Region II.
This ‘ ‘r ’)rL is produced by EPA to c’ocwncnl prr jress achieved in
Irnproe1n ( 1 at : quality. ontrthutions ol ,nlorrnatinn for sirni ar reports
amin (ft(I. ;.:easecon(acLE F DrabkocL’ A,, EPA, 1 DSD. WH-553,
401 M Street SW, Washington. D.C. 20460(202) 82 7056.
20
15
CBOD
10,
—50 —40 —30
—20 —10
-------
United States
Environmental
Protection Agency
Febnj ry 1987
Water Quality Progress Report
_ - -
w w
Koshkonong
Creek,
Wisconsin
Over the past 5 years, the Wiscon-
sin Department of Natural Re.
sources (WDNR has conducted
more than 50 monitoring surveys
to document water quality impacts
resulting from the construction of
new or upgraded wastewater treatment plants. The primary objec-
ti ’ of these surveys is to ensure that water quality standards are
bcing met: a secondary objective is to evaluate mathematical
models that may have been used to assign effluent limits. Most of
the surveys to date have been carried out on small streams where
advanced wastewater treatment (AWT) was required.
WDNR has developed a streamlined procedure in which effluent
and instream water samples are collected during a 1- or 2.day
pt ’i iod of low stream flow both before and after the startup of a new
or upgraded treatment plant. Biomonitoring is included in the
procedure through the use of a macroinvertebrate screening sur-
vey. This survey is usually carried out during the spring and fall prior
to, and I or 2 years after the improved treatment processes have
been in use.
This report documents one such monitoring study, conducted on
Koshkonong Creek in southern Wisconsin. This creek is one of
many small streams that has been subject to gross organic pollu-
tion due to inadequate wastewater treatment. Construction of the
rlt’w Sun Prairie publicly owned treatment works (P01W), an AW
facility with ammonia removal and tertiary sand filters, has brought
about greatly reduced pollutant loadings and a corresponding
improvement in ambient water quality.
THE STUDY AREA
Koshkonong Creek is a channelized stream draining 138 square
miles near Madison, Wisconsin, The creek originates in the town
of Sun Prairie (population approx-
imately 14.000) and flows 42 miles
through agricultural areas to join the
Rock River. The creek is effluent
dominated with most of the flow
originating as the discharges from
Sun Prairies P01W and cooling i
water from two industrial facilities.
While much of Koshkonong Creek
has been dredged and straightened
to facilitate agricultural drainage, the
stream has a very low gradient and
many channelized portions are
clogged by vegetation. Consequent-
ly. flow in the creek is sluggish, and
most of the creek bottom is covered
by a foot or more of silt. Figure 1
shows the configuration of the creek
along with the locations of water
mr r tr’ .,i ‘ ns iced in the
Figure 1. Koshkonong Creek
showing the location of
Largely as - result of these physical constraints (channelizat ion and
low flow), the creek’s current designated use, from the headwaters
to station C-5. is ‘marginal surface water:’ As such, the stream is
considered suitable for only very tolerant aquatic insects and forage
fish. Below station C-5, the creek is designated for warm water sport
fishing, full aquatic life, and surface water recreation such as
canoeing.
PREVIOUS MONITORING STUDIES
Water quality monitoring conducted on Koshkonong Creek during
the 1970’s showed severe organic pollution below the original Sun
Prairie POTW. This treatment facility was hydraulically overloaded
and discharged high levels of biochemical oxygen demand (BOD)
and nutrients into the creek. Because of the low natural stream flow,
dilution of the wastewater was insignificant, and the concentrated
organic wastes stimulated nuisance growths of filamentous bacter-
ia extending nearly 4 miles below the plant. Samples of the macro-
invertebrate population (bottom-dwelling aquatic and other
organisms) also indicated very poor water quality below the former
treatment plant.
Base flow samples collected in 1977 at station C-2 showed that the
creek was often devoid of oxygen in the summer, with BOD reach-
ing levels of 15 to 60 mgIL While ammonia levels were normal in
the short stretch of stream above station C-2, below that point, levels
were quite high with an average concentration of 7 mg/L and a max-
imum of 14.5 mg/L ammonia. The 1977 survey also showed
extremely high average values for conductivity, chloride, fecal coli-
forms, and total phosphorus.
“BEFORE AND AFTER” SURVEYS: CHEMICAL STUDIES
Water samples were collected in 1981 (before AWT operation I and
in 1982 (after AWT operation) at one upstream station (C-I), the
original POTW outfall(C.2), and 3 downstream stations (C-3, C.4,
and C.5). Samples were analyzed at the Wisconsin State Laboratory
of Hygiene.
Pre .operatlonal Survey. Consistent with previous monitoring
results, the 1981 pre-operational survey indicated significant water
quality degradation. Assimilation of the high strength organic
wastes lowered the upstream dissolved oxygen (DO) level of 8.5
mg/L to 2.5 and 3.4 mgfL at downstream sampling stations C-3 and
C-4. respectively. At station C-5, approximately 6 miles below the
former discharge point, the stream showed a partial recovery with
a maximum DO concentration (noon reading) of 6.7 mgIL
However, a significant daily fluctuation in DO levels was observed,
particularly at station C-5. Here, an early morning concentration of
only 0.9 mg/L was apparently caused by respiration of abundant
aquatic plants at this location. At station C-3, the daily DO swing
was less significant, probably because filamentous bacteria covered
all available substrates. At C-4, the substrate was covered with a
combination of filamentous bacteria and algae, and the DO fluc-
tuation was of intermediate magnitude.
High fecal coliform and ammonia concentrations during the pre-
operational survey also reflected polluted conditions below the
former POTW. Upstream (station C-i) concentrations were 700 or-
ganisnis/100 mL fecal coliforms and 0.06 mg/L ammonia. At C-3,
fecal coliform and ammonia concentrations were 50,000 organ-
isms/100 mLand 17 mg/L respectively. At station C-5. which marks
the starting point of the full fish and aquatic life classification z ne,
Monitoring and Data Support Division
Office of Water
Washington. DC 20460
C-4
-------
ceeded 1 6 mg/L to ensure a healthy environment for fish and
aquatic life
Post-opeiatlonal Survey. The new Sun Prairie P01W, with a
capacity of 3 1 million gallons per day, was built to accommodate
both municipal waste and substantial seasonal pollutant loading
from a local cannery The facility began operating in December
1981, and the post-operational chemical survey was performed in
July 1982 Effluent monitonng data showed that BOD and sus-
pended solids concentrations had been reduced by 45 percent and
92 percent, respectively Total Kjeldahl nitrogen (TKN, a measure
of organic nitrogen plus ammonia nitrogen) dropped from 23 mg/L
to 1 1 mg/L
Water quality in Koshkonong Creek below the new wastewater treat-
ment plant reflected the improved effluent concentrations. As
shown in Table 1, BOD, TKN, and ammonia concentrations were
low, and nitrate- plus nitrite-nitrogen (NO 2 +N0 3 —N) concentra-
tions were high. Decreased levels of TKN and ammonia and in-
creased NO 2 + N0 3 —N concentrations show effective assimilation
of nitrogenous wastes in the treatment plant
TABLE 1. Water Quality Improvement Downstream
of Sun Prairie P01W (mg/L)
Station
No.
Year
BOD
TKN
Ammonia
nitrogen
N02 +
N0 3 -N
C-4
1981
1982
8.0
3.3
120
08
95
006
0.15
5.7
C-5
1981
1982
6.8
24
120
10
100
008
006
61
The increase in DO concentration, as well as the relative magnitude
of the diurnal oxygen shift, is shown in Figure 2 The persistence
of low nighttime DO levels is caused primarily by aquatic plant
respiration although residual sludge deposits in the sediments also
may contribute significantly to the total oxygen demand.
Figure 2. 1981 and 1982 diel dissolved oxygen measurements.
“BEFORE AND AFTER’ SURVEYS: BIOMONI1OR ING SIC] DIES
To further assess water quality improvements following startup of
the Sun Prairie POTW, WDNR staff conducted biomonitoringsur-
veys during the spring and fall of 1980 and 1983 Macroinver-
tebrates were collected (at the same sampling stations used during
the chemical surveys) using a D-frame net and lock sampling to col-
lect these bottom-dwelling organisms Samples were preserved in
95 percent ethyl alcohol in preparation for laboratory sorting and
identification
Macroinvertebrate samples were used to determine the Hilsenhoff
Biotic Index at each station. Developed by W L Hilsenhoffat the
Llnivprsitv of Wiscnns i r 1 -n this method uses the first 100
c C-3 C-4 C-5 C-6
Sampling Station
-i
E
0
0
100
80
60
40
20 -
arthropods (insects. aniphipods. and isopods) in a sample to eval-
uate the water quality of a stream The index values, which are
based on the varying tolerances of different species to organic po 1 .
lution, range from 0 to 5, with lower values reflecting better water
quality A value ofO is assigned to species found in pristine streams
of high water quality, while a value of 5 indicates species tolerant
of severe organic pollution The biotic index is an average tolerance
value for the entire sample
Upstream of the P01W (station C-I), biotic index values of 288 to
374 indicated fair (1983) to poor (1980) water quality Such an
unbalanced macroinvertebrate community reflects the intermittent
and marginal characteristics of the Koshkonong Creek headwaters.
At station C.3, 1980 surveys resulted in biotic index values rang-
ing from 463 to 50 These values indicated very poor water qual-
ity and severe organic pollution In all cases, the benthic
macroinvertebrate communities were limited to two to four very
pollution-tolerant species In one sample, only eight specimens
could be found. Macroinvertebrate samples collected (luring the
post-operational surveys showed a slight increase in pollution-
intolerant species and greater species diversity Figure 3 shows the
average biotic index values for the 1980 and 1983 surveys at each
stream station The limited recovery of the macroinvertebrate com-
munity may be due in part to the continuing effect of abundant
filamentous algae and macrophytes that consume available
oxygen
c i
ci
C
i-i
a
5.0
40
3.0
2.0
C-i C-3 C-4 C-5
Sampling Station
Figure 3. 1980 and 1983 average biotic index values.
The most obvious water quality improvement following completion
of the new wastewater treatment plant was the elimination of
filamentous bacteria in the stream Effective treatment at the new
facility eliminated the nutrient rich conditions required for this
growth, and as a result, algae and other periphyton typical of similar
streams in the basin have returned to Koshkonong Creek
FUTURE IMPROVEMENTh
Koshkonong Creek is an example of a small stream that has been
saved from gross organic pollution by the construction of a new
P01W While the physical characteristics of the stream may limit
its biological quality, establishment of more diverse aquatic com-
munities in the upper 8 miles of the creek is expected to occur over
several years as sludge deposits are gradually reduced As this
progress continues, WDNR will penodically review the streams use
classification for possible upgrading Additional improvements in
the lower portions of the creek, which are classified for full fish and
aquatic life uses, are expected as the quality of the upstream water
is improved
Matenal for this report was furnished by Jem j McKersie, Chief, Eva!-
uatzon and Special Projects, WDNR, Water Resources Management,
Dave Marshall, WDNR Water Resources Management, Southern Drs-
tnct, arid Noel Kohl, US EPA Region V
This report is produced by EPA to document progress achieved in
irnprovuig water quality Contnbuborts of infomiation for similar reports
are invited Please contact E F Drabkowskt, EPA, MDSD. WH 553,
401 frI StreetSW, Washington, DC 20460 (202) 382 7056
80
83
— 1981
-—• 1982
1 = noon readings
2 = readIngs before sunrise
——.*
I I I I
-------
United States
Environmental
Protection Agency
Office of Water Regulations and Standards
Monitoring and Data Support Division
Washington, DC 20460
WW
rWL L
Whitewood
I Creek,
1 South Dakota
I J Whitewood Creek flows for
____________ 32 miles through the Black
Hills of South Dakota to its
confluence with the Belle Fourche River. For over 100 years, the
stream was polluted by mine tailings and mill wastes from a num-
ber of gold mining and milling operations and raw sewage from the
towns of Lcad and Deadwood. Although most mining companies
had ceased operations in the area by 1920, the ft rgest of them,
Homestake Mining Company. continues to operate. Wastewater
from the mine (containing, at various times, high levels of mercury,
arsenic, cyanide, and other metals as well as suspended solids)
effectively killed all life in the river and turned the Creek bleak grey
in color. As far as 20 miles downstream from Whitewood Creek, the
Belle Fourche River was devoid of aquatic life, and pollutants from
these discharges were detected through nearly 200 river miles to
the Missouri River. In 1972, Whitewood Creek was listed by South
Dakota as “the most polluted stream in the State.
During the past ten years, S0L!th Dakotas Department of Water and
Natural Resources (DWNR) and Department of Game, Fish and
Parks(DGFP), as well as the Lead-Deadwood Sanitary District, the
Homestake Mining Company (HMC), and the U.S. Environmental
Prot r •Aqt ncy(EP \, have worked together to resurrect White-
wood Creek. Significant activities have included creekbed resto-
ration, upgraded municipal wastewater treatment, the construction
of a tailings pond and an innovative waste treatment process for
cyanide and other heavy metals from the mine, and the use of tox-
icity testing by EPA and the State to diagnose remaining water qual-
ity problems in the river.
While the creek still shows some effects of past discharges, it has
improved enough so that trout and other pollutant-sensitive spe-
cies are now present throughout the stream. As a result of this
progress, DWNR has considered nominating Whitewood Creek for
inclusion in the State’s system of Scenic and Recreational Streams.
GOLD MINiNG ON WI-IITEWOOD CREEK
Gold was discovered in the Black Hills in 1874 during an expedition
led by Lt. Colonel George A. Custer. The first mine and mill were
constructed in 1876, and HMC was incorporated in late 1877.
Although mining and milling technology have changed over the
last 100 years, the basic approach for obtaining gold has remained
the same: underground rock containing the gold ore is pulverized
to a fine sand and treated with chemicals to dissolve, then precipi-
tate the metal. Early methods used mercury compounds to
separate gold from the ore, but in 1970 HMC switched to a com-
plete cyanidization process. Wastewater from the mine consists of
both a tailings slurry (crushed rock that remains after the gold has
been removed) containing the residual cyanide and other heavy
metals and mine water from drilling, cooling, and other processes.
IMPLEMENTING CONTROLS
Following the passage of the Clean Water Act amendments in 1972,
DWNR classified Whitewood Creek as a “cold water fishery” (suit-
able for reproduction and propagation of coldwater fish). In 1981,
a comprehensive bioassessment conducted by HMC was used to
reclassify the creek as a “cold water marginal fishery:’ This classifi-
cation requires water quality that is suitable for stocked coldwater
fish during portions of the year, but due to critical natural condi-
tions(e.g., low flow, siltation, or warm temperatures) is not expected
to support a permanent coldwater fish population.
HMC was issued an NPDES permit that required the company to
drastically reduce its discharge of suspended solids, cyanide, and
metal compounds in order to support the coldwater fishery stan-
dard. To reduce the discharge of suspended solids, HMC conS
structed a large tailings pond upstream of the mine using a dam
280 feet high and 1220 feet wide. The pond, which allows sedi-
ments to settle out prior to passage through a waste treatment
plant, was completed in 1977 and won a national award from the
Council on Environmental Quality and the Environmental Indus-
try Council.
To investigate methods for treating its wastewater, HMC constructed
experimental pilot plants as well as facilities to conduct bioassays
on effluents from the different processes However, many of the first
attempts to chemically remove the cyanide failed. Homestake’s
engineers found that the metal complex cyanides used in the mill-
ing process (e.g., iron cyanide and copper cyanide) could not be
removed using conventional cyanide treatment processes.
As a result of these initial difficulties, by 1980 Homestake was out
of compliance with its permit and still without an effective treat-
ment process. At this point a three-party consent decree signed by
South Dakota, EPA, and HMC permitted the company to extend the
deadline for NPDES compliance, but required the payment of
$390,000 to begin the rehabilitation of Whitewood Creek. While
the consent decree did not contain specific effluent limits for
cyanide or other metals, all parties did agree on a common goal:
that water quality should support the designated beneficial use of
the waterbody.
Working under the consent decree, HMC began testing new treat-
ment approaches. and in 1982 the company obtained encouraging
results with a simple and cost-effective biological treatment process.
Subsequent testing (using chemical analysis and onsite bioassays)
proved the effectiveness of the process. and in August 1984 a full-
scale treatment plant went on-line. The process, which has been
patented by Homestake, uses a mutant strain of bacteria that has
been acclimated to the high levels of cyanide found in the mine’s
effluent. (This bacterium, which has been designated Pseudomo-
nas mudiock, is named after Homestake’s chief environmental engi-
neer and chief chemist, T. I. Mudder and J. C. Whitlock.
respectively.) Water from the tailings pond is mixed with mine water
and pumped to the treatment plant where the bacteria break down
complex cyanides into comparatively harmless sulfates, car-
bonates, and nitrates. To handle the system’s design capacity of
nearly 5.5 million gallons of wastewater per day, the bacteria are
attached to 48 rotating biological contactors (RBCs), each of which
contains approximately 100,000 square feet of surface area. In addi-
tion to the RBC process, the plant also includes a sand filtration sys-
tem (for the removal of remaining suspended solids).
WATER QUALITY IMPROVEMENTS
Data provided by DWNR’s fixed station monitoring and by HMC’s
own instream monitoring network showed dramatic water quality
improvements following the implementation of controls.
EPA
June 1987
Water Quality Progress Report
Suspended Solids. During the two years prior to completion of the
-------
tailings pond, total suspended solids in Whitewood Creek just
below the mine’s outfall averaged approximately 62,000 mgfL Dur-
ing 1978 and 1979, after the pond was in use, monitoring at the
same station showed suspended solids levels of approximately 120
mgIL These concentrations were reduced even further during the
1980s by using sand filtration; in 1985, the average instream con-
centration at this station was 9 mg/L.
Cyanide. Instream cyanide levels (measured as total cyanide) fell
sharply in 1984 and 1985 following the start-up of HMC’s RBC treat-
ment system. Table 1 shows the annual average cyanide levels
measured at two sites downstream of the outfall. During the same
period, background cyanide concentrations (measured at stations
upstream) remained nearly constant at 0.01 mgIL
TABLE 1. Annual Average Cyanide Concentrations
Below the Homestake Outfall (mgIL)
Station
1981
1982
1983
1984
1985
1986
CS28GRa
WQM85b
0.87
0.25
2.02
0.20
1.33
0.13
0.18
0.06
035
0.08
a Monthly samples taken by HMC just below the outfall.
bQuarterly samples taken by DWNR 4 miles below the outfall.
Copper. Copper as well as other heavy metals were present in the
HMC wastewater, and these also have been reduced through the bio-
logical treatment process. The bacteria absorb significant quanti
ties of the metals during their lifetime, after which the organic
residue is recycled to the tailings pond. Table 2 summarizes annual
average copper concentrations upstream and at two stations below
the mine discharge. At present, South Dakota has not developed
a water quality standard for copper; however, it should be noted that
the waters in this region are extremely hard (Whitewood Creek
ranges from 500 to 800 mg/L hardness) so the relative biological
availability of total copper is low.
TABLE 2. Annual Average Copper Concentrations
Above and Below the Homestake Outfall ( ig/L)
St.ation
1981
1982
1983
1984
1985
1986
WQM86a
SC35BSb
2.4
—
4.0
300
<10
227
75
86.9
6.0
21.0
<13
28.5
WQM85C
414
120
94
55.
33.
38.
a Background concentrations measured upstream of HMC outfall;
quarterly samples by DWNR.
bsampling station located I mile below HMC outfall; monthly samples
by Homestake Mining Co.
cSampling stat,on located 4 miles below HMC outfall; quarterly samples
by DWNR.
BIOMONITORING
Under a cooperative arrangement, South Dakota DWNR and EPA
Region VIII have conducted surveys of ambient water toxicity in the
area surrounding Homestake’s discharge. Using an easily cultured
macroinvertebrate (Ceriodaphnia sp.), tests were run in July 1983,
and again in November 1984, several months after the new treat-
ment plant had come on-line.
In 1983, two series of tests were conducted. Initial screening tests
used undiluted water taken from six sites—three upstream and
three downstream of the mine discharge. The results of these tests
were unequivocal: organisms placed in the upstream water were
unaffected, while all downstream samples resulted in 100 percent
mortality. Dilution tests were then carried out using samples from
the three downstream sites. These tests resulted in the following
estimated toxicities expressed as a 48-hour LC 50 (the concentra-
tion of sample that is lethal to 50 percent of the test organisms over
48 hours): at the point of discharge, the LC 50 was less than 10 per.
cent (i.e, organisms could not survive in a mixture containing 10
percent streamwater and 90 percent upstream dilution water);
downstream from the discharge, just after mixing, the LC 50 was 25
percent: and 100 yards further downstream the LC 50 was 65
percent.
During post-treatment bioassays (November 1984), more sophisti-
cated studies were undertaken to test chronic and reproductive
effects below the outfall. Figure 1 illustrates the dramatic recovery
below HMC’s outfall. While in 1983, Ceriodaphnia could not survive
in samples taken downstream of the outfall, by 1984, over 70 per-
cent of the organisms survived in pure mine effluent—station
3—although they could not reproduce. At stations 4 and 5, just
downstream from the discharge, they lived and reproduced as well
as organisms tested in water from upstream, stations 1 and 2.
Source: U.S. EPA Region VIII
Upstream Downstream
Discharge
I I I
Figure 1. Cumulative mortality (left) and reproductive capa-
bilities (right) of Ceriodaphnia at stations upstream and down-
stream of the Homestake discharge (station 3). November 1984.
Additional evidence of water quality progress has been provided
by the return of trout to Whitewood Creek below the Homestake
mine. Although fish were present in waters upstream of the dis-
charge prior to waste treatment, they would not pass below the out-
fall. Within a month after the full treatment process was underway
however, biologists documented the presence of wild brook trout
below the outfall. By early spring of 1985. when DGFP added
Whitewood Creek to its regular stocking program. the numbers of
brook and brown trout were increasing steadily.
FUTURE STUDIES
Future monitoring studies in Whitewood Creek will be concerned
with elucidating the relative toxicities and degradation rates for the
different components of total cyanide. While it appears that free
cyanide is the most toxic component, this is also the one that Pseu-
domonas mudlock prefers. However, other components (e.g., iron
cyanide) contained in the Homestake effluent may be broken down
instream to release weak-acid dissociable cyanide, which is also tox-
ic. For this reason. Homestake’s revised NPDES limits (to be deter-
mined in November 1987) may be expressed in terms of both total
and weak-acid dissociable cyanide. These limits will be based, in
part. on extensive chronic toxicity tests conducted onsite by
Homestake’s biologists using fish and invertebrates found in the
Whitewood Creek Basin.
Material [ or this report was furnished by Joe Bower, South Dakota
DWtIR, Office of Water Quality; Fred Fox, Environmental Director,
and Ronald Waterland, thvironmental TecMician, Homestake Mining
Company; and Del Nimmo, LLS. EPA Region WI!.
This report is produced by EPA to document progress achieved in
improving water quality. Contributions of information (or similar reports
are invited. Please contact E. F Drabkow ki. EPA. MD.SD. WH-553.
401 M Street S.W. Washington, D.C. 20460(202)382-7056.
Downstream
.r
0
e
50
40 -
30 -
20 -
10
Upstream
Discharge
Stations
.
25
w
E
‘ 3, 20
0
z
15
I ’
.
.
.
5
1 2 3
Stations
4 5
-------
United States
Environmental
Protection Agency
Assessment and Watershed Protection Division April 991
Office of \V t r
Washington. DC 20460
Water Quality Progress Report
ww w
Duwamish
River,
Washington
The Duwamish River
is vital to Washingtons
commerce as a primary navigational route, a major contributor to
the States salmon and steelhead trout industry, and a potentially
attractive recreational resource. It is influenced by tidal action over
its lower 10 miles and is the principal source of fresh water to Elliott
Bay. near Seattle. The lower 6 miles of the once meandering river
is now a straightened navigational channel (the Duwamish Water-
way) that flows through a heavily industrialized area of Seattle
where airplane factories, shipyards, metal scrap yards, oil tank
farms, and port facilities are located (Figure 1).
SHIPYARC I
SW LANDER ST EE
FORMER SMELTER
SHIPYARD 2
Figure 1. The lower Duwamish Waterway.
Over the years. municipal and industrial wastes were dumped indis-
criminately into the river, affecting water quality in the river and
Elliott Bay. Improvements in conventional water quality parameters
were observed in the Duwamish in the late 1960s and early 1970s,
but metals and organic toxicants continued to flow into the river
from urban point sources.
Today. the Duwamish River is cleaner than it has been in decades.
This report describes water quality improvements that have
resulted from use of an innovative approach to identify toxic chem-
ical contamination from urban nonpoint sources in the lower
Duwamish watershed.
ASSESSING THE PROBLEM
When trace metal concentrations in the waterway were first meas-
ured in 1971, researchers found lead concentrations that were some
of the highest recorded in western Washington waters. High con-
centrations of other metals were also identified, and there was con-
cern that these metals put migratory fish, birds, mammals, and
other estuarine organisms at risk. Researchers first assumed that
the metals present in the water were contributed primarily b the
Renton Treatment Plant (RTP) effluents and by untreated sewage
mixed with stormwater runoff discharged by combined sewer over-
flows. However, in 1979 it wasdemonstrated that RTP effluent was
not a major source of metals present in the river. Other known point
sources of metals were industrial wastewater discharges permitted
under the National Pollutant Discharge Elimination System
(NPDES) program administered by the Washington Department
of Ecology (Ecology). However, these permitted discharges i .ere
limited to stormwater runoff and noncontact cooling water. Known
metal loadings from these point sources could not account for the
high metal concentrations observed in the river.
Thus. it appeared that nonpoint sources continued to degrade the
rivers water quality. Because of the rivers commercai and
aesthetic value. EPA and Ecology designated the Duwamish a high.
priority area for study and action. A Clean Water Act Section 208
grant was awarded to the Municipality of Metropolitan Seattle
(Metro) in 1979 to inventory pollutants entering the ri er and to
develop an abatement program.
WATER QUALITY AND SEDIMENT MONITORING STUDIES
In the early 1980s. water quality studies conducted during ow flow
months (July-September) revealed that copper concentratons
exceeded the EPA acute freshwater criterion( 18 gL) and lead con-
centrations exceeded the EPA chronic freshwater criterion (3.2
gJL). The highest concentrations of metals (e.g.. lead. copper.
arsenic, zinc, mercury, and cadmium) and other contaminants were
found unevenly distributed in the sediments of the Duwamisn
Waterway, which suggested that contaminants from locahz c
sources were contributing to the formation of ‘hot spots: Sec
rnents near a storm drain outfall in the West Waterway Figure Ii
were found to contain 18.000 ppm lead. Lead concentrations
decreased away from the storm drain outfall. At another site off r ar
bor Island. copper and zinc concentrations were as high as 3.000
ppm in marine sediments (Figure 1).
Metro continued investigating industrial discharges along the
waterway to identify chemical use and disposal at facilities that
might contribute contaminants to storrnwater runoff. EPA and Ecol.
ogy assisted Metro with regulatory actions leg., revised permt
requirements. best management practices (BMPs). finesi at fact
ties with pollution problems.
METRO’S STORM DRAIN ANALYSIS
In 1984, Metro received a Clean Water Act Section 205(j) grant.
which was used to conduct an innovative sediment sampling pro
gram in storm drain systems. Storm drain sediment sampling is
easier and more efficient than direct monitoring of stormwater dis
charges. for these reasons. First, sediments in storm drains
represent an accumulation of contaminants over time. whereas a
stormwater sample represents contaminant input to a system onI
at a given point in time, Second. toxicants are easier to detect in
storm drain sediments than in stormwater samples. Third. sedi
ment sampling efforts do not need to be coordinated with rainfall
events.
Metro collected sediment samples from the low energy sec ons
(i.e.. manholes) of 12 municipal storm drain systems that s
charged to the Duwamish Waterway. These samples were anal’ zed
for metals, organic compounds, and conventional contaminants.
EPA
Elliott Bay
SW FLORiDA STR
18.000 m
•10
AEN1ON TREATMENT
PLANT OUTFALL
-------
Sediments within four drainage systems contained substantially
higher concentrations of contaminants than those found in street
dust in the area. Additional sediment samples were collected and
pc 1lutants were tracked up through the drain systems until, by
process of elimination, the sources were identified.
At a storm drain on Southwest Lander Street. a major source of lead
to the river was discovered (Figure 1). Storm drain sediment
samples containing 350.000 ppm lead, or 35 percent lead, were
found in a storm drain adjacent to a former smelter that had recov
ered lead from used batteries. Metro. Ecology. EPA. and the City
of Seattle removed over 20 cubic yards of contaminated sediments
from the line and sent them to a recycler
In a storm drain on Southwest Florida Street. which empties into
the West Waterway. high concentrations of creosote, pentachloro
phenol. copper, and arsenic were traced to a wood.treatment facility
(Figure 1). The company later pleaded guilty to charges of illegal
dumping. High concentrations of PCBs in the same storm drain sys•
tern were traced to a scrap yard that recycled old PCB.contairiing
electrical transformers. Contaminated storm drain sediments were
removed and shipped to a hazardous waste disposal facility
SUCCESS OF THE STORM DRAIN MONITORING APPROACH
Cleanups of sediments in storm drain systems and reductions in
contaminant inputs from industrial facilities eliminated major
sources of contamination to the Duwamish Waterway Research
ers found lower contaminant concentrations in subsequent storm
drain sediment sampling efforts. In 1989. sediments in the South’
west Lander Street line contained 10.000.50,000 ppm lead, which
is substantially lower than the 350.000 ppm lead found in 1984
(cleanup at the former lead smelter is still in progress)
One can assume that the removal of contaminated sediments from
a storm drain system would reduce contaminant input from the sys
tern to the Duwamish Waterway and Elliott Bay However, it is
difficult to monitor nonpoint sources of pollutants and thus eval•
uate the effectiveness of pollution abatement programs.
To evaluate the success of this cleanup program, net loadings of
metals (lead, copper, and zinc) from the Duwamish Waterway to
Elliott Bay were quantified using data collected by NOAA from
1980.1986 Dissolved metal concentrations, salinity, and flow were
measured in the water column near the mouth of West Waterway,
where the majority of the river flow is discharged. and throughout
Elliott Bay. NOAA calculated net loadings of trace metals using a
water transport model Researchers knew that the net loading of
metals to the bay included metals discharged directly from indus’
trial sources and from the RTP outfall, plus metals naturally present
in the Duwamish River Therefore, to determine net loadings of
metals from industrial sources to the river (Figure 2). NOAA subS
traded known metal loadings from the RTP and natural metal load
ings in the river from the calculated total loadings.
NOAA found that metal concentrations measured in the water
column at the mouth of the West Waterway were far less than those
found in early 1970 studies Dissolved lead concentrations at the
mouth of the West Waterway were 0047 czg/L in 1985 and 0041
g/L in 1986. which is dramatically less than lead concentrations
of 54 zg/L found in 1971 Lead reductions may be attributable to
several factors, Between 1981 and 1985. the lead smelter ceased
operations and initial remedial actions were completed at the site
The decreased dissolved lead concentrations in the waterway fol.
lowing these actions, and the calculated net loadings of lead from
industrial sources suggest that these initial control measures were
effective in reducing dissolved lead discharges from industrial
sources.
FIgure 2. Industrial metal loadings from the Duwamish
River to Elliott Bay (NOAA).
Data on copper suggest that voluntary BMPs uncertaken by two
shipyards prior to January 1986 were not totally effective in reduc
ing net loadings of dissolved copper to Elliott Bay Between Janu-
ary and August 1986. one shipyard closed for economic reasons
Activities at a second shipyard were reduced substantially. and man
datory BMPs were instituted.
In contrast to changes in lead and copper loadings, little change
in the net loadings of dissolved zinc from industrial sources was
observed between the August 1981 and January 1986 samples
These data suggest that voluntary and mandatory BMPs at the ship
yards were ineffective in reducing net loadings of dissolved zinc
However, activities at the shipyards were decreasing between Jan
uary 1986 and August 1986. and loadings of dissolved zinc from
industrial sources to Elliott Bay were reduced by 92 percent from
194 lb/day to 15 lb/day Alternatively, investigative work conducted
by Ecology, which resulted in the implementation of BMPs at other
contaminant sources (e.g. salvage yards), may have resulted in
reduced loadings of zinc to the waterway.
FUTURE ACTIVITIES
Local agencies are continuing to explore options to further reduce
contaminant inputs to the Duwamish River Although water qual
ity has improved in the river, bottom sediments in the river and bay
have been contaminated with toxic substances for many years
Agencies are currently identifying potential remedial alternatives
(eg., dredging, detoxification, n’place capping) for the contam
mated sediments. in the near future, the Washington Urban Storm
Water Management Program will be implemented to reduce con
taminant inputs. One of its key provisions will be to issue NPDES
permits to cities for urban stormwater discharges.
Matenals for this report were furnished by Tom Hubbard. Municipal
ity of Metropolitan Sea We (Metro), Tony Paulson, NOAA Pacific
Marine Environmental Laboratory: John Armstrong. EPA Region X
Office of Puget Sound, and Lee Dongan and Dan Caigmll. Washing•
ton Department ol Ecology.
This report is produced by EPA to document progress achieved in
improving water quality Contributions oIinformat,on for similar reports
are ,nuited Ptease contact EPA. AWPD. Monitoring Branch, WH 553.
401 M Street SW, Washington. DC 20460 (202)382 7056
Land Use Practices
250
Siupyazd I Orydocus E npiy
Sh’ yard 2 ClOied
Lied Srnaii.q Cleanup
200 BP 4Ps— 1
I
ii
m l
gi
. ‘15O
‘ I
\ I
•—“—‘Copcer I
——SLea i I
—Zr.C I
100
I
0
Apr Au Apr Jan AuQ
80 81 85 86 8C
-------
EPA
United States
Environmental
Protection Agency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
April 1986
Water QuaHty.Pi grarn Highlights
- -- w --w
New York State’s Waéteload Ailocation Procedure
Section 303(d) of the Clean Water Act requires States to identify as
water quality limited: “those waters. . . for which effluent limita-
tions. . . are not stringent enough to implement any water quality
standard applicable to such waters.” For each water so identified,
States are required to establish “the total maximum daily load for
those pollutants that the Administrator identifies under section
304(a)(2) as suitable for such calculation.” Using the total maximum
daily load (TMDL), States assign individual wasteload allocations
(WLA) for point sources along water quality limited segments.
NEW YORK’S WATER QUALITY PERMIT PROGRAM
The New York State Department of Environmental Conservation
(NYSDEC) DMsion of Water administers approximately 30O NPDES
permits for surface water dischargers. Of these, approximately 1,600
are classified as “significant” dischargers; this group includes all pub-
licly owned treatment worl (POTWs), industrial plants, and any other
discharge that contains toxicpollutarlts. All significant permits are
processed by permit writers and water quality analysts at the State
offices in Albany. The remaining 1,700 permits are predominantly con-
cerned with conventional pollutant limits, and these are processed
and issued directly by one of the nine Regional Offices around the
State.
The central feature of New York’s water quality-based permit program
is the list of Ambient Water Quality Standards and Guidance Values,
which serves as a basic resource in the State’s regulatory and permit-
writing activities. The list contains quantitative water quality criteria
(based on either aquatic life or human health protection) for approx-
imately 210 toxic and nonconventional pollutants. At the present time,
criteria values for about half of these substances have been codified
as regulatory standards. The remaining criteria termed “water qual-
ity guidance values,” are used pending completion of the
administrative and review process specified for standards develop-
ment; once this process is complete, all guidance values will be
adopted as water quality standards. As new substances of concern
undergo sufficient reviev additional guidance values are developed.
In addition to the list of ambient standards and guidance values, the
State has also developed “threshold criteria” for screening effluent
concentrations of toxic and nonconventional pollutants. Existing
threshold criteria for toxics are 1.0 lb/day for total (or total recoverable)
metals, total cyanide, total phenols, volatiles, and acid and base!
neutral compounds; and 0.001 mg/L for pesticides. Threshold criteria
for nonconventional pollutants cover cyanide amenable to chlorina-
tion, sulfide, total residual chlorine, fluorides, and other “Substances
of Concern.”
Where existing or expected discharge levels exceed the threshold
criteria, then a BAT/BPJ determination must be made and a
technology-based limit established for that pollutant. (This limit may
be superseded if the pollutant is found to be water quality limiting.)
Where a threshold pollutant is present in an effluent, but does not
exceed the criterion (i.e., does not warrant additional technological
controls), then an “action level,” along with specific monitoring
requirements, is included in the permit. The action level is not a dis-
charge limitation, but rather a numerical reporting level developed
using a specified methodology to obtain a multiple of the reported
concentration. If the action level is exceeded, the permittee is required
to undertake a more intensive monitoring effort; where action levels
are consistently exceeded, the permit may be reopened to specify
either an effluent limitation or a revised action level.
Thus, for any effluent that contains toxic or nOncOnventional pollu-’
tants, a permit is developed using action levels, technology-based
limits; or water qu Tn -based limits.
WASTELOAD ALLOCATION PROCEDURES IN NEW YORK
Oxygen-Demanding Substances. For oxygen-demanding sub-
stances, TMDLsIWLAs are calculated using mathematical water qual-
ity (dissolved oxygen) models. These models apply to specific
waterbody segments, and whenever possible, they are calibrated and
verified using site-specific survey data. Where appropriate, dissolved
oxygen models consider the following factors: reaeration; photo-
synthesis; aquatic plant respiration; biological oxygen demand; car-
bonaceous/nitrogenous oxygen demand; sediment oxygen demand;
and advection and diffusion.
For oxygen-demanding substances, the TMDL is essentially equiva-
lent to the waste assimilative capacity of the water during a critical
period. Assimilative capacity analyses are conducted using the mini-
mum average 7-day flow with a recurrence interval of 10 years (7010
!Ow flow). Unless temperature records at the design flow are avail-
able, models assume a critical temperature of 25 °C for non-trout
waters’ and 24°C for trout waters. Complete mixing is assumed for
discharges to riverine systems.
Where the waste assimilative capacity is exceeded, and minimum
technology-based treatment has been implemented for all discharg-
era, then the segment is classified as water quality limited. Wasteload
allocations are assigned using the following principles:
• Reductions are required of the discharger(s) that most directly and
significantly affect the dissolved oxygen violation.
• For multiple discharges, allocations are proportioned according
to the relative contribution to the oxygen deficit at the point of vio-
lation. Consideration is given to costs and the degree of treatment
required by each discharger.
• New discharges that cause a segment to become water quality
limited are required to bear the full burden of maintaining water
quality standards (provided other dischargers meet minimum
technology-based treatment levels.)
• Water quality-based allocations are generally written as seasonal
limits for the period June 1 through October31, unless conditions
indicate that a different time period would be appropriate.
• Effluent limits may be expressed as 7-day average and/or 30-day
average for conventional pollutants depending on the sensitivity
of the waste assimilative capacity analysis.
Toxic Substances. For toxics and nonconventional pollutants, the
maximum allowable load for each water pollutant is determined for
an entire river basin. The State has 17 river basins, and an average
basin might contain 10 pollutants that are water quality limiting (some
basins may have 20, others only 2 or 3). TMDLsIWLP s are developed
-------
by assuming that all toxic substances behave conservatively (ía ,that
there are no losses due to degradation 1 setlling, adsorption, orvolatil-
ization)
To provide some balance to the assumption of no instream losses,
nonpoint source contnbulions and background levels in the waler are
assumed to be zero, unless river basin or stream-specific data are
available In most cases, downstream concentrations are calculated
using a standard mass balance equation, sometimes referred to as
the point-of-discharge-dilution model In a few cases where site-
specific data are available, nonconservative modeling techniques
may be used.
Figure 1 illustrates New York’s TMDLJWLA process for substances
other than oxygen-demanding wastes The procedure begins with
the State’s tech nology determination or action levels, which are specs
fied as loading rates in lb/day Minimum technology-based treatment
must be specified (aA,T, secondary treatment, BPJ) If threshold values
are not exceeded, toxics are still included in permits as action lev-
els Where threshold values are oroeeded and the substance appears
on the State’s list of standards and guidance values, then the analyst
goes on to determine whether the substance is water quality limit-
ing For substances that are not listed, minimum technology-based
limits are assigned, and the State may require effluent toxicity testing
To support its basin-wide toxics allocations, NYSDEC has developed
an inventory of the toxics discharged in each c i the State’s 17 drainage
basins Sources of information for the inventory include current and
proposed permits, pretreatment analyses, discharge monitoring
reports, compliance sampling, and industrial chemical surveys The
computerized inventory lists each discharger with its associated
parameters and effluent concentrations Then, when a new discharge
permit or a permit modification is received, water quality staff can
quickly assess the changes in total loading for a given pollutant
Each toxic pollutant is regulated according to its concentration in a
“critical segment” The cntical segment is defined as that portion of
the basin where water quality classification or standards are most
stnngent Allowable pollutant load in this segment is calculated using
the 7010 flow for standards that protect tor aquatic life, and the 30010
flow for standards that protect for human health Where standards
exist for both human health and aquatic life, the more stringent is
applied
Using the appropriate water quality standard (or guidance value) and
streamftow in the cntical segment, the analyst calculates the max-
imum allowable load (for each pollutant) for that segment. Then a
“baseline load” is calculated for each substance The baseline load
calculated for the critical segment reflects expected loading follow-
ing the adoption of technology-based treatment by all industnes and
municipalities Where technology-based limits are not specified for
a particular substance, the “action-level” concentration is used to cal-
culate baseline load
Companng the baseline load with the maximum allowable load
results in a determination of whether the basin or stream segment
is water quality limited or effluent limiled for each pollutant Where
the waterbody is effluent limited, discharge limits are developed using
technology-based requirements For each substance for which the
basin is water quality limited, wasleload allocations are established
as follows
• At the critical water quality limiting segment, the baseline load is
compared to the maximum allowable toad The amount by which
the baseline load exceeds the allowable load is the “excess load”
• The excess load is allocated among dischargers in the ratio of a
discharger’s baseline load to the total baseline load
• Individual effluent limits are set as the difference between the dis-
charger’s basel ine load and their proportion of the excess load
• The water quality-based effluent limit for a specific substance is
expressed as a maximum daily value when the cntenon is based
on 7010 flow The water quality-based effluent limit for a specific
substance is expressed as a 30-day average value when the cnte-
non is based on 30010 flow
• For certain waterbodies, naturally occurring or man-made condi-
tions may preclude the technical development and defensibility
of a water quality-based effluent limit for a specific toxic substance
In these situations, technology-based limits and effluent toxicity
testing are considered as a permit requirement in lieu of a water
quality-based limit
• Where a standard or cntenon value does not exist fora substance,
technology-based limits along with effluent toxicity testing are again
considered
PROGRAM PROGRESS
Dunng the fiscal year October 1984 through September 1985, New
York State reviewed about 1,000 discharge permits Of these,
NYSDEC’s Central Office in Albany reviewed 270 significant permits
to check permit limits against TMDL cntena The remaining permits
concerned conventional limits and were reviewed by the State’s
Regional Offices In FY 86, the State expects to calculate or check
approximately 2,500 wastetoad allocations for both toxic and nontoxic
pollutants This number is based on an average of 250 permits per
year that contain water quality limited parameters multiplied by an
average ot 10 water quality parameters for each permit.
Matenalfor this report was turnished by Albert W Brnmbetg, Chief,
Water Qualify Evaluation Section, N YSDEC and Robert Vaughn, US
EPA, Region!! Technicaland Operational Guidance Series (TOGS)
memorandums issuedtyjNYSflEC, 50 Waif Road, A(banx NY 12233
contain the standards and guidance values as well as specific
guidance concerning procedures discussed in this report
I
This report is produced by EPA to lnghiightmon,tonng and waslaload
allocation activities Contnbutions of inform ation for similar reports we
invited Please contact E F Drabkowsk,, EPA, MDSO, WH-ssa 401 M
Street SW, Washington, DC 20460(202) 382-7056
FIgure 1. TMDL/WLA process for toxics in New York State.
-------
EPA
United States
Environmental
Protection Agency
Water Quality Program Highlights
The Delaware River Cooperative Monitoring Program
The Delaware River forms the boundary between Pennsylvania, New
York, and New Jersey, and water quality management in the area is
the shared responsibility of a dozen or more State, local, and federal
environmental and public health agencies. In addition to the States,
the Delaware River Basin Commission (DRBC), a federal-interstate
agency, has junsdiction throughout the river basin, and the Nation-
al Park Service (NPS) operates within the boundaries of two scenic
river aroas.
In 19 4, the DRBC and the NPS initiated a
cooperative water monitoring program focus-
ing on the upper and middle sections of the
Delaware River These two sections include
the Upper Delaware Scenic and Recreation-
al River (UDSRR) and the Delaware Water
Gap Natk)nal Recreation Area (DWGNRA ,
both of which are part of the National Wild
and Scenic Rivers system. Together these
two reaches, administered by the NPS, span
nearly 120 river miles in the three States and
provide fishing, boating, and swimming
recreation for over 300fl )0 visitors annually
Prior to initiation of the monitoring program, water quality in both the
Delaware and in most tributaries was thought to be good due to the
rural nature of the area. Hc vever this assumption was based on scat-
tered, infrequent monitoring or on occasional special-purpose water
quality studies. The three States maint2ined only four routine (fixed)
monitoring stations along the upper and middle Delaware River, and
only 6 out of 75 tnbutanas were m-’iitored. Because of this, it was
determined that a coordinated mo- ering prognm was necessar’
to provide a consistent flow of water a’ ‘ality data.
DEVELOPING THE MONITORING PROGRAM
To make the best use of oxisting resources and skills, a water quality
“screening” approach was developed with monitoring responsibil-
ities shared between the DRBC and the NPS. The program uses a
limited number of parameters, but relies on frequent sampling at
numerous locations, and follow-up surveys (where necessary) to pro-
vide the greatest amount of useful information at the lowest cost.
The cooperative monitoring program was designed to expand
monitoring activities in three ways:
• Sampling was to be extensive (rather than intensive), covering,
to the extent possible, numerous Delaware River locations and
all skinfficant tributaries dunng the 3-month recreational season.
• Sampling was to be frequent (biweekly for many locations)
because water quality can vary rapidly over time.
• Data turnaround was to be rapid so the program could respond
quickly to potential public health probfr’ms.
MONITORING PARAMETERS
The parameters selected for screening water quality were fecal coil-
forms and fecal streptococci, water temperaturc, dissolved oxygen,
pH, conductivity, and species diversity (rnacroinvtrtebrates). Rain-
fall information is readily available from other agencies and was not
collected by the proqram itself. Stream flow measurements are being
phased in over time.
These eight parameters yield additional data when they are used
together The ratio of the two fecal bacterial parameters serves as an
indicator of possible human sewage contamination. Water temper-
ature and dissolved oxygen data are used to calculate the percent
dissolved oxygen saturation. Saturation information, in conjunction
with pH data, is used to assess aquatic plant activity. Finally, infor-
mation obtained from macroinvertebrate analyses is converted to one
or more biological indices that serve as a surrogate for measuring
the presence of toxics.
Biological monitoring focuses on benthic macroinvertebrates. All river
samples and several tributary samples are analyzed taxonomically;
a computer program is then used to calculate the Species Diversity
Index value and the equitability number for each sample. For other
tributaries, the Sequential Comparison Index, a technique that is rela-
tively simple, fast, and usable by nonbiologists, is applied.
LOCATION AND FREQUENCY OF SAMPLING
In 1984, the cooperative monitoring effort was implemented on a tria’
basis in the DWGNRA. This approach proved to be highly success-
ful, and in 1985, a decision was made to expand the program into
the UDSAR. During May through September 1985, a total of 88
locations were sampled (24 river stations, 64 tributaries). In almost
all cases, tributary sampling stations were located near confluences
with the Delaware River.
With the exception of three designated swimming areas, which were
sampled weekly, nearly all stations were sampled every other week.
During the 1985 sampling season, 550 station-visits were carried out
by DRBC and NPS personnel. In addition, seven follov:-uo surveys
were performed in response to data fin 1 ngs. Follow-up survey . v.hich
generally involve more intensive sampling in a limited area, have been
used to locate specific sources of agricultural and urban runoff and
natural causes of water quality problems as well as point sr•urc’
problems.
USE OF PARAMETERS TO SCREEN WATER QUALITY
The primary purpose of the screening program is to determine the
need for water quality management actions. These actions can be
short-term (e.g., closing a bathing beach or initiating an intensive sur-
vey) or long-term (e.g., establishing pollution control priorities).
Although questions have been raised about their public health
implications, fecal coliform (FC) and fecal streptococcus (FS) remain
excellent indicators for screening pollution in a watershed. For exam-
ple, completely undeveloped watersheds show bacteria levels that
are very low. In other watersheds with varying levels of development,
streams have correspondingly higher bacteria levels and FC/FS
ratios. In this way mean fecal coliforrn ared F /FS values tend to
“fingerprint” the level of development in a watershed. By sampling
frequently for these parameters, it is possible to compare similar
watersheds to screc n fr’ r potential water quality problems. By con-
trast, monthly, quarterly, r even lflt f ve surveys rarely “catch’ ih
almost random nature ef bacterisl fluctuations or the 1ingerprin ”
of typical watersheds.
Frequent data colV lion under varying dry and w- t conditions also
serves to pinpoint rroh rns. or e mple, two impaired trft :;’3ries
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
July 1986
Location of Study Area
-------
in the study area are Brodhead and Cherry Creeks Brod head Creek
receives discharges from two overloaded sewage treatment plants,
a combined storm and sanitary sewer system, and several industnes
Its watershed also contains large second home and resort develop-
merits Data from Brodhead Creek show that bactenal levels and
FCIFS ratios increase dramatically following rainfall, indicating that
these pollutants may originate from nonpoint sources and/or bypass-
es of untreated sewage. Conversely Cherry Creek’s bacterial levels
and FCIFS ratios increase during low flows, indicating a chronic
problem caused by raw sewage discharges (there is one municipal.
ity located in this watershed).
Based on DRBC-NPS focal bactenal data, wastewater treatment
improvement projects have been accelerated in both the Cherry and
Brodhead Creek watersheds. The DRBC-NPS cooperative monitor-
ing program continues to alert State and local jurisdictions when
problems arise, particularly problems resulting from point source dis-
charges that may require intensive analyses including effluent sam-
pling and toxicity testing The New York State Department of
Environmental Conservation, for mple, is planning a survey of Cal-
licoon Creek in 1986 as a result of screening data that showed higher-
than-expected levels of bacteria.
Dissolved oxygen measurements are most valuable when used in
conjunction with water temperature to calculate percent saturation,
which is generally a measure of aquatic plant activity (photosynthesis)
This is vcrufied by corollary increases in pH values. During 1985,
drought management activities led to drastic reductions in reservoir
releases and resulting flows downstream. Monitoring data demon-
strated that a major impact of the altered reservoir operations was
increased aquatic plant activity and violations of the pH standard
(above pH 9) during daylight hours in response to these observa-
tions, measurements were conducted at night to determine if
increased plant respiration dunng the night was causing violations
of dissolved oxygen standards. The data showed large drops in dis-
solved oxygen levels, but standards were not violated.
Conductivity is another parameter thnt appears to “fingerprint” the
a - ount of development in a watershed Generally conductivity values
increase with the degree of human activity in a watershed Monitor-
ing in one DWGNRA stream, White Btok (a mostly undeveloped
watershed), showed consistently high conductivity compared to
streams with similar levels of development and physical charac-
teristics It was speculated that a landfill might have been located in
the watershed during cleanng operations for a recreation area in the
early 1970’s A follow-up investigation was completed that determined
that the cause of the high conductivity was naturai—a small limestone
formation traversed the basin in the headwaters Still, the ability to
identify potential water quality problems and to determine their origins
is evident.
DETERMINATION OF MONITORING PRIORITIES
An important initial objective of the cooperative monitoring program
was to collect baseline water quality for as many tnbutaries and nver
locations as possible. To accomplish this, many stations were sam-
pled five or more times dunng the 1984 and 1985 sampling
seasons—a level that could not be sustained over the long run given
available resources To define a more efficient monitonng strategy
for 1986 the DRBC developed a method for determining reach-by-
reach pnorities.
Figure 1. Cntena for rating monitoring priorities on
designated segments of the Delaware River.
subject of intensive surveys, and macroinvertebrate studies will be
conducted in two locations in each segment Four segments scored
between 2 and 4 and are considered second pnonty; these will be
sampled biweekly to monthly, with one location subject to biomonitor-
ing. Low-pnonty segmonts (with scores below 2) are to be sampled
only once, with other monitoring to be conducted as time and
resources permit.
PROGRAM ADMINISTRATION
While the cooperative monitonng program is administered without
the use of formal interagency agreements, the DRBC and NPS staff
work together closely and solicit assistance from other governmen-
tal agencies as necessary Although most of the sampling and anal-
ysis is camed out by the two pnncipal agencies, State and county
officaaJs are involved in program planning, along with representatives
from EPA Regional Offices and the US Geological Survey
When monitonng results indicate an immediate problem, follow-up
actions are initiated by State or local agencies. All monitoring data
are entered into EPA’S water quality data system, STORET Al the end
of each sampling season, a report of findings and recommendations
is prepared and distributed to local government agencies, the pub-
lic, and the press.
Dunng the 1986 sampling season, monitoring staff will study river
foam found at various locations along the Delaware Although pos-
sibly natural in ongin, foam currently is perceived as a pollution
problem in the region Longer-range objectives for the program
include establishment of a scenic rivers research center with a water
laboratory and the institution of a land-use monitoring system, pos-
sibly using high-altitude and remote-sensing data.
Material for this report was furnished by Richard C Albert of the Ce/a-
ware River Basin Commission, John Kansh of the National Park Ser-
vice, and Charles Sapp, U.S. EPA Region Ill
The method selected divides the 120-mile study area into 10 seg-
ments, varying from 8 to 20 miles in length The 10 segments were
determined by evaluating hydrology land use, extent of urbanization,
known or suspected problems, and access points along the river.
Then, the segments were ranked in order of prionty using the rating
system shown in Figure 1
Based on the distnbution of the scores, segments sconng greater than
4 are considered highest pnority. Three segments qualified as high
priority segments and, along with significant tnbutaries, will be the
This report is produced by EPA to highlight monitoring and wasteload
allocation activities Contributions of information for similar reports are
invited Please contact F F Drabkcwsk,, EPA. MDSD, WH-553,
401 M Street SW, Washington, DC 20460(202) 382-705&
A WATER QUALITY IMPACTS
• Potentiai for problems (1 point)
• Problems known (2 points)
• Possibie problem warrants investigation (3 points)
B DEGREE OF URBANIZATION
• No significant urbanization (no points)
• Small town type of urbanization (1 point)
• Significant urbanization (2 points)
C PUBLIC BATHING OR WATER RECREATiON AREA
• Yes (1 point)
D AVERAGE NUMBER OF TRIBUTARIES PER MILE (fraction)
-------
United States
Environmental
Potection Acjency
Monitonng and Data Support Divis;on
Office of Water
Washington, DC 20460
September 1986
Water Quality Program Highlights
Arkansas’ Ecoregion Program
Traditionally, State water . uaiit1,, standards have been based upon
national criteria developed by the U.S. Environmental Protection
Agency (EPA). However, these criteria may be too general for some
local conditions, and States are encouraged to develop site-specific
criteria and designate more specialized uses where they ate
appropriate for indMdual water bodies. Yet the development c i numer-
ous site-specific standards can be resource intensive.
To resolve this problem, several States aria the EPA are developing
an intermediate method that employs a regional framework for water
quality management decisions. Instead of developing standards
around indMdualized waterbodies. homogeneous ecological regions
or “ecoregions” are identified. These ecoregions integrate charac-
teristics such as climate, land surface form, sods, vegetation, and and
use, to form distinct biological and chemical units.
in 1982, the Water DMsiOn Of A.rkaj-isas’ Department of F llution Con-
trol and Ecology (DPC&E) began a 5-year project to test the aquatic
ecoregions concept as a basis for reevaluating stream classifications.
The State is examining the physical, chemical, and biological charac-
teristics of carefully selected streams in each of six ecoregions within
Arkansas. As the project nears completion, DPC&E staff are con-
vinced that the concept is sound and that the resuhs will be directly
applicable during the 1987 triennial review Of State water quality stand-
ards. The State has already successfully employed ecoregion data
in the development of use attainability analyses.
THE ECOREGION CONCEPT
Ecoregion analyses have been applied for both research and regula-
tory purposes. For example, an ecoregion approach has been used
to identify surface waters that are sensitive to acidic deposition. Other
applications include the definition of aquatic uses in biological terms
and defining impact thresholds associated with EPA’S antidegration
policy. In Arkansas, the primary use of this concept has been to re-
examine State water quality criteria and standards.
To define realistic standards for a disturbed water body, it is neces-
sary to compare the water quality with a control or reference site where
the disturbance is absent. The most commonly used reference sites
are upstream and downstream ef a stressed water segment. By con-
trast, the ecoregion approach uses carefully selected “least-
disturbed” streams within the same ecelogical region as water quality
reference sites.
To carry out an ecoregion analysis, it is necessary to first designate
areas where physiographic and biological characteristics are similai
Then, within these ecologically homogeneous regions (ecoregions),
a series of reference streams are selected corresponding to differ-
ent size drainage basins. Streams of roughly equal size in the same
ecoregion would be expected to have similar water quality and eco-
logical indicators and, thus, respond similarly when subjected to corn-
parable effluents or habitat disruptions.
The advantages of a regional framework for stream classification
are that it provides an objective ecological basis to clef me specifc
subcategories of aquatic uses and it allows such uses to be developed
efficiently, without conducting numerous site-specific surveys for sun-
liar water bodies. Subcategories alk the expansion of general aquat-
ic life expressions—e.g., ‘warmwater fishery,” or “cold-water fish-
ery’! .to reflect diversity in species composition, within a categoiv,
from region to region; this allows States to establish realistic biolog-
ical uses and goals.
DEVELOPING THE ECOREGION PROGRAM
The motivation for undertaking the ecoregion program in Arkansas
was the knowledge that many of the States cleanest streams and
lakes did not meet national water quality standards—not because of
pollution, but because of naturally occurring physical and chemical
conditions. Rather than enforce inappropriate standards, State offi-
cials undertook an ambitious program to assess water quality and
ecological conditions in representative least-disturbed streams. These
least-disturbed streams will be used as “reference streams” to re-
fine use classifications and associated water quality criteria for sim-
ilar streams and rivers around the State.
A workplan and grant request were submitted to fund the study under
provisions of Section 205(j) of the Clean Water Pet. Funding was
approved by the EPA Regional Office in Dallas, bcas; however, before
stream surveys could be initiated, it was necessary to complete four
preliminary tasks.
1. Define Ecoregions The approach used for defining regions and,
uttimately, the specific sites to be investigated, was based on tech-
niques developed by James Omemik, Robert Hughes, and others
at EPA’s Environmental Research Laboratory in Corvallis, Oregon.
This method incorporates consideration of geographic characteris-
tics including land surface form, soils, potential natural vegetation,
and land use. By determining the dominant types ci these four charac-
teristics, then mapping their different combinations, 76 relatively
homogeneous ecoregions were identified in the cotemiinous United
States. Six of these ecoregions are within the State of Arkansas
(Figure 1).
2. Select l4 tershed&za Within each ecoregion, DPC&E designated
three watershed sizes: small watersheds of 20 to 50 square miles;
medium watersheds ci 100 to 200 square miles; and large watersheds
draining 300 to 500 square miles. It was felt that this range would
include the beneficial uses for those State streams for which reclas-
sification would be proposed.
EPA
FIgure 1. Aquatic Ecoregions in Arkansas
a Select Survey Penods. Two sampling periods were selected to cap-
ture critical and seasonal dissolved oxygen (DO) levels: late summer
-------
August or early September) dunng the high-temperature. low-flow
penod when DO levels should be minimal, and dunng spnng when
DO requirements for fish reproduction are cruciaj. The exact timing
for the spnng sampling penod was chosen by monitonng stream tem-
perature to determine appmpnate fish-spawning conditions.
4. Select Sur #ey Sites. Reference streams were selected by revi-
ing the location of known dischai ers and using the field expenence
of DPC&E stafl to eliminate streams with known pollution sources.
.Ail potentiai watersheds were outlined on a map and reviewed b non-
point source pollution Extensive field evaluations of potential sites
were conducted to confirm their suitability and final selection as
representative least-disturbed streams
STREAM SURVEY METHODOLOGY
In 1983-84, DPC&E collected water quality and ecological data from
least-disturbed streams in small watersheds of all six ecoregions.
Medium-size and large watersheds were surveyed dunng 1984-85
and 1985-86 respectively. Each stream survey followed identical
procedures, and the same 1-week werk schedule was canied out to
assess physical, chemicai, and biological parameters. In most cases
t representative streams for each size watershed were sampled
in each ecoregion. The following parameters were measured:
• DLsoA edO en. Dissolved ex en was measured continuously
dunng the 5-day survey period, with DO probes placed in a pool (at
mid-depth) and in a nffia Both temperature and DO data were re-
corded, and a computer was used to calculate the percent satura-
flon of DC as well as the daily maximum, minimum, and average DO
concentration.
• Chemicaa Chemical analyses were performed on three grab sam-
ples taken at least 1 hour apart within an 8-hour penod. Four con-
tainers were used to test for parameters as follows:
— Coliforni bottle—focal cohform
— Filtrate vial—ammonia-nitrogen and ortho-phesphate
— Dark bottle—chlorophyll a
— Light bottIe—turbidit ç total suspended solids, total dissolved
solids, biochemical cygen demand (BOD 5 and 8OD ), total
phosphorus, nitrate +nrtrite-nitmgen, chloride, sulfata, l ã m,
specific conductivity, aJl 1tnrty, hardness, and manganese.
• Physical. Evaluation of physical parameters was an uiiportant pail
of the survey and included both hydrologic measurements and hatX
conditions. Hydrologic parameters assessed for each site included
stream flow and stream velocity, stream gradient, and mean width
and depth. Habitat measures included an assessment of stream sub-
strafe, instream and cano er vegetation, bank stabdhy and npa
ian vegetation.
• Biological. DPC&E conducted biological studies of both benthic
macroinvertebrates and fish populations. Samples were used to tax-
onoimcally chai ei -ize the aquatic oommunit ides dity 1uatb.,r tam,
and determine relative abundances. For both types of bk a, the
Shannon-Wiener diversity index and Indices of nnees, variety and
dominance were calculated to assess overall community health.
SURVEY RESULTS
Results of the srnail watershed surveys reveal the diversity of water
quality conditions among some of the most pristine waters in the State.
Daily minimum summertime DO values tanged from less than 3 mglL
in the Mississippi Afluviai Plain, South Central Plans, and Arkansas
Valley streams (well below the existing water quality standard of
5 mg/L) to over 6 mglL in the Boston Mountain and Ouachita Moun-
tarn Regions.
H ver, even in streams where minimum DO values were less than
3 mgIL, significant numbers of fish species were collected (28 to 31
species) arid all included black bass (Centrarchidae family), which
are particularly sensrtnm to habitat disruptions. Three of the streams
sampled had minimum flow at the time of sampling, yet they sup-
ported 28 or more species of fish, including black bass. Table 1 lists
the results for selected parameters in the small watershed streams.
TABLE 1. Selected Results of Small Watershed Surveys
Ai1ian
Ecoregions
Minimum
DO duiing
crftlcai period’
(mg/U
Averege No.
fish species
coiiected
Predominant
stream
subeireti
Ovefflead
canon
(%)
Mssissippi Afluvial
Plain
27
31
mud sdt
>90
South Central Plains
1 0-1 4
30
sand
>90
As$ainsas Iley
Ouachita Mountains
O s1c Highlands
Boston Mountains
22-30
56-68
45-47
58-64
28
27
17
24
bedroc li
gravel
gravel
gravel
50
50
>75
50
ciitkal period us du
ring the late summer
(high tempera
lure, low flow)
In addition to the seasonal grab samples collected from least-
disturbed streams, in 1984, DPC&E selected representative sites in
each ecoregion to be part of the ambient monitonng network. Monthly
samples collected from these sites were analyzed for the full com-
plement of chemical parameters as well as 11 heavy metals. These
additional data substantiate the seasonal grab samples.
PROGRAM APPUCATIONS
The Arkansas ecoregion program was designed to provide a sound
basis kir reclassifying streams where existing cntena and standards
are either too stringent or too lenient. Afthough the project is not yet
complete and the triennial review at water quality standards will not
take place until 1987 the State has successfully incorporated the
ecoreglon approach into the preparation of use attainability analyses
(UAA.). Under the State’s UM procedures, DPC&E staff conduct a
brief field survey to evaluate a stream where change of use is pro-
posed. The char eristics of this stream are then compared to the
detailed data from the Ieast-distuftied reference stream of similar size
In that region.
For aia’nple, UAA conducted for Caney Creek, a low-flow stream sub-
ject to a single municipal discharge, recommended continuation of
the existing designated use (perennial fishery). This was based on
the strong similarity between fish samples and the macrobenthic
population in both Caney Creek and the least-disturbed reference
stream In the Mississippi Alluvial Plain region. How veç as a result
of the surve Caney Creek was reclassified from “warmwater fish-
ones” to the more explicit descriptor, “small watershed, Mississippi
AIkMaI Plain warmwater fisheries.” Along with the modification in use,
the DO criterion was also revised from 5 mg(L to 3 mg/L to reflect
levels found in similar least-disturbed streams in the region.
lb assist the State in its plan to revise water quality standards in 1987
EA is worlong with DPC&E to develop a national criteria and stand-
ards database and interactive procedure that would allow the State
to easily access and update designated uses and critena for all
reaches or porlions of reaches in EPA’s Reach File.
FUTURE STUDIES
Arkansas has demonstrated the utility of the ecoregion approach for
developing and evaluating water quality standards, particularly those
concerned with the designation of fisheries and DO cnteria. In the
future, the State plans to use the same approach in its development
of water quality standards for tooc pollutants. For example, back-
ground levels of pH and hardness may vary from region to region with
correspondingly different effects on metals texicit
Material bthrs report was umisliedby ,.ioflnGlese; Ai*ansas Depart-
ment of Pollution Control and Ecology; David P Larsen, EPA Erw,ron-
mental Research Laboratot% CorialsYs; and Larry Champagne; U S
EPA Region VI.
This report is produced by EPA to l7lghl,g!rt monitoring and wastoload
allocation activitIes. Contnbuz,ons at information for similar reports are
,rwitad. Please contact E. F Dtab vsk,, EPA, MOSO, WH-55a
401 M Street SW, V tishington, DC 20460(202) 382-705&
-------
EPA
United States
Environmental
Protection A ency
Monitoring and Data Support Division
Office of Water
Washington, DC 20460
November 1986
Water Quality Program Highlights
EPA’s National Dioxin Study
Dioxin” is the generic term for a group of 75 related compounds
known as polyehlonnated dibenzo-p-diodns; h veç in common use
the name refers to the most toxic and thoroughly studied of these
compounds: 2 3 7$4ettachIorodibenzo.p- in, or 2,a7,8-TCDD. This
compound, which has caused toxic effects at concentrations lower
than any other man-made chemical, is not produced intentionally but
rather is a byproduct in the manufacture of several pesticides, chiefly
2,4 ,5-trichlorophenol (2,4,5-1CP). Environmental contamination can
occur (1) during the production of 2,4,5-1CI9 (2) during the production
of pesticides derived from 2,4 ,5-TCP (principaJI ç the herbicides 2,45-T
and silvex), or (3) where any of these pesticides have been applied.
Certain types of combustion sources have also been found to emit
dioxins.
While the use of many products that may cause dioxin contamination
has been suspended, once the compound is in the environment it is
persistent. In addition, dioxin bloaccumulates so that even if present
in extremely low concentrations, it can concentrate in organisms to
much higher levels, increasing the likelihood of hazard.
During the past three years, the U.S. Environmental Protection
4A ency (EPA) has conducted a national monitoring program, using
new and sensitive analytical techniques, to estimate the extent of
2, 7,8-1CDD contamination in the environment. This report highlights
the scope of the National Dioxin Studyand the results for those
portions of the study that were carried out by EPAS Office of Water,
Monitoring and Data Support DMsion.
LEVELS OF CONCERN
Criteria or levels of concern regai ing dioxin contamination have been
established by several Federal agencies for different environmental
media. EPA has estimated an increased lifetime risk of one additional
cancer per 1 million people from drinking water and eating fish from
waters containing dioxin at 1.3 x 10-8 igIL (0.013 parts per
quadrillion). For soil, the Centers for Disease Control considers 1 part
per billion (ppb) to be a level of concern in residential areas (where
there is a potential for ingestion by children). However, this level vanes
depending on land use. Soil concentrations as low as 6 parts per
trillion (ppt) could be of concern in areas where dairy cattle are
grazing, while concentrations above 1 ppb could be acceptable in
many industrial areas.
THE NATIONAL DIOXIN STRATEGY
In 1983, the EPA issued its National Dioxin Strategy. This strategy was
designed to provide a frame rk for the study of dioxin-related prob-
lems, including the nature and extent of dioxin contamination
throughout the country and riska to people and the environment. The
strategy also addressed the clean-up of contaminated sites and the
destruction or disposal of existing dioxin. To implement the
information-gathering portion of the strategy, EPA defined seven
categories (or tiers) of sites for investigation. The tierawere believed
to exhibit a decreasing potential for 2,3,78-TCDD contamination.
Tier 1 — Facilities where 2,4,5-TCP was produced and associated
waste disposal sites.
Tier 2 — Facilities and associated waste disposal sites where
2,4,5-TCP was used as a precursor to manufacture other
pesticide products.
Tier 3 — Sites where 2,4,5-TCP and its derivatives (2,4,5-T, silvex,
erbon, ronnel, hexachlorophene, and isobac 20) were
formulated, blended and packaged.
Tier 4 — Combustion sources.
Tier 5 — Sites where suspected contaminated pesticides were
commercially applied.
Tier 6— Sites where the manufacture of certain organic chemicals
and pesticides could have resulted in the inadvertent
formation of2,3,7 ,8-TCDD.
Tier 7— Control sites where contamination from 2,3,7,8-TCDD was
not suspected.
EPA conducted a complete investigation of all sites in tiers 1 and 2,
which was managed by the Office of Solid Waste and Emergency
Response. However, because of the large number of sites in tiers 3
through 7, only a representative sample of these was investigated in-
ftiall Studies of sites in tiers 3,5,6, and 7 were carried out by EPP s
Office of Water; tier 4 studies were managed by EPA5 Office of Air and
Radiation.
THE NATIONAL DIOXIN STUDY
The National Dioxin Study was designed to characterize the extent
of 2 3,7 8TCDD contamination in tiers 3,5,6, and 7. To accomplish this,
over 4,000 samples, in various media, from 862 sites across the na-
tion were collected and analyzed.
Several approaches were used to identify the potential sampling sites.
For tiers 3 and 5, samples were taken at both statistically selected sites
(using EPA and industry databases) and at sites of particular interest
to States and the EPA Regional Offices. Sites in tier 5 were selected
based on pesticide use information provided by EPAs Office of
Festicide Programs, the Regional Offices, and State agencies. In this
tier, a statistical sample was not practical, and sites were selected to
represent a wide range of conditions and uses. For tier 7, sampling
sites for soil, fish, and shellfish were statistically selected from three
national environmental monitoring neMorks; in addition, Regional Of-
fices selected fish sampling stations of particular interest.
Sample analysis in the National Dioxin Study was carried out by both
commercial laboratories under EPA contract and EPA research
laboratories. Analytical methods used in the commercial labs had a
nominal detection limit of 1 ppb for soils, while methods used by the
EPA had a detection limit of approximately 1 ppt for all media other
than water and approximately 10 parts per quadrillion (ppq) for water.
GeneralI ç commercial laboratories analyzed soil samples from tiers
3 and 6; EPA analyzed soil samples from tiers 5 and 7, as well as
samples from other media in all tiers.
STUDY RESULTS: TIERS 3,5, AND 6
Sites in tiers 3,5, and 6 were suspected of showing 2,3 ,7,8-TCDD con-
tamination. Sampled media in these tiers included soil, water, stream
sediment, and biological tissue. Table 1 summarizes the results for
tiers 3,5, and 6, while the following paragraphs highlight the conclu-
sions for each tier.
Tier 3. This study was designed to evaluate the percentage of all tier
3 facilities expected to have soil contamination above 1 ppb or at any
detectable level in other environmental media. To accomplish this,
EPA statistically selected 61 facilities that formulated one or more tier
3 compounds between 1976 and 1981. Of these, 41 were actually
sampled. (Information request letters revealed that 20 facilities either
did not handle tier 3 compounds, or that soil on the site had been ex-
-------
TABLE 1. Results of the National Dioxin Study for Tiers 3, 5, and 6.
of spray areas is not warranted.
Tier 3, statistically selected sites
No. of
sites
No. of
sites
sampled
No. of
sites
contaminated
61
41
Tier 3 additional sites
23
23
Tier 5 sites
26
26
15 b
Tier 6, statistically selected sites
25
15
Tier 6, additional sites
3
3
l
aS tes with soil concentrations greater than 1 ppb or detectable levels in
other media.
bSites with detectable levels (ppt) in soils, stream sediment, or biological
tissue.
tensively disturbed or paved over.) Additionally, 23 sites were iden-
tified by EPA Regional Offices based on known activities atafacility
or on previous contamination incidents.
Based on the survey results, EPA estimated that of all tier 3 facilities
nationwide (approximately 300), about 10 percent would have con-
centrations of 2,3,7 ,8-TCDD in soil above 1 ppb or detectable levels
(above approximately 10 ppq-lppt) in other media.
Of the 12 sites where contamination was detected, all were at or near
facilities handling the pesticides 2,4,5-TCP, 2,4,5-1, and/or silvex. As
shown in Figure 1 (six statistical and six regionally selected sites),
only two tier 3-sites were extensively contaminated. As a resul’ of
these findings, EPA concluded that the immediate investigation of
the remaining tier 3 sites is not warranted, although the Agency will
conduct further evaluations of specific large pesticide manufacturers
where 2,4,5-TCP, 2,4,5-1, and silvex are formulated.
: L
14 16 20 25 10 36 26 10 10 19 53 40
I
0
S
2
S
100
Statistically Selected
Sites
Number of Soil Samples Analyzed
Cr tamination was in media other than soil
u: 1. Percentage of soil samples from tier 3 contaminated sites
y greater than 1 ppb 2,3,7,8-TCDD.
Tier 5. For tier 5, the objective was to determine whether 2,3,7,8-TCDD
1 h ’ ‘ ete -ted at any level in soils, stream sediment, or biological
‘i ues fl areas where pesticides suspected of containing dioxin had
h r’n ‘ d. Twenty-six sites were sampled including forested areas,
rice and sugarcane fields, rangeland, and aquatic sites.
At the 15 sites where 2,3,7 8-TCDD was detected, soil and sediment
contamination was extensive, with over 40 percent of the samples
analyzed at each site containing levels above the ppt detection limit.
Two sites had detectable levels in fish, and at one of these, all fish
samples were contaminated. While contamination in soil and sedi-
ment was widespread, concentrations were generally cuite low (less
than 5 ppt). Levels detected in fish fillets were between 8 and 23 ppt,
and dioxin was not detected in other animal tissue or n vegetation
samples.
Of particular interest was the fact that, with the exception of loading
areas, 2,3,7,8-TCDD levels were much lower, and in most cases, net
detectable in areas where pesticides were uniformly applied by spray-
ing. Due to low levels found at tier 5 sites where 2,4,5-TCP-based
spraying occurred, EPA concluded that further national investigation
Tier 6. The objective in tier 6, as in tier 3, was to determine the percen-
tage of facilities nationwide expected to have soil contamination abc e
1 ppb or at detectable levels in other media. EPA identified 67 facilities
thought to manufacture one or more of the 60 compounds whose
production can create dioxin. Of these, 25 sites were selected for
sampling; three additional tier 6 sites were selected by EPA Regional
Offices based on known activities or previous contamination
incidents.
None of the three sites where 2 7 ,8-TCDD was found were extensive-
ly contaminated (detectable levels were limited to one or two
samples). As a result, it was concluded that additional investigation
of the tier 6 sites was not warranted.
STUDY RESULTS: TIER 7
EPA used tier 7 sites to evaluate background levels of 2,3,7$-TCDD
in soil and fish tissues. Thus, sampling was carried out at sites that
had no previously known sources of dioxin contamination. Using
established national databases, EPA randomly selected locations for
testing rural soils, urban soils, and fish tissue. Also, additional fish
sampling sites were selected based on proximity to population
centers or recreational fishing activity.
The results, summarized in Table 2, show that 2,3,7$-TCDD was
detected infrequently and at very low levels in soil, while background
fish contamination was somewhat higher. Of the fish samples, the
highest proportion of contaminated samples was found in Great
Lakes sites. This is consistent with previous findings and is a result,
to some extent, of the long water retention times (which tend to
increase bioaccumulation potential) and the many pollutants entering
the lakes.
TABLE 2. Results of Tier 7 Dioxin Study:
Background Levels of Contamination
No. of
sites
sampled
No. of
sites
contamInated
Maximum conc.
detected
(ppt)
Urban soils
221
Rural soils
138
17
11.2
Fish tissue
1
0.5
Fish tissuea
305
17
95
19
85
aSites were not statistically selected.
The two sites with the highest levels of dioxin contamination were the
Androscoggin River in Maine (maximum of 29 ppt) and the Rainy
River in Minnesota (maximum of 85 ppt), and in both cases, the rivers
were subject to upstream pulp and paper mill discharges. Further
investigations at these and similar sites are being conducted by EPA,
the States, and the paper industry to determine the sources of
2 ,3,7,8-TCDD within the mills. As a result of the study, fish consump-
tion advisories were issued by the two States.
MAJOR STUDY ACCOMPLISHMENTS
The National Dioxin Study has produced a major increase in EPA’s
knowledge of 2,3,7,8-TCDD levels in the environment, and it has also
helped to refine the tools needed for futher study. Guidance to en-
sure uniform sampling procedures for dioxin monitoring and the
development of uniform review procedures to assess analytical data
are two important accomplishments. In addition, the study has
resulted in the development of more reliable and less costly analytical
methods to routinely measure dioxin at concentrations that are close
to the levels of concern for human health.
Material for this report was furnished by Stephen Kroner of EPA’S
Monitoring and Data Support Division.
This report is produced by EPA to highlight monitoring and wasteload 1
allocation activities. Contributions of information for similar reports are
invited. Please contact E. F. Orabkowskj, EPA, MDSD, WH-553,
401 M Street S. W., Washington, DC 20460 (202) 382-7056.
Regionally Selected
Sites
-------
&EPA
United States
Environmental
Protection ency
Monitonng and Data Support Division
Office of Water
shington, DC 20460
January 1987
Water Quality Program Highlights
— p p p p p p p
‘ - W L L% — - -
The Massachusetts Fish Toxics Monitoring Program
Toxic water pollutants can affect human health through several path-
ways including the ingestion of contaminated fish and shellfish In
addition, toxicants can have significant adverse impacts on the health
of aquatic life Since the late 1970’s, the US Environmental Protec-
tion Agency (EPA) has encouraged States to monitor and control the
discharge of toxics An important part of such efforts includes the
determination of contaminant levels in fish tissue and the risks
associated with these levels Of the many known toxicants, numer-
ical criteria for determining fish edibility currently exist for only a hand-
ful of substances These numerical criteria are the ‘action levels”
developed by the US Food and Drug Administration for eight pesti-
cides and for mercury, pol hIonnated biphenyls (PCBs), and dioxin
A program currently underway in the Commonwealth of Massachu-
setts is monitoring toxic substances in fish tissue This report
describes the program’s implementation and demonstrates how toxic
rrionitonng results were used to assess human health nsks associated
with the consumption of fish from the Sudbury River
IDENTIFYING THE NEED FOR A FISH TOXICS PROGRAM
The construction of municipal and industrial water pollution control
projects has significantly reduced pollutant loadings to Massachu-
setts rivers and coastal waters The resulting water quality improve-
ments have led to the reestablishment of viable and diverse fish
populations in many nver systems that previously sustained few, if any,
species of fish This enhancement of recreational fishing oppor-
tunities coupled with the increased awareness of toxic pollution led
to concerns about the health risks associated with consuming
freshwater fish from some Massachusetts waters Major sources of
toxics in Massachusetts’ waters are industrial discharges such as
those from electroplating facilities, discharges from municipal
wastewater plants that receive significant industrial input, and non-
point sources including urban runoff and in-place sediments.
In 1984, the Massachusetts Department of Environmental Quality
Engineering’s Division of Water Pollution Control (MDWPC) and the
Division of Fisheries and Wildlife (MDFW) established the Toxic
Chemicals in Fish Program To address the problem in a comprehen-
sive way, a Toxics in Fish Committee was formed, with membership
consisting of representatives from several State and Federal agen-
cies Early discussions revealed that while existing data on fish con-
tamination in the State were sparse, such information would be useful
in the development of surface water quality standards, NPDES per-
mits, and human health risk assessments.
IMPLEMENTING THE PROGRAM
Massachusetts implemented the Toxics in Fish Program to
accomplish three major objectives
• Develop a Statewide database of levels of toxic contaminants in
freshwater fish
• Identify waters where levels of toxic chemicals in fish may Impact
human health
• Identify waters where toxic chemicals may impact fish populations
and other aquatic life
To date, fish surveys generally have been restricted to waterbodies
where discharge information (eg , NPDES permit applications) or
previous studies have suggested a potential toxics problem In addi-
tion, sampling has been conducted in areas where heavy industnal
development or hazardous waste disposal are present Because of
limited resources, human health concerns have received highest
priority in the surveys carried out thus far, for this reason, fish tissue
analysis has been restricted to edible fish fillets
The fish toxics monitoring program is carried out using a three-
phased approach for data collection
1 Screening Survey During this initial survey, five to ten fish from at
least two species are collected for analysis Target species include at
least one bottom feeder (e g , bullhead) and one resident game fish
(eg , lax emouth bassorpickerel) Leftfilletsarepooled andanalyzed
as a composite, right fillets are archived as individual samples
2 Confirmatory Analysis If levels found in screening composites are
high relative to existing criteria (where they exist) or baseline literature
values, the archived fillets are individually analyzed to provide a range
of concentrations, and a mean value is calculated
a Follow-up Survey If contaminant concentrations are at levels of
concern in individ ial fillets, additional fish of several different species
are collected and individually analyzed
In some cases, MDWPC has begun analyzing both individual fillets
and composites in screening surveys This change was necessary
because of variability in both the size of fish collected, and in the
number of fish per sample Fish are now individually analyzed Cur-
ing screening surveys if (1) only one fish of a desirable species is cap-
tured, or (2) if one of five samples is significantly larger than the other
four In the second case, the other four are combined for a composite
analysis, and, as before, the right fillets are archived for future
reference
Uniform protocols, designed to assure accuracy and prevent cross
contamination of samples, are followed for fish collection, process-
ing and shipping Fish are taken with electroshocking gear or gill nets
Lengths and weights are measured, and fish are visually examined
for tumors, lesions, or other indications of disease Scale samples or
other hai parts (e g , pectoral spines) are obtained from each sample
to determine the approximate age of the fish
Tissue samples are frozen before being transported to the laboratory
for analysis Preparation of the samples differs depending on the
nature of the suspected toxicant where organics are suspected,
samples are wrapped in aluminum foil, for metals, plastic wrap is
used Once at the laboratory, samples may be analyzed for specific
toxicants (where a particular source is suspected) or for a broad spec-
trum of heavy metals, pesticides, or organic chemicals Laboratory
analytical methods include atomic absorption spectroscopy and gas
chromatography/mass spectrometry
ADMINISTERING THE PROGRAM
The fish toxics monitonng program is a cooperative effort that invetves
several State agencies MDWPC presently employs one full-time
aquatic biologist to direct the program (this position is funded under
the Clean Water Pct’s Section 106 Program Grant), and he is assisted
with field work and sample preparation by one or more seasonal
employses during summer months. In addition, MDFW provides a
biologist to assist with sampling and furnishes necessary equipment,
such as electroshocking gear and nets Massachusetts’ Lawrence
Experiment Station analyzes fish and related samples for toxic
-------
chemicals and also advises the program commitlee on analytical
methods and data interpretation Finally, the agency’s Office of
Research and Standards coordinates with State and local public
health agencies who communicate health risks to the public when
necessary
In 1986i annual operating costs associated with the Toxics in Fish Pro-
gram included approximately $27,000 for sample collection, prepa-
ration, and data management, $15,000 for chemical analyses, and
$3,000 for administrative costs, data analysis, and the preparation of
health advisories These costs include the salary of one full-time
biologist
TOXICS MONITORING TO DATE
MDWPC performed its first fish toxicswerk in response to public con-
cern about the environmental impact and human health effects of
2,3,7$-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD or dioxin) The
presence of dioxin was suspected in several Massachusetts lakes
and ponds that had been treated with the herbicides Silvex, Kuron,
or 2,4 -T (dioxin can occur as a contaminant in these herbicides), and
in 1983, MDWPC and MDFW conducted a screening survey of six
ponds where these three herbicides had been used
In five of the six ponds surveyed, dioxin levels in both fish tissue and
in sediment samples were below the detection limits However, in Lake
Winthrop, the composite fish sample measured 71 parts per trillion
(ppt) — exceeding the FDA’s dioxin advisory for consumption of fish
caught in the Great Lakes (50 ppt) As a result, the Massachusetts
Department of Public Health issued a health advisory concerning
fish caught in Lake Winthrop
Follow-up studies, currently underway in Lake Winthrop, include fur-
ther analyses of sediment, fish, and aquatic vegetation for the
presence of dioxin, the examination of historical records of herbicide
applications, and the investigation of other possible sources of the
contaminant
In 1984, fish tissue monitonng was carned out as part of a water qual
ty survey designed to support the reissuance of 17 major NPDES per-
mits in Massachusetts’ Ten Mile River Basin The basin contains a
number of metal plating and finishing plants, and the development
of wasteload allocations and permit limits required a thorough evalua-
tion of the water quality impacts from existing toxic discharges Five
composite samples of fillets from nine sites were analyzed for ten
metals, and results were compared with concentrations reported in
the literature as having adverse impacts on fish Chromium, copper,
lead, nickel, and zinc were present in fish at concentrations only
slightly below those literature va es, and of the metals sampled, lead
and mercury approached levels that could cause public health im-
pacts Findings from this survey resulted in a fish consumption ad-
visory (because of elevated lead levels in Ten Mile River fish) and the
justification for advanced waste treatment and stringent NPDES per-
mit limits for many industrial facilities in the river basin
In 1985, the Toxics in Fish Program was expanded to include ten dif-
ferent fish flesh monitoring surveys Two surveys sampled lakes
located near hazardous waste sites The remaining eight surveys ad-
dressed rivenne waterbodies, and all but one were screening surveys
The one exception was the Sudbury River where previous work by
MDFW suggested the need for follow-up surveys
A total of 12 surveys were performed during 1986, the majority
designed to screen for metals contamination, although some anal-
yses for PCBs and other organics were also done The Toxics in Fish
Program was also expanded in several directions A comprehensive
microcomputer database was created for all fish toxics data Also,
sampling is taking place in clean-water or “least-impacted” water-
bodies to begin the creation of a reference database containing back-
ground levels for various chemicals This reference database will be
enlarged as additional clean water sites are sampled in the future
SUDBURY RIVER CASE STUDY
The Nyanza Chemical Company, a manufacturer of organic dyes,
discharged wastes near the headwaters of the Sudbury River be-
tween 1917 and 1970 Downstream of the discharge are two large
reservoirs, several small impoundments, and an extensive winding
wetland, part of which is included in the Great Meadows National
Wildlife Refuge Although only the headwater segment of the river is
stocked with trout, the entire nver, above and below the reserveirs, has
long been popular with anglers, many of whom consume their
catches of bass, perch, and bullheads Sampling of fish tissue per-
formed in the Sudbury River in 1972 revealed high levels of chromium,
mercury, and other toxic substances A second fish sutvey conducted
in 1981 confirmed the high metals concentrations, and in 1983 EPA
identified the Nyanza properly as a Superfund hazardous waste site
Follow-up sampling was conducted in 1985 to determine whether fish
from the Sudbury River contain mercury levels in excess of the FOAs
10 part per million (ppm) market standard and to evaluate mercury
concentrations in the water column and sediments MDWPC chose
six sampling locations that coincided with the stations used in 1981
One station was upstream of the Nyanza site, with five stations spread
along the approximately 20 miles below the site Five individual fillets
were analyzed at each station, and water and sediment samples were
collected from all but the most downstream site
Table 1 presents the results of the 1985 study The highest contami-
nant levels were found in three resident species often taken by anglers
in the two reservoirs Six of the ten fish sampled at stations 2 and 3
had mercury concentrations greater than 1.0 ppm and as high as a2
ppm Although the mercury concentrations declined in fish taken fur-
ther downstream, at four of the six stations at least one fish exceed-
ed the FDA market standard
TABLE 1. Mercury Contamination in the Sudbury River, Upstream
(Station 1) and Downstream (Stations 2-6) from the Nyanza
Hazardous Waste Site, 1985
Station
No
Fish Fiiiets
Water
Column
Sediment
(ppm)
Ran9e
(ppm)
Mean % of Samples
(ppm) >1 ppm
1
012-063
038
0
ND
024
2
030-22
140
60
ND
13
3
002-32
119
60
ND
72
4
042-ic
067
20
ND
13
5
002-i 2
053
40
ND
0022
6
019-070
042
0
—
—
ND — Not
detected (deiect
on iimii = 02
ppm)
All data from the Sudbury River Study were submitted to the Office
of Research and Standards, which then coordinated with the Statds
Department of Public Health and other members of the Toxics in Fish
Committee The result was a decision to post an advisory against
eating fish along the entire length of the Sudbury River
FUTURE CONSIDERATIONS
Creation of a baseline or reference database along with continued
monitoring of fish tissue in the Commonwealth’s nvers and lakes are
long-term goals of the Toxics in Fish Program Experimentation with
akernative preparations, e g , skin off vs skin on, will also be address-
ed in future ysars In addition to addressing human health considera-
tions, MDWPC hopes to standardize techniques for assessing the im-
pact of toxic pollutants on aquatic life To do this, methods for process-
ing and analyzing fish may have to be modified to include the analysis
of whole fish and/or specific target organs known to bioaccumulate
toxicants
Material for this report was furnished by Arthur Johnson, Robert
Majetta, and John Jonasch, Massachusetts Division of Water Pollu-
tion Control, Technical Services Branch and MichaelBilge US EPA
Region I, Environmental Services Division
This report is produced by EPA to highlight monitoring and wasteload
allocation activities Contnbutions of information for similar reports are in-
vited Please contact E F D,abko.vski, EPA, MDSD, WH-55 401 M Street,
SW, Washington, DC 20460, (202) 382-7056
-------
EPA
United States
Environmental
Protection Agency
Office of Water Regulations and Standards
Monitonng and Data Support Division
Washington, DC 20460
Apnl 1988
- _ w
Water Quality Program Highlights
Maine’s Biologically Based Water Quality Standards
Section 101 of the Clean Water Act slates that “it is the objective of
the Act to restore and maintain the chemical, physical, and biologi-
cal integnty of the Nation’s waters” Of the three characteristics, bio-
logical integnty has been the least considered, however, it may be
the most important since organisms not only integrate the full range
of environmental influences (chemical, physical, and biological), but
complete their life cycles in the water and, as such, are continuous
monitors of environmental quality
To take advantage of this fact and to improve its surface water
management capabilities, the State of Maine has developed biolog-
ically based water quality classifications and water quality critena
In Apnl 1986, after four years of negotiation with industry and environ-
mental groups, the Maine Legislature enacted a revised Water Quality
Classification Law. The new law includes language specifically
designed to facilitate the use of biological assessments. Thus, each
water class contains aquatic life standards that descnbe the minimum
conditions necessary to attain that class. To implement the new clas-
sification system, the Maine Department of Environmental Protec-
tion (MDEP) is now developing specific biological measures that will
be used to support the statutory aquatic life standards in the Water
Quality Classification Law
This report summarizes Maine’s biologically based water classifica-
tion system and the associated aquatic life standards for freshwater
streams and rivers Also summarized is the State’s instream biologi-
cal monitoring program that will be used to ensure compliance with
the new law.
A BIOLOGICALLY BASED CLASSIFICATION SYSTEM
The 1986 law that revised Maine’s water classification system was
not designed to change existing water quality levels but to improve
MDEP’s ability to monitor and manage rivers and streams Under a
previous law, the same aquatic life standard, “Discharges shall cause
no harm to aquatic life:’ applied to four classes of waterbodies (A,
B-i, B-2, and C) However, countless biological studies demonstrated
the impossibility of enforcing such a restrictive standard across all
classes of effluent-receiving waters Maine waters that were clearly
attaining the minimum chemical and physical standards of Class C
could not meet the ‘no harm to aquatic life” criterion due to the dis-
placement of some sensitive indigenous species
The revised classification system recognizes the necessity of hav-
ing waters of different quality, including pristine recreation-onented
waters and waters of lesser quality for economic and social needs
Table 1 descnbes the aquatic life standards for waters under the new
classification system In addition to these standards, the Water Quality
Classification Law specifies designated uses, dissolved oxygen lev-
els, and allowable bactena concentrations for each of the four classes
Class AA, the class with the highest degree of protection, is intended
for waters of special value to the State No impoundments or dis-
charges of any kind are permitted, consequently, no change in the
biological community is expected and the standard stales that aquatic
life should be “as naturally occurs” This is interpreted to mean that
essentially the same species and numbers of organisms should be
found as in similar habitats that are free of human influence
TABLE 1. Aquatic Life Standards for the State of Maine
Water quality
class
Biological standards
AA
No direct discharge of poiiutants, aquatic life shall be
as naturally occurs
A
Natural habitat for aquatic life, aquatic life shall be as
naturally occurs
B
Unimpaired habitat for aquatic life, discharges shall
not cause adverse Impact to aquatic life in that the
receiving waters shall be of sufficient quality to sup-
port all aquatic species indigenous to the receiving
water without detrimental changes in the resident
biological community
C
Habitat for aquatic life, discharges may cause some
changes to aquatic life, provided that the receiving
waters shall be of sufficient quality to support all spe-
cies of fish indigenous to the receiving waters and
maintain the structure and function of the resident
biological community
Class A waters would be managed much as Class AA, although
hydropower and some highly treated effluents would be permitted
(discharged effluent must be equal to or better than existing quality
of receiving waters) Because of the expected high level of treatment,
the same standard is used that aquatic life shall be as naturally
occurs
The third level, Class B, requires that discharges have no adverse
impact on aquatic life, but allows for some changes in the residen-
tial biological community This standard has two distinct parts The
first is that the receiving water be of sufficient quality to support all
indigenous species—to be determined through effluent toxicity testing
(using Ceriodaphnia and brook or brown trout) where effluent is
appropnately diluted in receiving water This does not mean that a
species has to exist in the nver or stream—only that water quality can-
not be the limiting factor The second part is that changes in the resi-
dent community must not be detnmental, under the new law,
detnmental change is defined as a significant loss of species or
excessive dominance by any species or group as a result of human
activity (The previous law stated that the composition of the bottom
fauna could not be altered) Class B waters typically receive dis-
charges from municipal wastewater treatment plants. in such cases,
increased nutnent and organic loading, in the absence of toxic com-
pounds, has an ennching effect on the biological community, result-
ing in increased numbers of individuals Generally, however, such
effects are not harmful and should be differentiated from detrimen-
tal changes.
Class C waters are the lowest level in the Maine system in these
waterbodies, discharges may cause some changes to aquatic life,
provided that water quality continues to support all indigenous spe-
cies of fish and the biological structure and function of the corn mu-
nity is maintained. The first part, which cart be determined through
effluent texicily testing (using Cenodaphnia as the most sensitive sur-
rogate species for trout), is necessary to meet the requirements of
-------
the Clean Water Act Again, it is not necessary that all indigenous
fish are actually found; factors other than water quality (e g , compe-
tition, predation, lack of habitat) may preclude the presence of some
species. The second part of the Class C standaiti is that community
structure and function be maintained Bnefly, structure is the num-
ber of species and individuals within a community, while function is
the means by which they interact to utilize food and other resources.
Within Class C waters, significant losses and shifts in species would
be allowed Although it is expected that pollutant-sensitive species
may disappeai it is essential that there is some replacement by more
tolerant species and that these tolerant species fulfill all vital func-
tional roles in the aquatic community
These biological standards allow Maine to assess the success of its
overall water quality program in terms of cumulative biological effects
and provide the statutory framework to directly protect the actual con-
dition of aquatic life
MACROINVERTEBRATE MONITORING PROGRAM
In order to develop specific numeric and descnptive cntena to sup-
port the aquatic life standards in the State’s Water Quality Classifi-
cation Law, the MDEP initiated a statewide biological monitonng
program in 198 Under this program, benthic macroinvertebrates
(eg., aquatic insects, snails, clams, worms, and crayfish) were
selected as the pnmary indicators of biological integnty in streams
and nvers. These organisms were chosen because of their
• Limited mobility(compared to fish) which makes them less able
to avoid the effects of pollutants
• Longer, more complicated life cycles (compared to algae or bac-
tena) which allow them to reflect or integrate water quality over
time
• Wide range of pollutant tolerances among the venous species
• Importance as a food source for freshwater gamefish
• More diverse feeding and energy use strategies than higher level
organisms (e g fish) and thus the ability to provide information
about disturbances in nutrient cycling through the ecosystem
• Ease of colleoton using accepted and wellestabhshed sampling
and analysis procedures
Sampling locations were chosen with two major considerations
(1) to represent the range of water quality conditions in the State (e g,
on different size streams, both with and without discharges) and
(2) to provide information on the presumed worst-case condition of
all nver and stream reaches known to be significantly affected by
human activity Thus, the benthic macroinvertebrate data will serve
the dual purposes of generating qualitative and quantitative biologi-
cal classification cntena and assigning reach-by-reach biological das-
sifications
To date, MDEP has collected macroinvertebrate samples from 161
sites on 55 nvers Samples were taken above and below all signifi-
cant discharges in the State, as well as from some pnstine areas, and
this information is now being subjected to extensive statistical anal-
yses All decisions about specific numenc critena will be made after
the data have been thoroughly analyzed. MDEP, which is responsi-
ble for proposing waterbody classifications and assessing compli-
ance (actual assignments are made by the State legislature), expects
to complete this process during 1988.
BIOMONITORING PROTOCOL
Measures of biological integnty are fundamentally dependent on the
sampling methods used to collect raw data Thus, standard practices
must be set for all evaluation procedures including both environmental
factors (e.g., season and type of habitat sampled) and methodologi-
cal factors (eg , sampling device and sieve size used to collect or-
ganisms) To ensure consistency, the MDEP has prepared a manual
(Methods for Biological Sampling and Analysis of Maine Waters,
January, 1987) that establishes specific guidelines and procedures
for site selection, sample collection, and sample analysis
For example, biomonitonng efforts are to focus on benthic macroin-
vertebrate communities of flowing streams and rivers having a hard
eroded substrate which is characteristic of the majority of effluent
receiving waters in Maine Sampling sites are to be representative
of the stream or river reach as a whole and should not be influenced
by man-made structures, river bank effects, or slackwater areas
Reference (control) and effluent-impacted sites must be matched for
similarities in water velocity, substrate composition, canopy cover,
water depth, and all other upstream influences except the discharge
source to be evaluated.
DEVELOPMENT OF NUMERIC AND DESCRIPTIVE CRITERIA
With its welkiefined biological classification system and standardized
benthic macroinvertebrate database in place, Maine is now in the
process of developing the specific numeric and descnptive critena
necessary to identify the biological classification attained by a given
waterbody The cntena are being developed (in cooperation with biol-
ogists from industry and environmental interests) by examining the
responses of various biological indices and measures across the
range of water quality conditions represented in the database Table
2 presents the general types of measures that will be applied to each
classification
TABLE 2. Potential Measures for Biological Statutory Cntena
Water
quality
class
Standard
1 rpe of biological measure
AA & A
As naturally
occurs
Comparative measures
(eg , percent similarity)
B
No detnmental
change
Population and community measures
(e g , coefficient of community loss,
species population reduction, reten-
tion/recruitment)
C
Maintenance
of community
structure and
function
Structure measures (a g richness,
abundance, diversity)
Function measures (eg ,feeding
groups, specialist/generalist ratio)
The measures chosen will be those which demonstrate a consistent
correlation to known water quality conditions and which best address
the specific statutory language for each classification Measures that
are selected will be incorporated into an hierarchical evaluatioh
process that proceeds from the least ambiguous conditions (e g ,total
absence of aquatic life or 90 percent dominance by one type of orga-
nism) to conditions requinng more rigorous biological interpretation
Past expenence has shown that no single biological measure or index
can provide an adequate summary of community status, so MDEP
will employ several evaluation measures to confirm each decision
In most cases, the interpretation of biomonitonng results will depend
upon an examination of the relative differences between a down-
stream test community (below a discharger) and a matched reference
community (usually upstream) that is not impacted In other cases,
a matched reference site may not exist, and the site must be eval-
uated on the degree to which descriptors of the benthic community
are comparable to charactenstics of communities from similar natural
areas
FUTURE CONSIDERATIONS
The pnmary goal for MDEP’s instream biomonitoring program is to
provide feedback concerning the State’s efforts in protecting its
aquatic life resources The program is not expected to have a sig-
nificant role in permitting, although effluent toxicity testing has been
required in some discharge permits for the last three years Informa-
tion from the program will be used, however, to assess the degree
of protection afforded by effluent limitations. As the freshwater bio-
monitonng program matures, the State will be devoting increasing
attention to the development of a marine biomonitoring program to
complement similar biological standards adopted by the State for
coastal waters
Material for this report was furnished by Susan Davies and David
Courtemanch, Maine Department of Environmental Protection,
Bureau of Water Quality Control, and Mike Bilge US EPA Region
I, Environmental Services Division
This report is produced by EPA to highlight monitoring and wasteload
allocation activities Contributions of information for similar reports are
invited Please contact E F Drabkowski, EPA, MDSD, WH-55a
401 M Street SW, Washington, DC 20460, (202) 382-7056
-------
United States
Environmental
Pm on Agency
Office of Water Regulations and Standards
Monitoring and Data Support DMsion
Washingto DC 20460
Water Quality Program Highlights
w W
Minnesota’s Nonpoint Source Assessment Program
Sect n 319 of the 1987 amendments to the Clean Water Act author.
ized a new direction and significant Federai financial assistance br
the expansion of Stale nonpoint source (NPS) control piograrns. The
U.S. Environmental Protection Agency (EPA) is encouraging States
to build on existing intormation to develop NPS programs that ndude
three major elements:
• A cornprehensrve assessment of State waters impacted by NPS
pollutants
• A procedure kx targeting high-pnority geographic areas
• A State Management Program describing ions that will be taken
on a watershed-by-watershed basis.
The Minnesota Pollution Control Agency (MPCA) is developing a
procedure to rank the NPS pollution potential of all watersheds in the
State The procedure will indude three s of analysis L el I, com•
pleted in November 1986, assessed the general vulnerability (to non -
point pollution) of s n ecoregions within the State and identified
the major NPS pollutants. Level II, which is n ’ bein completed,
will assess NPS problems at a smaller scale by identifying the poten-
tial for water quality impacts in the State’s 5,611 “minor watersheds.”
Fina1 Level Ill analyses wiH be condu d as neces by local unrts
of government to address source-specific nonpoint loading prior to
the implementation of controls.
This report highlights the approach and the specific methods to be
used by the MPC.A in each level of its NPS assessment procedures.
LEVEL I: ECOREGION NONPOINT SOURCE ASSESSMENT
Both the Level I and Level II assessments are based on the concept
of aquatic ecoregions as developed by EPA’s Environmental
Research Laboratory in Corvallis, Oregon. Using this approach, areas
with similar land use, sods, topography, and potential natural vege-
tation are categonzed into ideritthable geographic areas or
ecoreg”ions. Waterbodies within these areas generaily exhibit physi-
cal, chemical, and biological characteristics that are rrore similar to
each other than to walerbodies in other ecoregions. Thus, streams
and lakes within the same ecoregion are likely to react similarty when
subjected to NPS pollutant loads. Figure 1 shows the seven aquatic:
ecoregions identified within Minnesota.
Source MPCA
The Level I study was designed to pr ’ide information about major
problems r broad geographic areas. For mple. the Level I anal-
ysis suggested that, while lakes r the North Central Hard od Forest
are not currently experiencing widespread problems, many water-
bodies in this region are near critical nutrient levels so that minor
additional pollutant loading may result in use impairment. This is in
contrast to lakes in the Western Combeft Plains ecoregion where
nutrient levels are uniformly h h and pollution related problems corn-
mon. Information such as this helps the State locus its efforts and
programs on protective or restorative activities within a given region.
The MPCA assessed existing and potential NPS pollution for each
ecoregion by examining three types of information in each area land
use, topographic features, and water qualit Land use and topo-
graphic leatures were assessed by using Minnesota’s Land Manage-
merit Information Center data base, which was de e3oped from aerial
photo interpretations of each 40-acre parcel in the Stats Water quality
data were taken from EPA’s data storage and retrievai system,
STORET. The general approach used for the ecoregion analysis is
outlined in the following paragraphs.
Land Use Assessment. It was assumed that land uses maybe rated
according to their “intensity” and that high-intensity uses encourage
greater NPS pollution. Urban development, mineral extraction, trans-
portation, r rops (corn and soybeans), and small grain cuttivation
were categorized as intensive land use actMties. Pasture and open
lands, forested areas, water and marsh were considered to represent
lower intensity land use activities.
bpogrsphical Assessment MPCA selected four topographical fea-
tures associated with NPS pollution potential: water orientation, slope,
soil texture, and soil hydrologic group. Each 40-acre parcel was de-
fined as water oriented if it either contained or adjoined a waterbody.
Because an inverse relationship exists between NPS pollutant de liv
ery and the distance to the nearest watercourse, nonpomt loading
is likety to be greater where a large proportion of the 40-acre parcels
are water oriented—assuming that each land area generates equal
pollutant loads.
The generation of runoff and the transport of NPS pollutants to water-
bodies is strongly influenced by the other three topographic features
that were evaluated. Steep slopes, which are associated with
increased runoff potential, were evaluated for each ecoregion by
grouping the average slope for each 40-acre parcel into five slope
categories: <1%, 1-2%, 2-3%, 3-6%, and >6°/o. Soil texture affects
runoff because fine-textured soils generally limit infiltration; in addi-
tion, once suspended, fine soils are more easily transported, and their
larger surlace area enhances the adsorption of pollutants. Soil tex-
ture was evaluated by grouping 40-acre parcel data into tour
categories: sand, silt, clay, and water (including peat and mine
dumps). Finally, soil hydrologic groups were evaluated as a direct
measure of soil permeability. For this evaluation, each 40-acre par-
cel was grouped into one of five categones according to the rate at
which their soils uld transmit water high, moderate, slow, very slow,
and no rating.
Water Quality Assessment. MPCA examined monitoring data col-
lected over a 12-year period (1973-1985) from 149 ambient monitor-
ing stations segregated by ecoregion. Other monitoring stations
whose drainage area included large areas of more than one
ecoregion or did not include at least 4 years of data were excluded.
IS
May1988
FIGURE 1. The seven aquatic ecoregions in Minnesota.
-------
The resuPts of the mrtiai assessment sh both increasing and
decreasing water quality trends over the 12-year period. As expected,
waterbodies in ecoregions that are more intensively developed tend
to have high and increasing concentrations of pollutants such as total
suspended solids, n Itrite, and nitrate, while the less developed
ecoregions are not as affected by these pollutants.
STORET data was used to calculate the mean values (by ecoregron)
for 10 water qualrty parameters (temperature, pH, conductivity,
suspended solids, turbidity, nitrate + nitrite, ammonia. total phos-
phorus, BCD 5 , and fecal colrforms). It was then necessary to deter-
mine statisticaily significant differences among these means. The
usual test for companng means is the West, but it was necessary to
simultaneously compare means from several ecoregions, so Dun-
can’s multiple range test was used. This procedure identified a set
of ecoregions for each water quality parameter where the mean value
for that parameter is statistically different. Because Duncan’s multi-
ple range test is a parametric measure and water quality data is sel-
dom normally distributed, the MPCA is currently im#estigabng the use
of a nonparametnc measure of statistical difference.
LEVEL 11: MINOR WATERSHED NONPOINT SOURCE
ASSESSMENT
The Level II analysis, designed to assess NPS pollution potential for
each of the State’s 5,611 minor watersheds, is based on statistically
significant relationships between water quality parameters and the
venous ecoregion characteristics. The analysis can be divided into
two steps as descnbed bel .
Step 1. The ecoregion means, calculated in the Level I water quality
assessment, were correlated with detSed land use and topography
information to test the relationship between these ecoregion charac-
teristics and water quality Two types of correlation coefficient were
calculated the Pearson product moment correlation (a parametric
measure) and KendaJi ’s tau-b (a nonparametnc measure). Although
the water quality values themselves usually are not normally distnb-
uted, the ecoregion means that we analyzed are likely to be
normally distnbuted HoNever because there are only seven obser-
vat ons (ecoregions), a normal distribution of the data may still be
questionable For this reason, both Pearson and Kendall’s correla-
tions were calculated, and a relationship was considered significant
only if it was significant for both measures.
Because land resources data represented the period 1965-7a MPC*
s arnined relationships between ecoregion ch eristics arid water
quality parameters for the same time period This procedure resulted
in 34 relationships where either the water quality mean, median, or
both were significantly correlated with an ecoregion characteristic.
For aarnple, total suspended solidswere correlated to stream onen-
tation, and total phosphorus levels were correlated to the percent of
land in agnculture
Step 2. Not all of the 34 relalionships identified are applicable or
appropnate for an NPS assessment of minor watersheds. Some fac-
tors used to develop the relationships are not provided at the neces-
sary level of detail (e.g , Department of Agriculture statistical data are
recorded by county and minor watersheds are smaller than coun-
ties) Other factors are related to water quality parameters that may
not pose a senous threat to the aquatic environment at levels typi-
cally observed in Minnesota With these limitations in mind, MPCA
has made a preliminary selection of nine factors or “predictors” of
NPS pollution potential These predictors are’ (1) percent of land in
cultivation, (2) percent of land that is urbanized, (3) percent of land
with forest cover, (4) sift sods, (5) sandy soils, (6) stream onentation,
(7) lake orientation, (8) slopes of 3106%, and (9) slopes greater than
6% The objective is to map the pollution potential of each minor
watershed based on these nine predictors.. The selection of these
predictors and the Level II assessment will help MPQA meet require-
ments under Section 319 of the Clean Water Act Amendments.
the nine predictors (individual scores for stream and lake orientation
were doubled to reflect the importance of these factors) The
watershed with the highest score was considered to have the greatest
NPS pollution potential. Finally, the scores were displayed graphi-
cally by mapping the percentile scores of each minor watershed.
It should be noted that the map produced by this exercise does not
Identify waterbodies with N f l pollution, but rather identifies areas
where such problems are likely—la, where watershed conditions
encourage NPS pollution in the absence of proper land use manage-
rnent. The MPCA hopes to use this procedure to identify areas bath
where further study is required and where protective measures may
be needed.
LEVEL.lll: SITE-SPECIFIC NONPOINT SOURCE ASSESSMENT
Under the Minnesota Clean Water Partnership Mt, enacted in May
1987 the MPCØL will fund up to 50% of the costs for both diagnostic
studies (to assess the need for site-specific NPS controls) and the
Implementation of such prbjecs Level Ill NPS assessments will be
undertaken pnmanly by local authorities (e.g., counties, municipalities,
or watershed distr icts) in conjunction with prqects to be funded under
this State program. Level Ill studies will include on-site water quality
monitoring and the use of analytical tools such as the Minnesota
Feedkl Model and the Agricultural Nonpoint Source Model (AGNPS)
developed cooperatively by MPG I and by the US. Department of
Agriculture.
The feedlot model may be used to estimate the mass loading and
concentrations of nitrogen, phosphorus, and chemical acygen
demand discharged from a feedlot into receiving waters after a storm
These computations can be performed manually on a hand-held cal-
culator. It the calculated discharge eceeds State standards for these
pollutants, a numerical rating based on the potential seventy of the
hazard Is assigned to allow comparison and ranking with other
feed l e
AGNPS is a computer simulation model (for use on personal and
mainframe computers) developed to analyze the quality of runoff from
Minnesota watersheds. The model predicts runoff volume and peak
rate; eroded and delivered sediment; and nitrogen, phosphorus, and
chemical ocygen demand concentration in runoff for single storm
event AGNPS is formulated on a gnd-cell basis, so that the model
can provide water quality information for specific locations within each
watershed. Most of the information required for model input is avail-
able from the Minnesota Land Management lnformation Center
Within Minnesota, AGNPS has been utilized in the Big Stone Lake
restoration project to identity critical areas in two sub-watersheds
where upland erosion and runoff resulted in high concentrations of
sediment and nutnerrts at the outlets of these watersheds. Once these
critical areas were identified, Soil Conservation Service personnel
conducted on-site inspections and recommended best management
practices to reduce specific sources of pollution
PROGRAM SUMMARY
Level I of the nonpoint source assessment revealed NPS prqblems
throughout Minnesota As a result, the State developed the Clean
Water Partnership program to address this problem Through this pro-
gram, MPCI4 will administer both State and Federal nonpoint source
control funds, and Levels II and Ill of the assessment will help direct
resources to areas where the most benefit will be obtained
Material for This report was furnished by Gary Fandtei and Gay/en
Ree , Minnesota Pollution ConUof Agency Division of Water Qua! ,-
ty- Robeit )bung, US Department of Agriculture, Agricu l tural
Research Service; and Thomas Davenport, US EPA, Region 5
For addrtional information, contact Mr Fandre , (612) 296-7363
TESTING THE LEVEL II PROCEDURE
The Westem Corn Ben Plans ecoregion was used to test the MPCdi5
proposed Level I I NPS assessment procedure. First, the nine predic-
tors were rank ordered for each of the 1238 minor watersheds in this
ecoregion Second, the pollution potential score for each minor
wpterthed was calculated by summinc the rank scores for each of
This report is produced by E P A Ic h ,ghlqhf monitoring and wastelbad
allocation activities. Contr ibutions tnt rmaton for similar r eports a r e
invited Ptease contact E F Drabkowsffi, EPA, MOSO, WH 553,
401 M Street SW, Vstishington, D C 2046Q (20Z1 382-7056
-------
4 ’
United States
Environmental
Protection Agency
Assessment and Watershed Protection Division
Office of Water
Washington, DC 20460
May 1990
Water Quality Program Highlights
w w W W W W -
— a — a a — S
_ w w w w
Ohio EPA’s Use of Biological Survey Information
BACKGROUND
The Federal Clean Water Act has been amended numerous times
since it was completely rewritten in 1972, but its pnncipal goal remains
unchanged to restore and maintain the physical, chemical, and bio-
logical integrity of our surface waters To achieve this environmental
goal, we — EPA, the States, the regulated community, and the
public — need information about the aquatic environment We need
information to help us develop meaningful yet workable water qual-
ity goals, wisely direct our limited resources to those waters that will
benef it most from restoration or Control efforts, and ensure that our
efforts result in measurable environmental improvements
EPA and State monitoring programs have historically relied on one
type of information — measurements of individual pollutants in the
water column Only slowly have chemical analyses of the water
column been supplemented by other monitoring methods There
seems to be a consensus today that to manage our remaining water
quality probtems cost-effectively, we must employ a vanety of methods
capable of assessing the impact of chemicals in tissue and sediment
as well as in the water column, the condition of the physical habitat
as well as water column chemistry, and the response of resident biota
as well as laboratory test species
HIGHLIGHTS OF OHIO EPA’S MONITORING PROGRAMS
The Ohio Environmental Protection Agency (Ohio EPA) uses a com-
bination of chemical, toxicological, and ecological approaches to
monitor the quality of its rivers and streams This Highlight focuses
on Ohio EPA’s Biological and Water Quality Survey (BWOS) Program,
and briefly discusses Ohio’s tong-term ambient water quality monitor-
ing network (NAWOMN) Both programs make use of integrated
chemical and biological monitoring In the early 1980s, the highest
priority in the BWQS was evaluating the need for publicly owned treat-
ment works (P01W) to install advanced treatment, the current pri-
ority is to evaluate nonpoint sources and assess toxicity due to point
sources The main use of the NAWQMN data is to evaluate the effec-
tiveness of selected pot lution control projects
Because the biological component of Ohio EPA’s programs is likely
to be of greatest interest, this Program Highlight focuses on the poten-
tial uses, advantages, and limitations of the biological survey infor-
mation collected in these two programs
Ohio EPA has found that incorporating biological survey methods into
its water quality assessment program produces several benefits com-
pared with relying exclusively on chemical-by-chemical or whole
effluent toxicity monitoring First, biological assessments can detect
water quality problems that other methods might miss or underesti-
mate The resident biota act as continuous monitors of environmen-
tal quality, increasing the likelihood of detecting the effects of episodic
events (e g , spills, nonpoint sources) or other highly variable impacts
that monthly or even weekly chemical sampling might miss And sam-
pling need not be conducted at critical low flow or under other worst
case conditions Second. biological surveys can detect problems
such as habitat degradation that are not stnctly water quality problems,
but can prevent attainment of uses Third, biological surveys direct-
ly assess biological integrity, providing information needed to iden-
tify high quality waters deserving special protection or confirm
instream impacts predicted by fate and transport modeling (eg,
wasteload allocation) and toxicity testing (i e, bioassays)
The power of biological assessments is their ability to assess aquat-
ic ecosystem health (i e , biological integrity) They can supplement,
but not replace, chemical and toxicological methods that are neces-
sary to predict risks (particularly to human health and wildlife) and
to diagnose, model, and regulate problems once they are detected
USES OF THE DATA
Three major uses of the chemical and biological data derived from
the BWQS and NAWOMN Programs have included
• improving water quality standards (including refinement of stream
use classifications and development of biological criteria),
• identifying impaired waters and assessing attainmentinonattain-
merit of beneficial uses
• evaluating the effectiveness of pollution controls
Use #1- Improving Water Quality Standards
A major use of BWQS data has been to improve Ohio’s water quality
standards by refining existing use classifications and developing
numeric biological criteria to supplement existing chemical-specific
and toxicity criteria The development of biological criteria required
descriptions of the type and condition of aquatic life thought attain-
able in streams and rivers throughout the State Ohio recognized that
biological critena needed to account for intrastate differences in attain-
able quality due to regional variation in land surface form, land use,
vegetation, soils, and climate, but realized that it was infeasible to
develop site-specific criteria for each of the hundreds of waterbod-
es in the State Their solution was to monitor streams least affected
by human activity in each of several regions of the State ( ‘least dis-
turbed streams”) and analyze the data to establish criteria specify-
ing attainable conditions within each region
Ohio EPA could not have developed biological criteria without first
developing standardized biological assessment methods Ohio has
accomplished both — the development of assessment methods and
criteria — through an iterative process of monitoring, the development
of initial criteria, additional monitoring, and the subsequent develop-
ment of more rigorous criteria Ohio was fortunate to have a fairly
extensive historical database dating back as far as 1979
The process began in 1980, when Ohio EPA used the available data-
base of about 150 sampling locations and the experience of its biol-
ogists to develop biological criteria for two aquatic life uses
(exceptional warmwater habitat and warmwater habitat) These early
criteria included both narrative and numeric requirements (eg a
stream met the exceptional warmwater habitat use only if there were
more than 30 taxa present and pollution-sensitive species were
“abundant”)
The process continued in 1983 and 1984, when Ohio EPA and
USEPA’s Environmental Research Laboratory in Corvallis, Oregon,
carried Out the Stream Regionalization Project The project involved
delineating the five distinct ecological regions (‘ecoregions”) illus-
trated in Figure 1, identifying ‘least-disturbed” watersheds in each
ecoregion, and conducting extensive field work to characterize the
health of fish and macroinvertebrate communities (and water qual-
ity) in the least-disturbed watersheds Ohio chose fish and macro-
invertebrates as its indicators of biological integrity because the
-------
interior Plateau
IP)
H ron/Erie Lake
Plain (HELP)
Eastern Corn Belt
Plains ECBP)
Western Alie-
ghenv Plateau
(WAP)
Erie/Ontario Lake
Pain (EOLP)
distribution, environmental tolerance. and importance of these com-
munitIes in lotic ecosystems were well known and because their
health also reflects the health of lower trophic groups. More than 250
small stream sites and about 100 large river Sites were sampled.
These reference sites represent roughly the highest quality 5 percent
of stream and river habitats in the State.
The field measurements were analyzed to determine various met-
rics of the health of the fish and macroinvertebrate communities such
as species richness. trophic composition. diversity, the presence of
pollution-tolerant individuals or species. abundance or biomass, and
the presence of diseased or abnormal individuals. These metrics were
in turn used to calculate values of three different biological indices:
the Index of Biotic Integrity (181) for fish. the Modified Index of Well
Being (lwb) for fish, and the Invertebrate Community Index (ICI) for
macroinvertebrates.
The next step in the analysis was to select values of each index
thought attainable for each ecoregion. Figure 2 illustrates in a box
and whisker’ plot the analysis conducted for one of the three indices:
the IBI. The plot shows the distribution of IBI values calculated for
least-disturbed streams in each of the five ecoregions. In the
Huron/Erie Lake Plain ecoregion. for instance, lBl values for least-
disturbed streams varied between 24 and 30. Ohio EPA established
that a stream surpassing the 25th percentile value of the fBI scores
of the reference streams in its ecoregion has attained the warmwater
habitat use. in this case an IBI of 26. Ohio EPA established that a
stream surpassing the 75th percentile value of the entire statewide
reference site data set has attained the exceptional warmwater habitat
use. These values serve as Ohio’s numeric biological criteria. Gener-
ally, a waterbody is reported to fully attain its use only if all three index
Note: See Figure 1 tar definition of acronyms
scores (161. Iwb. and CI) surpass the ecoregional criteria. Ohio reports
partial use attainment if only one or two index values are met and
nonattainment if none of the indices meet applicable criteria or if one
organism group indicates poor or very poor performance.
Ohio has now established reference values for each of its three bio-
logical indices for each of the five ecoregions in three of its five aquatic
life use categories. In addition. because attainable fish community
characteristics vary with stream size and sampling method. reference
values have been established separately for headwater streams
(streams with drainage areas less than 20 mi /. nonheadwater
streams sampled by wading (drainage areas uetween 20 and 500
mi 2 ), and streams and rivers sampled by boat (drainage areas
between 200 and 6.000 mi 2 ).
The five aquatic life uses included in Ohio EPA’s refined water qual-
ity standards are: warmwater habitat, exceptional warmwater habitat.
modified warmwater habitat, coldwater habitat, and seasonal
salmonid habitat. Warmwater habitat is designated where waters are
believed capable of supporting balanced reproducing populations
of warmwater fish and associated organisms; exceptional warmwater
habitat is designated where more sensitive and diverse biological
communities. or rare species. are possible: coldwater habitat is desig-
nated in waters capable of supporting coidwater fish and associat-
ed organisms or where sal monids are regulartv stocked: and seasonal
salmonid habitat app) ies between October and May in tributaries to
Lake Erie used by migrating salmonids The modified warmwater
habitat use designation is intermediate between the existing warm-
.‘iater habitat and limited resource water categories. Limited resource
waters are those that have extremely limited physical habitats due
to natural Imitations or extreme alterations of anthropogenic origin.
The modified use was adopted after integrated assessments identi-
fied a number of stream segments where irreversible impacts pre-
cluded the attainment of the warmwater habitat use. but documented
that these segments were able to sustain a semblance of a warm-
water biological community. A use attainability analysis and USEPA
approval are required prior to designating a stream as a modified
warmwater habitat. There are, in addition. designations for aesthet-
ics, water supply, and recreational uses, but the aquatic life use desig-
nations generally have the more stringent chemical criteria.
Ohio devoted a substantial fraction of its monitoring resources for 10
consecutive years to improving its water quality standards. Ohio
expects in future years to sample about 100/o of the reference sites
each year to detect any broad-scale changes in background condi-
tions that might prompt a recalibration of the biological indices. revi-
sions of the biological criteria, or both.
Ohio EPA’s approach to developing biological criteria is but one of
several approaches used by State water quality agencies to define
and measure achievement of biological integrity. States may choose
to conduct crash efforts and monitor reference sites statewide in a
year or two, or follow Ohio’s example and spread the sampling over
a 5- to 10-year period. The level of effort required to develop criteria
varies from State to State—more ecologically homogeneous and
sparsely populated States might find tens of reference sites sufficient:
more heterogeneous and densely populated States might need more
than the 300 or so sites monitored in Ohio. See the Proceedings of
the First National Workshop on Biological Criteria (December 1988)
for a discussion of other approaches.
Use 2: Identifying Impaired Waters
Biological assessments offer a powerful tool for identifying waters
degraded by sources and causes of impairment that other
approaches are likely to miss. In the middle segment of the Little
Cuyahoga River. for example, fish and macroinvertebrate sampling
indicated severe, but unexpected. impacts indicative of toxicity. These
findings were unexpected because point source dischargers in the
segment claimed to discharge only noncontact cooling water and
small quantities of sanitary wastes. Accordingly, their permit did not
require monitoring for toxic pollutants.
A followup investigation revealed that most of the dischargers in the
river basin were involved in plastic and rubber manufacturing and
therefore handled organic chemical products on the premises. Ohio
EPA plans further work to identify the source of toxicity reaching the
II
.
L ii
Figure 1. Ohio’s Five Ecoregions.
HELP iP
EOLP WAP ECBP
ECOREGION
Figure 2. Notched Box-and-Whisker Plot of Reference Site
Results for the IBI (headwater streams).
-------
stream (e.g. spills, contaminated surface runoff, sewer system over-
flows. or unauthorized discharges).
In many other situations as well. Ohio EPA’s increased reliance on
biological methods has improved its ability to detect instream impacts
(see Figure 3). The results of a survey of 431 stream segments found
that instream chemical analyses for conventional pollutants. NH 3 ,
and five heavy metals were in agreement with biosurvey results at
58°/a of the sites (at 17°/a of the sites both methods showed no impair-
ment: at 41% of the sites both methods showed an impairment). At
6% of the sites, chemical data implied that there was an Impairment
while the instream biota showed no impairment. The most interest-
ing finding, however, was that at 360/o of the sites. instream chem-
ical data implied no impairment while the instream biological
communities showed impairment. The waters in this last category
were degraded by “nonchemical” causes including sedimentation
and/or habitat degradation i43°’b). subtle enrichment/dissolved
oxygen impacts (31/n). unknown toxicity (7°/o), and other causes
(19° a).
chemistry Implies
No im ct:
B4ota Show Im ct
(36%)
chemistry Implies impscl;
Bio Show No Impact
(6%)
Chemistry and
Biosurvey ResJts
Agree
(58%)
Figure 3. Biosurvey Results Usually Agree with Instream
Chemistry or Reveal Unknown Problems.
An interesting consequence of Ohio EPA’s improved abilityto detect
water quality problems is that the percentage of rivers and streams
reported to fully support designated uses decreased between 1986
and 1988 from 61°n to 32°/o. Ohio attributes this change to the adop-
tion of revised biological criteria and more sensitive assessment
methods, not to changes in water quality. Most of the waters newly
designated in the 1988 §305(b) report as not supporting their uses
experience “slight” to minor” impairment, lending weight to Ohio’s
assertion that integrated assessments are capable of detecting
increasingly subtle impacts.
Use #3: Effectiveness Evaluation
Ohio EPA also monitors a network of 36 NAWQMN sites to evaluate
the effectiveness of selected projects. Each year. 10 of these sites are
assessed for macroinverlebrate community health. When plotted ver-
sus time (which Ohio did for 11 rivers in its 1988 §305(b) report), the
trends in ICI values from these sites present a meaningful indicator
of environmental improvement. Where intensive survey data are also
available to interpret observed trends (e.g., to correlate trends with
program actions), these plots provide a measure of program success.
Figure 4 shows the results for two of the 36 sites, on the Mohican and
Olentangy Rivers. At the Mohican site, four samples were collected
between 1977 and 1987. Macroinvertebrate sampling shows an
improving trend in biological condition since 1978. Warmwater habitat
communities have been present in all years. with the most recent data
suggesting that the site has the potential to achieve the exceptional
warmwater habitat use. Ohio EPA attributes these improvements to
industrial waste pretreatment requirements imposed in upstream
cities and wastewater treatment improvements made by several in-
dustrial and municipal treatment plants.
At the Olentangy site, five samples were collected between 1977 and
1986. Biological condition at the site has steadily improved through
this period. The most dramatic increase in ICI values occurred
Figure 4. Long-term Trend of the Invertebrate Community
Index (Id) at Ohio EPA Annual Monitoring Stations.
between 1977 and 1979 when an ICI value ref lectina nonattainment
of the warmwater habitat aquatic life use improved to an CI value
reflecting full attainment. The CI improved further in 1983 when the
macroinvertebrate community was scored exceptional or near excep-
tional, but this improvement may have been an artifact of moving the
sampling location upstream. Construction of a new advanced treat-
ment plant and improvements at several existing plants are the most
likely explanations of the overall improvement.
DESIGN CONSIDERATIONS RELEVANT TO
ALL USES OF MONITORING DATA
Ohio EPA conducts 12 to 15 intensive surveys per year during the
June to October sampling season. Intensive surveys can be as short
as 1 week, but usually last several months They include chemical
analyses of samples collected at between 3 and 80 sites at frequen-
cies ranging from three times during the survey to once each week.
They typically include fish sampling at each site between one and
three times during the survey. Macroinvertebrate sampling is typically
conducted at between 6 and 80 sites. with artificial samplers remain-
ing instream for 6 weeks at each site sampled.
Where monitoring data are needed to calibrate and validate water
quality models, more intensive sampling is done for selected physi-
cal/chemical parameters.
Ohio’s biocriteria describe attainable conditions. In addition, one or
more reference sites located upstream of all known sources of pol-
lution are typically sampled to sort out the effects of multiple discharg-
ers, but not as an arbiter of attainable condition. In a typical point
source evaluation, one site is located upstream from the outfall,
another site is located within the mixing zone, and additional sites
are located at intervals downstream to determine the extent and
severity of impact.
Fish Sampling Methods
The Ohio EPA has developed and documented standardized proce-
dures for fish sampling. Pulsed DC electrofishing is used to obtain
a representative sample of the fish community, either by wading into
the stream or using a boat, depending on the size of the waterbody.
In a survey, field personnel conduct repetitive sampling based on dis-
tance (rather than time) to avoid bias that would result where fish differ
in spatial distribution due to differences in available habitat. Field per-
sonnel also weigh fish, identify each fish to the species level, and
record external abnormalities. A three-person crew is required. Anal-
ysis of data collected attest sites indicates that spatial and temporal
variability are low if standardized procedures are followed.
MOHICAN RIVER (16.4)
C 3°
20
0— ’ —.
73 75 77 79 81 83 85 87
YEAR
OLENTANGY RIVER (19.6)
i :i E .:... :
20 /
a
73 75 77 79 81 83 85 87
YEAR
-------
Macroinvertebrate Sampling Methods
Ohio relies primarily on a modified Hester-Dendy multiplate artificial
substrate sampler for quantitative sampling of macroinvertebrates in
streams and rivers The Ohio EPA uses a composite set of five sam-
plers, supplemented with a qualitative sample from the natural sub-
strate that provides a more complete inventory of all taxa present The
Ohio EPA prefers artificial substrate samplers because they work in
locations that cannot be sampled effectively by other means, require
lower operator skill requirements, are nondestructive to the environ-
ment, and reduce the influence of the natural substrate Results col-
lected over the past 15 years confirm that sampling variability is low
1 there is strict adherence to standardized procedures
Chemical Analyses
Ambient waler samples (usually grab samples) are collected during
integrated surveys These samples are analyzed for dissolved oxy-
sen, nutrients, solids, oil and grease, total organic carbon (TOC).
methylene blue activated substances (MBAS), fluoride, organics, met-
als, pesticides. cyanides and phenols, as appropriate Effluent and
sediment samples are collected as necessary
Quality Assurance
Quality assurance is of paramount importance to the Ohio BWOS
Program In September 1989. the Ohio EPA published an updated
version of Biological Criteria for the Protection of Aquatic Life This
document published in three volumes, details all aspects of sam-
pie collection and analysis of biological samples including
• minimum staff training ri sample collection and species identifi-
cation needed to ensure adequate data quality.
• methods for selecting and evaluating sampling sites.
• sampling procedures including the design and use of sampling
equipment species identification, field counting and weighing
procedures. sample preservation, and “chain-of-custody”
procedures,
• habitat evaluation procedures.
• laboratory procedures for handling and identifying preserved
specimens,
• data management and storage procedures,
• data analysis methods (including statistical tests and calculation
of metrics)
Ohio EPA uses the USEPA’s S1DRET database to manage its chem-
ical data and its own Fish Information System (FINS) and Macro-
invertebrate Data Gathering and Evaluation System (MIDGES) for its
biological community data Personal computers are used extensively
to analyze data and prepare reports
PROGRAM RESOURCE REQUIREMENTS
Out of an estimated 52 workyears available for Ohio EPA’s monitor-
ing activities in FY88 (field sampling, field and laboratory analyses
data analyses, and reporting), 95 workyears (184%) went toward
BWQS (surveys). 02 workyears (03%) went toward the biological por-
lion of Ohio EPA’s NAWQMN, and 08 workyears (1 6Vo) went toward
the chemical portion of the NAWQMN The remaining monitoring pro-
gram resources went toward wasteload allocation modeling/permit-
ting, toxic contaminant monitoring, compliance/enforcement
moriitonng, water quality criteria, §4.01 certifications, and other needs
Collection of biological data has the reputation of being resource-
intensive and too costly for routine application in State monitoring pro-
grams In 1989, Ohio EPA compared the cost of different approaches
assumed to provide the same analytical and evaluative ‘power” Ohio
EPA believes that, on a per-site basis, sampling fish and macro-
invertebrate communities can be equal to or lower in cost than chem-
ical sampling or toxicity testing More comprehensive chemical
monitoring, such as priority pollutant scans and sediment analyses.
further increases costs for chemical data
THE FUTURE
Improving the ability of its monitoring program to produce the type
of monitoring information needed to support water quality program
decisions has, in turn increased the demand for Ohio EPA’s biolog-
ical monitoring resources Managers of Ohio EPA’s permits nonpoint
source, hazardous waste and other environmental programs now
compete for Iimiied monitoring resources In its most recent 5-year
monitoring strategy (Ohio EPA 1985), Ohio estimated that at current
staffing levels it would take 13 years to satisfy its outstanding monitor-
ing needs
Material for this report was furnished primarily by Chris Vode Ohio
EPA Figures 1 and 3 were prepared using data from Ohio EPA’S 1988
§305(b) report Figures 2 and 4 were taken from Biological Criteria
for Protection of Aquatic Life. February 28, 1988 Ohio EPA. Division
of Water Quality Monitoring and Assessment, Surface Water Section
Columbus. Ohio For more information, contact Chris Yoder
This report is produced by EPA to highlight EPA and Stare monitoring
activities Contributions of information for similar reports are invited
Please contact Monitoring Branch, EPA AWPD WH-553 401 M Street
SW Washington, DC 20460 (202) 382- 7056
-------
United States
Erwironmental
Protection Agency
Assessment and Watershed Protection Division
Office of Water
Washington, DC 20460
March 1991
Water Quality Program Highlights
w w w w
_‘- r r r - - - -r
w w w w —
w
Multimedia Toxics Study of the Calcasieu River Estuary, Louisiana
The Water Quality Act of 1987 reemphasizes the need to control
toxicpollutants from point and nonpoint sources. In particular. Sec-
tion 304(l) of the Act requires States and EPA Regions to identify
waterbodies impacted by toxicpollutants; where known toxic prob-
ems exist due to point source discharges, Section 304(l) requires
development and implementation of individual control strategies
(ICSs)for dischargers. States are encouraged to collect additional
information to characterize the nature and severity of these impacts.
BACKGROUND
The Calcasieu River is acomplex estuarine ecosystem that receives
discharges from more than 10 industrial facilities, primarily petro-
chemical plants (Figure 1). Based on results of earlier studies and
under provisions of Section 304(l), the Louisiana Department of
Environmental Quality (LDEQ) designated portions of the Lower
Calcasieu River and one of its tributaries (Bayou d’lnde) as
waterbodies with toxics-related water quality problems. Earlier
studies identified individual comoonents of the water quality prob-
lem but did not provide a complete perspective of the sources, ex-
tent, or magnitude of water and sediment degradation in the entire
system. Tofully characterizethe geographic extent and magnitude
of the toxics problem, LDEQ and EPA Region 6 initiated a mu time-
dia toxics study in June 1988 in conjunction with the EPA Environ-
mental Research Laboratory at Narragansett, RI (ERL-N), the EPA
Houston laboratory, and the U.S. Geological Survey. Extensive
concurrent sampling of effluent, ambient water, and sediment was
conducted. These media were evaluated using physical/chemical
analyses methods and toxicological testing methods employing
multiple test species. The 1988 Calcasieu River toxics study repre-
sents a type of multimedia, multimethod approachthat has become
possible only recently.
MULTIMEDIA, MULTIMETHOD APPROACH ADVANTAGES
Based on its
Figure 1. • experience in
Calcasseu the Calcasieu
4 River River (and in a
, similar study in
—s. the Hou ston
Ship Channel),
EPARegion6believes
that this monitoring ap-
Ba proach has several fea-
dinde - tures that make it desirable
for examining large, complex
- aquatic systems. The approach
can be used for(1 )estimating whole-
effluent toxicity and chemical-specific
concentrations for individualdischargers, (2)
determining the geographic extent of ambi-
ent toxicity and concentrations of priority pol-
.‘ \ ) lutants in receiving water and sediments, (3)
assessing the contribution of individual
\\j dischargers to ambient toxicity, (4) car-
relating chemical-specific concentra-
tions to ambient toxicity, and (5) as-
‘ / sessing the contribution of sediment
• . - - toxicity to use impairment. Because of
l\ \ the high resource requirements, the
(I 1’ . \ use of this approach is feasible only
‘ \ in the most complex and critical of
situations.
MONITORING APPROACH
The 1988 toxics study was designed to synoptically sample efflu-
ents, receiving waters, and sediments for both chronic toxicity and
specificchemicalcontaminants. Water and sediment sampling was
conducted at 38 stations, and effluent sampling was conducted at
15 industrial outf aIls throughout the Calcasieu system (Figure 1).
Monitoring activities were conducted during two 1 -week intervals
in a warm weather low-flow period. Effluents, receiving waters, and
sediments for chemical-specific analyses were collected one time
per location and shipped to the laboratory for appropriate chemical
analyses. Some followup sedirnenttesting was also conducted for
Bayou dlnde in April 1989 and July 1990. Effluents and receiving
watersto betestedfortoxicity were sampled threetimes during each
week and sediments were sampled once at each station; ailsamples
were shipped overnight to the laboratory.
Chemical Analyses—Chemical-specific analyses included the
EPA-designated priority pollutant organic chemicals (extractable
organics, volatile organics, and total phenolics) and selected toxic
metals (aluminum, arsenic, cadmium, chromium, copper, iron, lead,
manganese, mercury, nickel, and zinc). In addition, ammonia, al-
kalinity. hardness, chloride, turbidity, total suspended solids, total
dissolved solids, total organic carbon (TOC), and suit ides were also
assessed for ambient water and effluents. TOG was also analyzed
for sediments.
ToxIcity Testlng —Eff luents. receiving water, and sediment were
tested using a battery of standard EPA bioassay tests. In addition
to estimating toxicity, Region 6 was also interested in evaluating the
sensitivity of EPA toxicity methods for aquatic life in an estuarine
system. Chronic testing was conducted using four test species ex-
posed in static renewal systems:
• Sheepshead minnow (Cyprinodon variegatus)— Larvae (<24 h
old) were exposed for 7 days to effluent (LDEQ).
• Mysid shrimp (Mysidopsis bahia)—Juveniles (7 days old) were
exposed for 7 days to effluent or ambient water (ERL-N).
• Inland silverside (Menklia beryilina)— Larvae (7-9 days old) were
exposed for 7daysto effluent or ambient water (ERL-N).
• Benthic amphipod (Ampelisca abdita)—Juveniles were ex-
posed for lOdaysto sediment (ERL-N).
Growth and survival were the biological endpoints measured for all
water and effluent tests; survival was measured in sediment tests.
STUDY RESULTS—BAYOU VERDINE
Results of chemical analyses and toxicity testing are summarized
in Figure 2for Bayou Verdine, the most deg raded portion of the es-
tuary. Results for each discharger are shown attheiopof the figure,
with the position of the outfall in relation to the ambient station indi-
cated by an arrow. Foreach effluent, onlythosepollutants exceeding
EPA criteria and/or State standards are identified. Exceed ances of
ambient criteria by dischargers for this bayou are based on simple
dilutioncalculations; aO% upstream dilution was assumed and tidal
dilution was assumed to be negligible. The most sensitive biological
endpoint, chronic value (ChV), is reported as percent effluent con-
centration for each of the three species tested. The ChV is an esti-
mate of the presumably safe or no-effect concentration lying
between the no-observed-effect concentration (NOEC) and the
lowest-observed-effect-concentration (LOEC). The ChV is derived
by calculating the geometric mean of the NOEC and LOEC values.
-------
co,m I
Ct .sA!, LCD I
—a
C
0C2 nip i .P, nA n. E
tJ! !t TflA 7 (C !V) A tflk .,A
aa I e A L. :0
C enA .2A
2O
Beyou VeTdine
C. EPA LADPC
0 —EPA
E . EPA Sure PAn! PAWn!
C. OAA i• DI n rA . ,
I. !, , WS• —— DanA
I — Sc. AflS !
I . EPA rAre, ,, . .
ACE - Dare..
*.EPACA ,ere
V . EPA . 0 .. ,,a,P . t a . C O D a
Sqo.rAp , A . !DI iAOp A., 00 COD
- P C -. C O O
21 22
4t Pc0004 I
rik _ . Cl .A LCD 00 !
t o COOl
E0%0 ‘ DIADo AIPA. ,
‘a e . ‘I ,
C, 5C,0 5,. , , . !0O
1V23
CaIe e ii. o
Rip.,
PCI
WOOtnOOp API
WsCOeis .p LCD
AVDVCI A!
AIDCA A
WA.C 1e , PCI
Ar re . . A
AACC £
* C , .A ACE
Afl?D !1
AACA A
A
A ,..ei A
CD
tn. , CD
Cn n
A
C CO
an.. CO
tACO. CO
M . V 0104 C
MLAAnS
C
W ,Aar .SA C
Liçei.
Ma,r AA
PA..
C a l CDCI
t IC Ca
An .D C
too,
P000 C
a OprerA, !.,. I
‘ S t .. C
I A OAADCrW.
I 2 VS i0 .onIlr. C
‘A CAre .’C
A
ii COACADP.la . I
Ppr. £
iS OA ,npn,., .,
Aa aoAIwfl
AIVDXW £
n. e. T.ny
A Ir ,.Aa C
py, . ,. C
CA. TDaEA
en P0.007
en. !., o ,p
LA bASS ‘C lIDDI
LA OniAp. 00I PC
CO
‘ ,D ,q ’ al Di
U ytnI !EA.q .OSDD,
0 DIPS SOAP vein!,
AADC ‘S D VWc.l On!!!
C DSS 00% neon,
Al alnffOUP ,
as p sp C pa,
U A D , !tS an verse,
C tCW////////// Y////J//////i///J///A
,-n.— AACOCS ODE
#////////7////7// i/// MVñY/// //A
bAdDa.., çCOT All
/4’/il// 7J //1//il7///////AW#/////.I
ISOO It AC!
5 ADInb CCOIC, AC!
Ctq,J,///,J////////7////////M
e, 0C 5 , ! PCI
PA*PCADaA
AICOnI YnO D p
D P AC O D I I
S. P0, 10*,
P !.PCI IDC , . I
110.00,
P!DACCDCflD I
IOPCO reID,
CA .WDTVD ’D C
D 00• ’ !ACDy
A COO 00% non.
A A *9% v e .AA e,
A 00% VDIDISPp
A COt 00% POPSI ! ,
A COS 00% ret. .!
Figure 2. Chemical exceedances and toxicity results for Bayou Verdine.
Below the sampling station line, chemical oxceedances and toxic-
ity results are summarized forwater and sediment.
Effluent — Chemical-specific data showed high concentrations of
zinc in Vista 001 with lesser concentrations found in PPG 004 and
Conoco 001. Zinc loadings wiire calculated to be37, 5, and 1 lb/day
for Vista 001, PPG 004, and Conocio 001. respectively Nickel and
copper appear to be contributed primarily by Vista 001. Vista 001
and Conoco 001 effluents were highlytoxic, resulting in ChVs of <5%
effluent for mysid shrimp and <5% 1015.2% effluent for silversides.
Compared to mysids and silversides, the sheepshead minnow was
relatively insensitive to these same effluents, exhibiting ChVs of <25
to >100% effluent
AmbIent Waler — Water quality in Bayou Verdine was degraded
by ammonia, several heavy metals, and organics. EPAcriteria and/
or State standards were exceeded at all five stations for arsenic,
copper, manganese, nickel, and 1 ,2-dichloroethane and at several
stations for zinc, ammonia, pyrene, and anthracene. Particularly
noteworthy are those chemicals found in exceedance of EPA water
quality criteria for protection of aquatic life that may contribute to
ambientwatertoxicity Nickel, copper, zinc, and ammoniaexceeded
EPA chronic marine criteria; copper, zinc, and ammonia exceeded
EPA acute marine criteria at some stations. Several of these pollut-
ants were detected in effluent of the bayou’s three dischargers Ex-
posure to ambient water from stations 19, 20, and 21 produced a
significant toxic response in at least one test species. Mysid shnmp
exhibited the highest mortality—from 59to 100%.
Sediment Quafity — Sediment quality of Bayou Verdine was de-
graded by organic and metal pollutants. At four stations, the EPA
interim manne sediment criterion was exceeded for phenanthrene.
Currently there are no sediment criteria for metals. However, chro-
mium, copper, lead, manganese, and zinc concentrations great9r
than 50 mg/kg were detected at several stations and a zinc con-
centration of 1,234 mg/kg was detected at station 20. Sediment
metal concentrations were generally highest (>50 mg/kg) at sta-
lions 20 or 21, decreasing downstream Chromium, copper, man-
ganese, lead, nickel, and zincwere detected in the effluents of Vista
001, and zinc and manganese were also detected in effluents from
the other dischargers Exposure to sediment sampled from all sta-
tions produced high mortality (>98%) in the benthic aniphipod.
STUDY CONCLUSIONS AND RECOMMENDA11ONS
A primary finding was the identification of ambient water and sedi-
ment contamination and toxicity in Bayou Verdine Analytical data
reveal violations of EPA acute and chronic marine watorquality cri-
teria and State chronic marine standards for zinc, EPA chronic ma-
rine waterquality criteria for nickel and copper, State human health
standards for 1 ,2-dichloroethane and EPA human health criteriafor
arsenic and manganese, toxicological data indicate ambient toxic-
ity of both receiving water and sediment at specific stations Based
on study results, EPA Region 6 included Bayou Verdine on the
304(l)(B) list and Vista Chemicals on the 304(I)(C) discharger list for
nickel and zinc. Steps to control these pollutants are being taken
through an ICS and modifications to the NPDES permit No action
has been taken concerning highly contaminated sediments.
Anotherfinding was the detection of high sediment concentrations
ot hexachlorobenzene (HCB) and hexachlorobuladiene (HCBD) in
Bayou d’Inde. The point source believed to be responsible for
widespread bioaccumulation of these two compounds in the lower
Calcasieu estuary was determined to be PPG 001. EPA Region 6
approved the State’s listing of Bayou d’Inde, a portion of the
Calcasieu River, and Prien Lake underSection 304Q)(B) PPG was
listed under Section 304(l)(C) for discharges of halogenated au-
phatic and aromatic pnority pollutants, principally HCB and HCBD
EPA added bromoform and recommended more stringent controls
for HCBD. Controls are being required through the lOS/permit
modification process. Best available technology controls forvolatile
organic compounds (VOCs) will be implemented when the permit
is reissued later in 1991 or 1992 The LDEO is negotiating with PPG
concerning cleanup of contaminated sediments
EPA Region 6 believes periodic ambient toxicity testing at selected
sites is warranted, particularly in Bayou Verdine and Bayou d’lnde
Mysids werethe most sensitive test species, followed by stiversides,
sheepshead minnows were the least sensitive The State supports
use of the silverside in monitonng and permit requirements. and EPA
Region 6 is incorporating this species in its whole effluent toxicity
NPDES permit requirements.
Heavy metals, including zinc, copper, and arsenic, were detected
in effluents discharged to the two bayous studied and the Calcasieu
mainstem and were also found at relatively high concentrations in
ambient water and sediments at some sites. Region 6 and the State
recognize the need to investigate metal loading and impacts
throughout the Calcasieu system. As a result of this study, Louisi-
ana is currently establishing standards for copper, cadmium, lead,
nickel, and mercury.
Material for this report was furnished by Philip Crocker, EPA Re-
gion 6 WaterManagement Division. The study s conduct eo’ with
the support of EPA’s AWPD and the State of Louisiana. AToxIcs
Study of the Lower Calcasieu River is available from NTIS,
#PB90226 150/AS.
This report is pmduced by EPA to highlight monitoring and
wasteloadallocation activities. Contributions of information for
similarreports are invited. Please contact EPA AWPD Moni-
toring Branch, 401 M Street, SW(WH-553), Washington, DC
20460, (202)382-7056.
-------
It
United States
E nvi ron mental
Protection Agency
A5sess ” ent ar’ Watersr,ec otec or Division
Office of Water
Washington. DC 20460
Ma 1991
Water Quality Program Highlights
Eutrophication Management in North Carolina
In North Carolina, several noteworthy nutrient management actions
have resulted from eutrophication monitoring and assessments.
including 1) development of a special classification. Nutrient Sensi-
tive Waters (NSW), that both restores and protects designated waters:
(2) development of a State water quality standard for chlorophyll a:
(3) a ban on the manufacture, sale, and se of high phosphate deter-
gents, and 141 initiation of a cost-sharing program for implementing
agricultural best management practices (BMPs) Significant nutrient
reductions have been achieved already because of the detergent ban.
Fur’ ner reductions will occur as effluent controls and BMPs continue
to e mpiemented. This report discusses some of the tools the State
nas used to develop an integrated approach to eutrophication assess-
ment and management
THE EUTROPHICATION PROBLEM
Excessive discharge of nutrients (phosphorus and nitrogen) from
wastewater treatment plants and rionpoint sources into lakes, reser-
voirs, rivers, and estuaries can cause undesirable blooms of
microscopic algae and excessive growth of larger plants (macro-
c’ esi Accelerated growth of plants frequently taxes the water’s oxy-
gen resources and disrupts the food chain. Eutrophication can
‘herefore interfere with a waterbody’s intended use, such as recrea-
tion. The objectives of managing nutrient inputs are to restore water
qLality in impaired eutrophic waters and to reduce the potential for
future problems in unimpaired waters.
ASSESSING THE PROBLEM
Eutrophication in Coastal Rivers—The earliest documented
eutrophic response to high nutrient concentrations in North Carolina
waiers occurred in the 1970s in a coastal plain waterbody. the Cho-
wan River (Figure 1). The Chowan is a major alewife and hemng nurs-
ery and fishery area and has recreational uses important to the local
economy Occurrence of massive blue-green algal blooms focused
public attention on eutrophication and resulted in an intensive water
ality monitoring effort. Biweekly monitoring of phosphorus, rtilrO-
gen. chlorophyll a a measure of algal production), phytoplankton,
and other water quality measurements was used to document the
.:ajses and severity of algal blooms in the river. int source monitor-
ing and land use data were used to develop estimates of point and
nonpoint source Contributors to nutrient loading. Subsequently,
screening models relating phosphorus to chlorophyll a were devel-
ooe to determine necessary reductions and to assess the probable
e iects of phosphorus control strategies.
During the early 1980s. surface blue-green algal blooms were also
being documented on the lower Neuse River Research supporec
DY the State resulted in estimates of desirable nutrient reductions’
for inorganic nitrogen. a 30% reduction in spring and summer con-
centrations: and for phosphorus, a 50°/o reduction Developing these
estimates involved innovative techniques using both laboratory and
fl Situ algal assays focused on the phytoplankton species responsi-
bl for the worst blooms (Microsy tis aerugirtosa). Inorganic nutrients
were added to enclosed “hydrocorrals” in the river to determine whiCh
ut!!ents were limiting phytoplankton growth. In a majority of cases.
no enhanced growth was observed indicating a hypereutrophic con-
dition where nutrient supplies exceeded growth requirements. There-
fore, a bioassay technique designed to address potential nutrient
limitation was developed by diluting ambient water to determine the
concentrations at which nutrients would become limiting to Mictocys-
tis growth. Using effluent and stream nutrient monitoring data and
land use analyses, the NC Division of Environmental Management
(DEM) determined the annual average Contributions of nitrogen and
phosphorus from point and nonpoint sources and assessed strate-
gies to accomplish the necessary reductions.
In the Tar-Par-nlico River basin, as in the Chowan and Neuse basins,
heavy nutrient loads are delivered to the free-flowing nver in the upper
and middle basin before the river slows and broadens as it enters
the Parnlico Sound. Unlike the other basins, blue-green algal blooms
and excessive summer chlorophyll a levels have not occurred in the
Tar-Pamlico basin. However, current nutnent levels, symptomatic eco-
logical signs (dinoflagellate algal blooms, fish kills, loss of submerged
aquatic vegetation), and land development patterns clearly suggest
the probability of increasing degradation. The State thus developed
a proactive strategy by requiring nutrient removal before severe signs
of eutrophication appeared rather than taking the reactive stance
required for the Chowan and Neuse basins.
Eutrophication in Piedmont Reservoirs—In addition to monitor-
ing activities in coastal plain rivers, ongoing studies indicated that por-
tions of two Piedmont reservoirs, Falls Lake (Neuse basin) and Jordan
Lake (Cape Fear basin), were impacted by heavy nutrient loading.
Falls Lake is currently the principal drinking water source for the City
of Raleigh: Jordan Lake is a planned future water supply source
Although the intensity of surface algal blooms on the reservoirs has
not matched that found on the coastal rivers, sampling at many lake
stations has indicated that a majority of summer chlorophyll a values
exceeded 40 Mg/L. An additional concern is the observed dominance
of blue-green algae.
DEVELOPING A SOWTION
A State water quality standard for chlorophyll a (40 g/L in the sum-
mer) was adopted during the Chowan study. While chlorophyll a
standards have not been adopted by other States, DEM has found
the standard to be an effective tool in nutrient management The
standard has been particularly useful in providing a decision criteria
for models relating nutrients to chlorophyll a.
The Chowan River monitoring effort prompted the State to designate
a new surface water classification—Nutrient Sensitive Waters—for
surface waters “experiencing or subfect to excessive growths of
microscopic or macroscopic vegetation.” The NSW classification
allows the DEM to develop nutnent reduction strategies for individual
waterbodies arid basins. These strategies are based on existing data
and model projections indicating use impairment in the absence of
nutrient reductions. A Water Quality Management Plan for the
Chowan basin was published in 1982
COASTAL PtAIi
—I . ,,,
-
5” i. wate’s
Figure 1. Nutrient sensitive waters in North Carolina.
-------
The Chowan Management Plan provided a foundation for nutrient
management efloils in other basins As a result of the monitoring and
algal studies described above, the Neuse River received the NSW
classification in 1988. and a schedule f r implementing nutrient con-
trots was developed The Falls Lake and Jordan Lake watersheds
were designated NSW when they were impounded in 1983 a revised
NSW strategy was adopted after a 5-year assessment in 1988 In Sep-
tember 1989 the Tar-Panilico wate shed received the NSW desig-
nation
For the two Piedmont reservoirs sufficient data were available to acact
nitrogen phosphorus and chlorophyll a loadinglresponse models
and to assess the impact of predicted population growth and changes
in wastewater inputs and land use A model developed at Duke
University for southeastern lakes and reservoirs was used to predict
the annual average lakewide phosphorus and nitrogen concentra-
tions based on nutrient loading A model developed by the Army
Corps of Engineers for large reservoirs was used to predict the sum-
mer lakewide chlorophyll a concentrations based on the predicted
phosphorus and nitrogen concentrations Strategies for phosphorus
reductions based on these two analyses were finalized in 1987 since
then OEM. in cooperation with EPA’s Ecological Support Branch in
Athens. Georgia. and the City of Durham. has completed intensive
examinations of the impact of nutrient loading from several large
municipal wastewater treatment plants on localized areas in both
lakes EPA’s Athens laboratory provided support in completing algal
tests to study the maximum potential algal growth the biological avail-
ability of nutrients, and the limiting nutrient
IMPLEMENTING A SOLUTION—
NUTRIENT MANAGEMENT EFFORTS
Eutrophication assessments have also contributed to two major
statewide nutrient management activities agricultural BMP cost-
sharing and a ban on phosphate-bearing detergents
Point source controls—End-of-pipe nutrient controls required in
NSW watersheds include limitations of 1 mg/L for phosphorus and
3 mg/L for nitrogen for all dischargers in the Chowan basin In the
other basins, a phosphorus limit of at least 2 mg/L is required for alt
significant dischargers The stringent provisions of the Chowan
Management Plan, for example. caused virtually all dischargers in
this rural basin to construct land application systems for wastewater
disposal
BMP cost-sharing—The NSW classification also triggered a State-
funded, voluntary agricultural cost-sharing program for implement-
ing BMPs for waste, water. and fertilizer management and erosion
control This program has grown from 16 counties in NSW
watersheds. funded by the State legislature at a level of about $2 mil-
lion in 1984, to statewide funding in 1990 with an annual budget of
S12 million guaranteed for 10 years About 500.000 acres of land were
treated with BMPs through the cost-sharing program between 1984
and March 1989 OEM estimates that this program could reduce phos-
phorus inputs from treated agncultural lands by roughly 300/0 and
nitrogen inputs by 15-20%, with the additional benefits of pesticide
and sediment reduction and long-term economic returns
and algal assays additional reductions are being sought througn
effluent limitations and nonpoint source Controls
Table 1 Nutrient Sensitive Waters Annual Phosphorus (Pt
and Nitrogen (N) Reduction Goals and Achievements’
Waterbody
Target P
Reduction°
Observed P
Reductionc
Target N
Reduction
Chowan River
INC portion)
35.40 0 /c
0
1 5 ,25 0 c
Lower Neuse River
500/0 or
greater
26°c
3000 50 , ngI
summer
Upper Neuse River
(Falls Lake)
430/a
21 0 /c
—
Upper Cape Fear
River (Jordan Lake)
38°’o
2 4 °’c
—
Pamlico River
e
13°’c
e
a 290 10 reduction as been documented in Chowan N target
reduction is being achieved Target compliance dates f r other
basins range from 1990 to 1993
DAdditional P reduction needs are being studied
Cp reduction from P detergent ban
0 Municipai point sources are using nondscnargirig systems
eN O increase from point sources Implement agricultural BMPs
CURRENT AND FUTURE EFFORTS
In addition to monitoring the effect of management actions on NSW
waters, OEM is currently involved in both screening and intensive
eutrophication studies of other coastal and inland waters The Norm
Carolina Lakes Program focuses on screening level assessments of
14.4 public lakes greater than 100 acres An algal bloom monitoring
network, involving citizen and DEM staff investigation of bloom events
is being used for geographic tracking of bloom occurrence The most
notable management result of this monitoring network has been the
identification of a coastal river (New River) with excessive blooms
resulting in intensive sampling and point-source phosphorus controls
Material for the report was provided by John Dorney and Jim Over-
ton of the Water Quality Section of the NorTh Carolina Division of Emit
ronmerttal Management
Detergent ban—Another malor initiative to reduce phosphorus has
been the adoption of a statewide ban on the manufacture, sale, and
use of high phosphate detergents, effective January 1, 1988 DEM
has evaluated the effect this law has had on the discharge of phos-
phorus from municipal wastewater (Table 1) Data from 23 ma;or
municipal facilities indicate that influent phosphorus loading has
decreased by 10°/a to 58°/a, and effluent phosphorus loading has
decreased by 250/0 to 85°/o Municipalities are being polled to deter-
mine differences in treatment methods. which may explain the higher
reductions in effluent phosphorus relative to influent phosphorus It
is estimated that phosphorus effluent loading has decreased by
4 million lb/year since the ban became effective In the Neuse River
Falls Lake. and Jordan Lake watersheds. this reduction is equivalent
to as much as one-quarter of the total loading However, since reduc-
tions of about 50% or more are needed based on water quality models
This report is produced by EPA to highlight monitoring arte wasteloac
allocation activities Contributions of information for similar reports are
invited Please contact EPA AWPD Monitoring Branch (WH 553) 401
M Street SW Washington DC 20460(202) 382-7013
-------
EPA
United States
Environmental
Protection Agency
Water Quality Program Highlights
Florida’s Method for Assessing Metals
Contamination in Estuarine Sediments
Among the toxicants found in sediments, heavy metals (e.g.. lead.
cadmium, mercury, nickel) have been a continuing concern. These
contaminants are frequently found in nonpoint source runoff. ard ele-
vated metals concentrations often signal the presence of other types
of pollution Metals are, however, a natural component of the coastal
environment and, in small amounts, are necessary to the existence
of marine and estuarine organisms. The EPA is currently working to
develop sediment metal criteria that address the issue of bioavail-
ability of sediment-bound metals to aquatic life. This is an important
and complex issue because high concentrations of metals in sedi-
ment in and of themselves are not necessarily toxic—the metals may
be tightly bound to sediment particles and/or organic matter here-
by reducing their availability to be metabolized by aquatic organisms.
Conversely, seemingly low concentrations of metals in some sedi-
ment types may be toxic to various aquatic organisms because the
metals are more available for metabolism
Florida’s Department of Environmental Regulation (DER) has devel-
oped a method to identify sediments with elevated metals contami-
nation using data on metals concentrations in sediments from
unimpacted areas. The method does not identify sediments that are
toxic (i.e., where benthic communities are impacted), but it permits
targeting of areas of elevated metals concentrations for further chem-
ical, toxicological, or biological assessments. This State’s approach.
which is based on natural relationships between metals in uncontami-
nated sediments, may be useful in other regions of the country where
there is a need to investigate trends, manage cleanup efforts, or
develop permits for dredging and disposal of sediments.
THEORETICAL BASIS FOR THE FLORIDA APPROACH
The background or ‘natural” concentration of metals in sediment can
vary widely among marine waters, with concentrations generally
being a function of watershed mineralogy, sediment grain size,
organic content of the sediment, and the amount contributed by
human activities, Such variability makes it difficult to define numer-
ical criteria for sediment or compare sediment metals concentrations
from different areas.
Florida’s approach to interpreting sediment data is based on natur-
ally occurring relationships between the sediment concentrations of
various metals, Specifically. geochemists have found that several
abundant crustal elements occur in a relatively constant relationship
with heavy metals. which allows researchers to “normalize” or refer-
ence various metals to a single conservative element. This method
offers the additional advantage of requiring data any on sediment
metal concentration; other methods (sediment toxicity testing. parti-
tioning. and benthic community structure) require data on numerous
parameters.
Florida DER chose aluminum as the reference element because (1)
it is highly refractory (i.e., it does not degrade or alter in form in tie
environment); (2) heavy metal/aluminum ratios are relatively constant
in unimpacted estuaries: and (3) aluminum concentrations generally
are not influenced by human activities.
DEVELOPING REFERENCE DATA
To set up their analyses the DER sampled reference stations to docu-
ment metals concentrations at 103 “unimpacted” (i.e., clean water)
sites throughout the State of Florida. Duplicate samples were col-
lected from the upper 5 cm of sediment for up to nine metals: alum-
inum. arsenic, cadmium, chromium, copper. lead. mercury. nickel.
and zinc. A total of 785 data points were obtained (all nine metals were
not measured at each site).
Of the nine metals, mercury presented unique problems because
(1) it is more volatile than the other metals; (2) natural background
concentrations of mercury were near the analytical detection limit,
where higher measurement error can be expected: and (3) an inverse
mercury/aluminum ratio was observed. As a result, mercury was
excluded from further analysis.
For the remaining eight metals, several steps were taken before
statistical analyses were attempted. To satisfy the requirement that
data exhibit a normal (i.e.. bell-shaped) distribution, analyses were
carried Out using the logarithms of the raw data. In addition, proba-
bility plots were used to assist with decisions to discard some outlier
values. This step was necessary to minimize the possibility of includ-
ing data from sites with some enrichment and to satisfy statistical
assumptions: a total of 18 out of the 785 data points were discarded.
Finally, metal vs. aluminum regressions were completed. and 95°’o
prediction limits were calculated.
Regression is a standard statistical method of fitting a line to data that
describe the relationship between two variables: in this case, the
dependent variable is the concentration of one of seven heavy metals.
and the independent variable is the concentration of aluminum, The
95°/a prediction limit defines a range where the concentration of a
metal can be expected to occur (with 95°o confidence) under natur-
al conditions. An example of the leadlaluminum relationship is shown
in Figure 1. Similar plots showing a statistically significant relation-
ship across the range of observed aluminum concentrations
(47 to 79,000 ppm) were observed for all seven metals.
Figure 1. Lead/aluminum regression line (a) and 95% prediction
limits (b). Dashed line indicates extrapolation beyond
data range.
Assessment and Watershed Protection Division
Office of Water
Washington, DC 20460
June 1991
LEAD/ALUMINUM
102
-I
10 -b
a
-J I
10-1
10-2
10 10 ’
November 987
Aki num (PPml
102
I 0
I I
-------
u i iLl t or i n ivi u riou
Florida DER has found the metal vs. aluminum plots to be useful in
a number of ways. A key use is simply to flag stations where metal
values may be in excess of naturally expected concentrations. An ob-
servation outside of the 950/0 prediction range should trigger closer
investigation of possible pollutant sources. In addition, the method
provides the ability to track trends in metallaluminum ratios and has
been used to determine where additional monitoring may be needed.
For example. where sediment contamination is suspected. the State
may perform sediment bioassays. Sediment bioassays involve expos-
ing an organism to metal-contaminated sediments and measuring
the effects of exposure on a biological endpoint (e.g.. morlal:t-y growth.
or reproductive effects).
Experience gained in the studies of unimpacted coastal sites has
enabled the DER to improve their sampling. laboratory. and interpre-
tive activities. Key findings include the following:
• Duplicate or triplicate samples are crucial to reducing sampling
variability.
• Total digestion of sediment samples in the laboratory is neces-
sary to ensure release of metals that are tightly bound with the
crystalline structure of sediment minerals.
THE BISCAYNE BAY STUDY
One application of the reference metal approach has been to assess
toxicant levels in the Miami River-Biscayne Bay area of Miami. The
river is a major tributary to the bay, is lined with ship berthing areas,
and receives urban stormwater runoff from Miami. Sediment sam-
ples were taken at 11 stations in the river and bay (see Figure 2), and
results were plotted over the regression lines and confidence inter-
vals obtained from unimpacted areas. Of the seven heavy metals,
arsenic and nickel fell within the expected natural ranges; chromium,
copper. and lead concentrations were both within and above the
predicted natural range. with most stations near the mouth of the river
having the highest concentrations; and cadmium and zinc exceeded
the predicted natural ranges at almost all the stations sampled.
‘ 1iAMi
- LOR IDA
—J10
Figure 2. Sediment monitoring stations in the Miami River-
Biscayne Bay area.
IIl tdI dilu yet e uuy I1dV rilyriCt iutiiuriurii values. irniiar conciu-
sions could not have been reached using water column data alone.
Figure 3. Lead concentrations in sediment from Miami River
and Biscayne Bay. Regression line and 95°/o range
from clean water sites are also shown.
The conclusion that the Miami River was contributing toxics to the
sediments of the river and bay and that expansion in the area ol con-
tamination is likely in the absence of remediation has resulted in an
increased commitment to reduce pollutant inputs to the system
Similar studies have been completed successfully in other estuaries
in Florida, including the Tampa Bay and Pensacola Bay systems. The
approach has been found to be especially valuable in surveys of large
systems. where it can be used to identify potential sources of pollu-
tion and assist in decisions regarding more intensive monitoring
efforts. In addition to the quality of information obtained from the refer-
ence metal approach. the State has been pleased with the methods
simplicity, cost-effectiveness, and the resulting improved consistency
in regulatory decisions and reduction of regulatory delays.
Materialfor this report was furnished by Dr Steven J. Schropc’ and
Fred D. Calder Florida Department of Environmental Regulation.
Coastal Zone Management Section. Florida Department of Environ-
mental Regulation. 2600 Blair Stone Road. Tallahassee, FL 32301.
Figure 3 shows the sediment concentrations of lead at the 11 stations
superimposed over the lines indicating the expected metals concen-
trations oased on statewide sampling of unimpacted areas. The num-
bered data points indicate stations, with 1 representing the most
upstream station and 11 representing the most downstream station.
The increasing degree of enrichment with increasing aluminum con-
centrations is thought to be due to grain size effects. Fine-grain sedi-
ments (typical of the Miami River) more readily adsorb contaminant
1,000 02 DD
- 01
100 08
10
0.1 10,000
1,000
Aluminum (ppm)
I
S
•8
0
N
Biscayne Bay
“S
This report is produced by EPA to highlight monitoring and wasteload
allocation activities. Contributions of information for similar reports are
invited. Please contact EPA, AWPD. Monitonng Branch ( WH-553), 401
M Street, SW. Vv shington. DC 20460(202) 382-7013.
------- |