United States
Environmental Protection
Agency
Great Lakes
National Program Office
230 South Dearborn Street
Chicago, Illinois 60604
EPA-905/9-91-006A<
GL-06A-91
&EPA
Agricultural IMPS Control of
Phosphorus in the New York
State, Lake Ontario Basin
Volume I — Delivery of Phosphorus to
Lake Ontario from Cultivated Mucklands in
Oak Orchard Creek Watershed
Printed on Recycled Paper
-------
FOREWORD
The U.S. Environmental Protection Agency (USEPA) was created because of increasing
public and governmental concern about the dangers of pollution to the health and welfare
of the American people. Noxious air, foul water, and spoiled land are tragic testimony
to the deterioration of our natural environment.
The Great Lakes National Program Office (GLNPO) of the U.S. EPA was established in
Chicago, Illinois to provide specific focus on the water quality concerns of the Great
Lakes. The Section 108(a) Demonstration Grant Program of the Clean Water Act (PL 92-
500) is specific to the Great Lakes drainage basin and thus is administered by the Great
Lakes National Program Office.
Several demonstration projects within the Great Lakes drainage basin have been funded
as a result of Section 108(a). This report describes one such project supported by this
office to carry out our responsibility to improve water quality in the Great Lakes.
We hope the information and data contained herein will help planners and managers of
pollution control agencies to make better decisions in carrying forward their pollution
control responsibilities.
Director
Great Lakes National Program Office
-------
EPA-905/9-91-006A
February 1991
AGRICULTURAL, NONPOINT SOURCE CONTROL OF PHOSPHORUS IN THE
NEW YORK STATE LAKE ONTARIO BASIN
VOLUME 1. DELIVERY OF PHOSPHORUS TO LAKE ONTARIO FROM CULTIVATED
MUCKLANDS IN THE OAK ORCHARD CREEK WATERSHED
by
Patricia Longabucco
Michael R. Rafferty, P.E.
Bureau of Technical Services and Research
Division of Water
New York State Department of Environmental Conservation
Albany, New York 12233-3502
R005725-01
Project Officer
Ralph G. Christensen
U.S. Environmental Protection Agency
Great Lakes National Program Office
230 South Dearborn Street
Chicago, Illinois 60604
1987
-------
Disclaimer
This report has been reviewed by the Great Lakes
National Program Office, U.S. Environmental Protection
Agency, and approved for publication. Approval does not
signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency
nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
-------
CONTENTS
Abstract i
Figures iii
Tables v
Acknowledgements vi
1. Introduction 1
Background 1
Objectives 3
Study Setting 3
2. Conclusions 10
3. Recommendations 12
4. Methods 13
Research Approach 13
Study Sites 13
Flow Measurements 14
Field Data and Sample Collection 15
Sample Processing and Analyses 16
Quality Assurance 17
5. Results 19
Hydrology 19
Water Chemistry 21
Pesticides 37
6. Discussion 39
Phosphorus, Sediment and Pesticide Losses
from. Cultivated Mucklands 39
Stream Transport of Phosphorus and Sediment
through Oak Orchard Swamp 45
Loading to Lake Ontario 49
References 55
-------
ABSTRACT
Cultivated mucklands in western New York State were investigated as a
nonpoint source of phosphorus to Lake Ontario. The 70,500-ha Oak Orchard
Creek watershed, which drains to Lake Ontario, was selected for the study
area. It is located in Genesee and Orleans Counties, New York, and contains
3250 ha of heavily fertilized muck cropland on which predominantly vegetable
crops are grown. The creek was monitored at several sites from May 1984
through April 1985 to determine the role of the mucklands in annual
phosphorus loading to the lake.
At an upstream site which drained approximately 10,200 ha, including the
majority of the muck cropland, the creek load was 18,000 kg of total phos-
phorus with 75 percent of it as dissolved reactive phosphorus. Two-thirds of
the annual load was delivered in the 3-month, high-flow period of February
through April. Runoff during the late winter-early spring period appears to
be the most important hydrologic factor in governing annual phosphorus
loading from the mucklands, greater than either total precipitation or total
runoff for the year.
The pesticide DDT and its metabolites, DDE and ODD, were detected in muck
soils and in creek suspended and bed sediments at this site. Although the
annual loading rate of these compounds was thought to be relatively small,
based on limited sampling, accumulation in the freshwater wetlands downstream
which contain both a federal and a state wildlife refuge, could pose a hazard
to sensitive species and warrants possible further investigation.
A number of impoundments on the creek downstream of the mucklands,
including the managed freshwater wetlands and two hydroelectric facilities,
did not appear to significantly affect transport of phosphorus through the
system during high-flow, late-winter months. In the largest impoundment,
Waterport Pond, which is located 10 km from the creek mouth, internal loading
of phosphorus from bottom sediments occurred during periods of hypolimnetic
anoxia. Uptake and removal of bioavailable phosphorus by algae in Waterport
Pond, rather than dilution by incremental flow, was thought to account for a
spatial phosphorus concentration gradient evident in summer months. The
overall effect of Waterport Pond on annual phosphorus loading to Lake Ontario
appeared to be removal of about 25 percent of the dissolved reactive form.
Total phosphorus loading to Lake Ontario from Waterport Pond was 37.
tonnes for the study year with 54 percent in the dissolved reactive form.
Half of the load was delivered during the high-flow months of March through
May. A phosphorus load mass balance for Waterport Pond indicates that the
greatest portion derives from the upper watershed containing the cultivated
-------
raucklands. Other smaller sources are the Village of Medina wastewater
treatment plant and seasonal diversions of supplemental flow from the Erie
Barge Canal. Control of phosphorus losses from the mucklands would appear to
offer the most significant and cost-effective opportunities for loading
reductions from this watershed.
ii
-------
LIST OF FIGURES
Number
la Map of Oak Orchard Creek Watershed 5
Ib Profile of Oak Orchard Creek 6
2 Map of Oak Orchard Swamp and cultivated mucklands 7
3 Runoff at Site 1, 2 and 3 for study period 22
4 Estimated relationship of Site 2 discharge and
time-of-travel in Oak Orchard Creek between
Sites 1 and 2 23
5 Time course of total phosphorus, dissolved oxygen
and chlorophyll a concentrations at Site 5 26
6 Time course of total phosphorus, dissolved oxygen
and chlorophyll a_ at Site 6 26
7 Time course of total phosphorus concentration at
Sites 1, 2 and 3 during study period 28
8 Changes in flow, phosphorus and suspended sediment
at Site 1 during the February 22-March 4 runoff event 29
9 Changes in flow, phosphorus and suspended sediment
at Site 2 during the February 22-March 4 runoff event 31
10 Phosphorus daily loading rates at Sites 1, 2 and 3
during study period 33
11 Monthly loads of phosphorus at Sites 1, 2 and 3 34
12 Temperature and dissolved oxygen isopleths
for Site 5 35
13 Temperature and dissolved oxygen isopleths
for Site 6 36
14 Relationship between phosphorus concentration in
sediments and sediment surface area 41
111
-------
LIST OF FIGURES(con't.)
Number Page
15 Relationship between Q /Q and instream phosphorus
concentration at Site l during the February-March
runoff event „ 44
16 Relationship between Q /Q and instream phosphorus
concentration at Site 2 during the February-March
runoff event 45
17 Time course of dissolved molybdate reactive
phosphorus concentrations at Sites 5, 6 and 3 53
IV
-------
LIST OF TABLES
Number Page
1 External quality assurance program summary
for analyses of water samples 18
2 Precipitation, runoff and flow for Oak Orchard
Creek (May 1984 through April 1985) 20
3 pH and alkalinity means and ranges of water samples
collected during study period 24
4 Total phosphorus and dissolved molybdate reactive
phosphorus concentration means and ranges for
sample sites during study period 25
5 Concentrations of DDT and metabolites in muck soils
and Oak Orchard Creek sediments and water 38
6 Estimates of phosphorus in excreta produced by visiting
waterfowl at Iroquois National Wildlife Refuge 49
la Mass balance of total phosphorus load (kg) entering
Waterport Pond on a monthly basis compared to actual
measured load at Site 3 for the study period 51
7b Mass balance of molybdate reactive phosphorus load (kg)
entering Waterport Pond on a monthly basis compared to
actual measured load at Site 3 for the study period 52
v
-------
In April 1983, an USEPA-sponsored workshop was convened at Cornell
University to review the problem of nonpoint source phosphorus loading from
New York State to Lake Ontario. The two major aspects considered by the
workshop were the magnitude of various phosphorus nonpoint sources and the
potential for their control. One of the two areas that ultimately received
the largest share of attention and reflected a workshop consensus concerning
priority, was phosphorus losses from muckland agriculture in the basin.
Although cultivated mucklands account for a small portion of agricultural
land use within the basin, the extraordinarily high losses of phosphorus per
unit area from these systems and the potential for low-cost management
indicated that further consideration of mucklands be included in any future
strategy to reduce nonpoint source phosphorus loading to Lake Ontario.
Although many people were involved, a number of individuals are
particularly notable for their roles in the accomplishment of the Oak Orchard
Creek research project. Dr. Mark Brown of the New York State Department of
Environmental Conservation (NYSDEC) was instrumental in the initial
development and inception of the project and provided continual guidance and
technical advice throughout. Ms. Donna Geissinger was responsible for sample
collection and we gratefully acknowledge the time and willingness she
provided to the project.
Mr. Ralph Christensen and Mr. Kent Fuller of the USEPA Great Lakes
National Program Office lent their interest and support to the project and
granted the funds to carry it out.
Chemical analyses were performed in a timely and accurate manner by the
New York State Department of Health (NYSDOH) Laboratories under the direction
of Dr. Liaquat Husain and Mr. Robert Weinbloom. Additional sediment analyses
were provided by Mr. Frankie Ramos of the Cornell University Soil Lab and by
Dr. Martin Wahlen of the NYSDOH. Pesticide analyses were performed by
Versar, Inc. in a timely manner.
Niagara-Mohawk Power Company graciously provided records of their water
releases from Waterport Pond so that loading estimates to Lake Ontario could
be calculated. Through the efforts of Mr. Jeff Auser, Niagara-Mohawk
Syracuse Office, access to three hydro sites for sampling was granted.
The NYSDEC Region 8 Office in Avon, New York, was invaluable in their
lending of staff, equipment, vehicles, lab and office space, and their
general support of the project. In particular, we wish to thank Mr. Bruce
Butler, Regional Water Engineer, Mr. Eric Seiffer, Regional Director, Mr.
vi
-------
Robert Passalugo, Chemist, and Mr. Phil Featherly, Mr. Bob Quackenbush, and
Mr. Matt Grohol at the Oak Orchard Maintenance Center.
Technical assistance and support were generously provided by Mr. Edwin
Chandler, Refuge Manager, Mr. Raymond Whittemore, Assistant Refuge Manager,
and their staff at the Iroquois National Wildlife Refuge.
We also wish to thank local staff of the USDA Soil Conservation Service
and Agricultural Stabilization an<3 Conservation Service, and Cooperative
Extension for their interest in and support of the project.
We sincerely appreciate the permission given us by various landowners in
the watershed to use their properjty for locating our monitoring equipment.
They include Mr. Joseph Ognibene,! Oak Orchard Creek Small Watershed
Protection District, Mr. John Joy, farmer, and Mr. Victor Caleb, former
Highway Superintendent for the Tcjwn of Shelby.
For the preparation of this jreport we gratefully acknowledge the word
processing skills and accuracy of Ms. Sharon Hotaling and the drafting
abilities of 11s. Karen Guetti. '
Vll
-------
SECTION 1
INTRODUCTION
BACKGROUND
Substantial progress has been made nationally in addressing the most
conspicuous water quality problems, largely through the treatment of
municipal and industrial point source discharges. In many cases, nonpoint
source pollution is responsible for a greater portion of the problems still
remaining. The high cost and limited potential of additional point source
treatment underscore the need for devising and implementing cost-effective
nonpoint source control programs to achieve designated uses of water bodies.
Through international treaty, the United States and Canada have agreed
to further reduce phosphorus loading to the lower Great Lakes. As one of the
involved states, New York has developed a phosphorus control strategy
designed to meet the target load of 7000 tonnes as established for Lake
Ontario in Annex 3 of the 1978 Great Lakes Water Quality Agreement. In
addition to certain point source reductions, New York has proposed a
comprehensive agricultural nonpoint source phosphorus reduction program. A
nonpoint source contributing area's priority in the program will be
determined by the magnitude of the contribution, its treatability, related
side benefits, and cost of treatment. Based on past research, cultivated
mucklands appear to be the greatest source per unit area of nonpoint
dissolved phosphorus to streams in the New York Great Lakes basin. For this
reason, the present study was undertaken to determine the significance of
muckland-derived phosphorus in the annual delivery of phosphorus by a
tributary to Lake Ontario.
Mucklands
Mucklands (Histosols; organic soils) were formed under water or in very
poorly drained areas and consist mostly of decomposed plant material. The
soil is described as peat if undeveloped or fibrous in texture, and as muck
when well decomposed (Lucas, 1982). There are approximately 2.8 million ha
of organic soils within the Great Lakes basin. Only a small percentage are
drained for agricultural use; of the 235,000 ha in New York State,
approximately 16,000 ha are cultivated for vegetable production, 8,800 ha of
which are in the Lake Ontario basin (Klausner, 1986) .
Mucklands are artificially drained usually through a network of
subsurface tile drains and open ditches which eventually discharge to an
existing stream. Often the stream is channelized and improved to prevent
flooding of the cropland during high-flow periods. Proper regulation of the
water table is important in agricultural utilization of organic soils.
-------
Adequate drainage must be provided for optimum crop yield, yet to prolong the
life of the soil, the maintenance of as high a water table as practical is
necessary (Lucas, 1982).
The predominant use of cultivated mucklands is for the production of
vegetable crops. Most virgin Histosols contain less than 0.1 percent
phosphorus (Lucas, 1982) and are heavily fertilized (commonly 100 to 130
kg/ha P00r in New York) to ensure good crop production. Reported annual
phosphorus leaching losses from organic soils range from 1 to 30 kg/ha and
concentrations in drainage waters have been measured as high as 10 mg/L
(Duxbury and Peverly, 1978; Erickson and Ellis, 1971; Hortenstine and Forbes,
1972; Miller, 1979; Nicholls and MacCrimmon, 1974). In contrast, annual
phosphorus leaching losses from mineral soils average <0.01 kg/ha, although
runoff losses can approach 1 kg/ha (Hanway and Laflen, 1974). The
variability in muckland leaching losses of phosphorus is likely due to a
combination of factors, including the rates of fertilization and
mineralization, soil and water management practices, the ability of the soil
to adsorb phosphorus, and the nature of the underlying mineral material
(Duxbury and Peverly, 1978; Cogger and Duxbury, 1984).
Past Research
Duxbury and Peverly (1978) investigated phosphorus losses from the Elba
muckland located in western New York. The muckland is drained by Oak Orchard
Creek which eventually empties into Lake Ontario. A deep muck (>107 cm) site
with tile lines laid in organic soil annually discharged up to 30 kg/ha of
dissolved molybdate reactive phosphorus (DMKP) in subsurface drainage waters.
In contrast, a shallow muck (45 to >107 cm) site with tile lines
predominantly in the underlying mineral substratum had leaching losses of up
to 0.9 kg/ha. Monitoring of Oak Orchard Creek at a point where it drained
about two-thirds of the 3000-ha muckland suggested that the phosphorus levels
measured in the deep muck drainage waters were more representative of the
entire system and placement of the drains in underlying mineral material did
not necessarily offer protection against leaching losses of phosphorus.
The factors affecting phosphorus leaching losses from, mucklands were
studied by Cogger and Duxbury (1984). Laboratory experiments employing a
series of leachings and incubations were performed to determine the
mineralization potential of soils collected from several sites in the Elba
muckland. There was a general lack of response in leaching losses to
environmental factors indicating that the release of phosphorus was
controlled by equilibration of inorganic phosphorus between more or less
labile forms, but not directly by biological mineralization. Adsorption of
phosphorus by the organic soils was much lower than levels reported for
mineral soils. The ability of the organic soils to retain phosphorus during
leaching processes appeared to be largely mediated by the content of total
and extractable iron and aluminum. Therefore, when applications of
fertilizer phosphorus exceed the amount needed for plant growth, the surplus
is either adsorbed by the soil or lost through leaching. Cogger and Duxbury
(1984) suggest that the key to controlling phosphorus leaching losses from
cultivated mucklands is the limitation of fertilizer applications in areas
where phosphorus response is small. Reductions in leaching losses would not
-------
be immediate, however, due to the build-up of phosphorus in the soil over
time.
The benefits that would accrue to Lake Ontario by implementation of a
muckland fertilizer management program in the Oak Orchard Creek watershed,
where the largest concentration of cropped muck soils are contained, are
difficult to ascertain. Phosphorus delivery from the mucklands to Lake
Ontario may be mediated by processes occurring in a relatively complex
transport system. This system includes a large, managed wetland (Oak Orchard
Swamp), two impoundments used for the generation of electricity (Waterport
Pond and Glenwood Lake), and water supplements from the Erie Barge Canal to
augment flow during dry summer months. In order to assess the benefits of a
phosphorus management program, a better understanding of phosphorus dynamics
in Oak Orchard Creek was needed.
OBJECTIVES
Duxbury and Peverly (1978), Cogger and Duxbury (1984) and Peverly (1982)
have shown that, per unit area, muckland drainage water could be a
significant source of phosphorus to streams. It has been suggested that the
phosphorus losses from mucklands can be considerably reduced by the adoption
of no/reduced phosphorus fertilizer practices. Since phosphorus additions do
not appear to increase muckland productivity in many cases, the costs of
practice implementation would be primarily related to an educational
program. Given the potential for significant reductions of phosphorus at low
cost, control of phosphorus from mucklands was seen as an appealing component
of the state and federal strategies to reduce phosphorus loading to the Great
Lakes. However, the relative magnitude of phosphorus contributions from New
York mucklands was uncertain.
The objectives of this research were to:
1) determine phosphorus losses from the muckland drainage
area of Oak Orchard Creek, a tributary to Lake Ontario;
2) assess the mediating effects of the stream bed, Oak
Orchard Swamp and Waterport Pond on the transport
of phosphorus through Oak Orchard Creek;
3) construct a phosphorus budget for Oak Orchard Creek
including point sources, runoff, intermittent
discharges and muckland contributions;
4) evaluate pesticide loading from cultivated
mucklands to Oak Orchard Swamp.
STUDY SETTING
Oak Orchard Creek, located in Genesee and Orleans Counties, winds a
90-km course from its source at 260 m MSL to the mouth at Point Breeze on
-------
Lake Ontario (Figures la and Ib). Rising just north of the City of Batavia,
the main stem and two smaller tributaries flow a few kilometers through a
cultivated upland area of about 35 km2 before entering the Elba rauckland. In
this upland segment, the creek is nearly dry during summer months except for
discharges from the wastewater treatment plant at Elba which has a flow of
approximately 350 m3/d. Upon entering the mucklands, the creek channel is
improved to aid drainage, and the bottom lies on calcareous till material or
marl (CaCCO . Water is pumped into the creek from collection pits in
adjoining fields, and there is little direct surface runoff or water-related
soil erosion.
West of Oak Orchard Road, the creek is joined by the Manning tributary,
a man-made channel providing a water course for drainage of the Manning
muckland which lies about 2 km to the north of the Elba rauckland (Figure 2).
After leaving the muck areas, the creek meanders westward in its natural
stream bed through Oak Orchard Swamp and receives Tributary 138-28-3d from
the south central portion of the drainage basin. Discharges to this
tributary include the Oakfield municipal wastewater treatment plant (approxi-
mately 700 m3/d) and U.S. Gypsum, a paper-board manufacturer.
Effluent from U.S. Gypsum plant is discharged to several small
tributaries of Oak Orchard Creek. The facility manufactures sheetrock which
involves the mining of gypsum and the production of paperboard. The
operation has eight permitted discharges, seven of which are for mine
drainage only. However, during most years, including the study period, only
two of the mine dewatering discharges are used. In addition, in all but the
wettest months only one dewatering point is required. The New York State
Department of Environmental Conservation (NYSDEC) requires the company to
maintain a minimum flow of 11,300 mVday of mine water drainage to
sufficiently dilute the paperboard plant waste because the receiving stream
is frequently dry. It is often necessary for the plant to direct flow from
within the mine to the wastewater discharge point in order to meet this
requirement.
Oak Orchard State Wildlife Management Area (WMA) (1011 ha) and Iroquois
National Wildlife Refuge (NWR) (4371 ha) are located within the Oak Orchard
Swamp segment and represent major waterfowl resting and feeding sites and
wildlife breeding habitat. Oak Orchard Creek flow is regulated in the
Iroquois National Wildlife Refuge by a number of stop-log control structures.
Four pools form the major development of the refuge and impound water from
the creek directly, or indirectly through other pools, springs, and local
surface runoff. They are drained prior to the spring thaw season in order to
provide the normal flood plain for the creek and thus prevent flooding of
Route 63 and surrounding areas. Drawdown generally begins in the latter part
of November, following waterfowl hunting season, and is complete by January.
The available storage in the drained pools serves to decrease the duration of
flood water downstream of the swamp. This decrease in length of flood time
serially allows an earlier draining of the upstream main stem with an obvious
benefit to the muck farms (USFWS, 1964).
Cattail (Typha latifolia L.) and reed canary grass (Pharlaris
-------
Lake Ontario
Oak Orchard
State Wildlife
Management
\ Iroquois National
\Wildlife Refuge
Oak Orchard Creek Watershed
FIGURE la. Map of Oak Orchard Creek watershed showing locations of point
source discharges from sewage treatment plants at Elba (A),
Oakfield (D) and Medina (C) , and from U.S. Gypsum (B); samp-
ling stations 1,2,3,4,5,6; and project field lab at Oak
Orchard Wildlife Management Area maintenance center (F).
-------
300
OAK ORCHARD CREEK PROFILE
20
30
4O 50 60
Kilome te rs
70
80
90 1OO
FIGURE Ib. Profile of Oak Orchard Creek.
arundinacea L.) dominate the plant cortmunity adjacent to the creek and around
several shallow pools in the swamp. The flood plain extends for 1 km on each
side of the channel and vegetation there includes red maple (Acer rubrum L.),
cottonwood (Populus deltoides Marsh) and willow (Salix nigra Marsh).
Downstream of the wetlands, Oak Orchard Creek runs northward dropping
more than 60 m over the Niagara Escarpment through the Village of Medina and
under the Erie Barge Canal to Glenwood Lake. During the low-flow period (May
through December), approximately 5.5 m3/s of Barge Canal water is diverted to
Oak Orchard Creek to augment flow. Glenwood Lake, the first of two
impoundments operated by the Niagara-Mohawk Power Corporation for the
generation of electricity, has a maximum depth of 15 m, surface area of 36 ha
at normal elevation, volume of 2 x 106 m3, and average retention time of 3 d
(Niagara-Mohawk Power Corporation, file report). The Village of Medina
sewage treatment plant discharges to Oak Orchard Creek below Glenwood Lake
and has an average wastewater flow of 9500 m3/d (NYSDEC, files). From
Glenwood Dam to Waterport Pond the stream is fast-moving over a rocky bottom.
The time of travel for this 15.3-km stream segment for the minimum required
release from Glenwood Dam (0.18 m3/s) is 48 to 66 h (NYSDEC, 1975).
-------
'AOak Orchard Swamp
© Site 1
® Site 2
O Peverly's (1982) upstream site
• Peverly's muckland site
FIGURE 2. Map of cultivated mucklands and Oak Orchard Swamp. Sampling sites 1 and 2 are
indicated.
-------
Waterport Pond (Lake Carlton, Lake Alice) has a maximum depth of 23 m, a
surface area of 136 ha, a volume of 11 x 106 m3, and an average residence
time of approximately 15 d (Niagara-Mohawk Power Corporation, file report).
The drainage area of the lake is 587 km2. According to New York State's
water body classification, the best use of Waterport Pond is for "fishing and
any other usages except for bathing or as a source of water for drinking,
culinary or food processing purposes." However, the lake is heavily used for
recreation including swimming (Turkow et al., 1982).
Below Waterport Dam, Oak Orchard Creek is deep and relatively
slow-moving over a muddy stream bed. The lower portion of Oak Orchard Creek
is considered an important fishing stream, ranking among the top twenty-five
areas within New York State for black bass, panfish, esocids, and
bullhead/catfish (NYSDEC, 1981). The major tributary entering this section
is Marsh Creek which drains a predominantly upland agricultural area of just
over 100 km2.
The total watershed area of Oak Orchard Creek is 705 km2 (70,500 ha).
Major land uses in the watershed are: cropland, 48 percent; woodland, 19
percent; open land formerly cropped, 10 percent; orchards and vineyards, 4
percent; pastureland, 3 percent; federal land, 6 percent; urban areas, 3
percent; other, 7 percent (SCS, 1974).
Large-scale agricultural activities began in the nuckland areas
approximately 50 years ago when a complicated system of drainage ditches was
installed to allow aeration and cultivation of the soils. These highly
organic soils were formed as a result of a receding glacial lake which was
gradually filled by deposition of sediment originating from the surrounding
uplands. Decaying organic matter in these wetlands produced the Histosols
which, in their virgin state, are 50 to 80 percent organic matter containing
2 to 3 percent nitrogen and 0.1 to 0.2 percent phosphorus.
The two cultivated mucklands in the watershed, the Manning muck and the
Elba muck, are respectively 250 ha and 3000 ha in size. In 1974, the
muckland areas were comprised of approximately 100 farms with an average size
of 30 ha (SCS, 1975).
A system is presently under construction that will improve muckland
drainage by installing an additional 136 km of tile drain to supplement
approximately 295 km of existing drains (SCS, 1975) . To protect fields from
flooding, 64 km of floodwater diversions to intercept upland flows, 51 km of
floodways to carry water from diversions to the main channel, and 30 km of
open channel are being constructed or upgraded.
Major crops grown on the mucks are onions, potatoes, sweet corn, and
lettuce. In 1972, the area produced 28 percent of the total onion crop grown
in New York State (SCS, 1975). Prior to the USEPA ban in January 1973, DDT
was the most widely used pesticide on muckland crops; nitrofen and aldicarb
were subsequently used until they too were banned. Currently, parathion,
diazinon and allidochlor are among the most frequently applied pesticides
(Starowitz, 1984). In the mid-1970's, the application rates used for
muckland agriculture exceeded those used on upland mineral soils by factors
8
-------
of two for herbicides, seven for pesticides, and eight for fungicides (SCS,
1975). Fertilizer application rates for muckland agriculture were similar to
those used on upland mineral soils for nitrogen and at least twice that used
for phosphorus (SCS, 1975).
-------
SECTION 2
CONCLUSIONS
Phosphorus sources to Oak Orchard Creek include three municipal
wastewater treatment plants and water diversions from the Erie Barge Canal,
but contributions from these are comparatively small and the potential for
load reductions limited. The major sources are nonpoint and result from
various land use practices in the watershed. The contribution representing
both the highest load rate per unit area and the greatest portion of the
annual load appears to be losses from cultivated mucklands.
Although representing only 27 percent of the drainage area, the mucklands
were estimated to account for between 56 and 86 percent of the annual
dissolved molybdate reactive phosphorus load measured in Oak Orchard Creek
before it enters the swamp. Depending on the rate used for non-muckland
contribution, the average areal DMRP loading rate for the mucklands during
the study period was calculated to be between 2.8 and 4.4 kg/ha which is in
the range of values reported by other investigators. Suspended sediments
collected in the muckland section of Oak Orchard Creek were enriched with
phosphorus with respect to particle surface area when compared to sediments
from a rural, non-calcareous mineral soil watershed; however, particulate
phosphorus accounted for only 16 percent of the annual phosphorus load in
this section of the watershed. Dissolved phosphorus comprised the bulk of
the phosphorus load (84 percent) measured in this stream segment, and most of
the phosphorus was delivered from late winter to early spring. Variations in
annual areal loading rates from the mucklands appeared to be largely
dependent on the amount of runoff during this critical period.
The relatively long time-of-travel through Oak Orchard Swamp likely
results in some exchange of phosphorus between water and sediments.
Phosphorus release frcm bottom sediments and senescing aquatic plants during
late summer to early fall was thought to occur, but when compared to loads
delivered under high-flow, runoff conditions, these amounts were negligible.
Overall, the data suggest a basically flow-through system with little net
retention in the swamp. Through assimilation processes, the bottom sediments
are enriched with phosphorus after many years of contact with cultivated
muckland drainage waters. Because of this, the swamp may continue to act as
a source of phosphorus for a time even if muckland phosphorus losses were
controlled and drainage waters became cleaner in the future.
Through an analysis of storm event characteristics at a stream sampling
location downstream of the swamp and calculation of a phosphorus budget for
this location, phosphorus delivered upstream appeared to be conservatively
transported through the swamp. Allowing for this, 66 percent of the TP load
and 75 percent of the DMRP load at the downstream site could then be
ID
-------
attributed to the phosphorus losses measured upstream.
The annual loading from Oak Orchard Creek to Lake Ontario for the study
year, May 1984 through April 1985, was 37. tonnes of total phosphorus, 20.
tonnes of which was dissolved molybdate reactive phosphorus. Half of the
total phosphorus load was delivered in three months, March, April and May,
when intensive rain and snowmelt produced major runoff and high creek flows.
If conservative transport from the mucklands to Waterport Pond outlet is
assumed, then as much as 72 percent of the DMKP load and 39 percent of the
TP load could be attributed to inputs from the mucklands.
There were indications of certain instream processes mediating transport
and delivery of phosphorus in this part of the system, however. Measurements
in Waterport Pond during summer stratification revealed increasing phosphorus
concentrations in the hypolimnion as dissolved oxygen concentrations
decreased to zero. This likely reflected internal loading from bottom
sediments under anoxic conditions which accounted for only a small fraction
of the total annual phosphorus budget of the impoundment. In the summer
months, a gradient in DMKP concentration existed in the epilimnetic waters of
Waterport Pond, with lower values being observed closer to the outlet.
Dilution from incremental flow does not appear to account for this, rather,
biological uptake and removal seem the most probable explanation. A mass
balance of phosphorus inputs and outputs for Waterport Pond suggests an
overall annual effect of 25 percent removal of the DMRP load entering the
impoundment.
The apparent significance of the mucklands in annual phosphorus loading
points to proper fertilizer management as a likely method for reducing
phosphorus losses to the lake from the Oak Orchard Creek watershed and other
muckland areas in the basin. Practices designed to limit leaching of
phosphorus fron muck soils to drainage waters could reduce the load delivered
to the creek and eventually Lake Ontario.
Limited pesticide analyses of environmental samples revealed that, in
muck field soils and in creek bed and suspended sediments, the greatest
portion of ZDDT was in the 4',4'-DDT form rather than in the weathered forms
of ODD and DDE. Concentrations in suspended sediments suggest that annual
loading to the wildlife refuges in Oak Orchard Swamp is likely relatively
small. However, many years of accumulation of these persistent compounds in
the swamp could pose a hazard to sensitive species which either seasonally
utilize the area or breed there. Further investigation of this situation may
be warranted.
11
-------
SECTION 3
RECOMMENDATICWS
The approximately 8800 ha of cultivated mucklands do not represent a
large fraction of the total cropland in the New York Lake Ontario basin and,
therefore, limited potential exists for reducing the phosphorus load to the
lake by controlling this source. From the results of this study and the
Muckland Fertility Study (Volume 2), it appears that the Oak Orchard Creek
nonpoint source phosphorus load could be reduced by limiting inputs from
muckland drainage waters through the adoption of a fertilizer management
program that controls phosphorus leaching losses. Such a program should
employ education and demonstration efforts to convince growers to fertilize
only when soil tests indicate the need. With the application of such e
program, some phosphorus could definitely be prevented from entering Lake
Ontario, although actual amounts are uncertain due to the variability in
muckland areal loading rates which seem to depend largely on runoff
quantities during the critical late-winter, early-spring period, as well as
stream transport processes, past fertilizer practices, depths of organic soil
and water table, underlying soils types, and infiltration rates. It has been
successfully shown that phosphorus additions can be reduced or eliminated on
certain mucks with no significant difference in yields when compared to
growers' application rates (see Volume 2). However, if significant water
quality improvements are to be realized, then continuation, expansion and
funding of such education and demonstration programs are necessary.
Because of the apparent persistence of DDT and its metabolites in this
system, it is recommended that additional research be performed to better
determine extent of contamination in the refuges, posible accumulation in
fish, birds and mammals, rate of degradation of DDT to ODD and DDE in the
muck soils, and loading rates to the creek.
12
-------
SECTION 4
METHODS
RESEARCH APPROACH
Assessment of the delivery of muckland-derived phosphorus to Lake
Ontario was approached through measurements of phosphorus flux in Oak Orchard
Creek at several locations and determination of certain physical, chemical
and biological properties of the system. Various sources and sinks of
phosphorus were identified and related to changes in phosphorus flux.
Known point sources of phosphorus in the watershed include municipal
wastewater treatment facilities and industrial effluents. The Erie Barge
Canal also discharges seasonally to Oak Orchard Creek through a penstock.
Nonpoint source phosphorus can be present in overland flow, subsurface
runoff, and groundwater discharge. Overland flow includes surface runoff
from agricultural, forested, and urban areas. Significant subsurface and
groundwater runoff sources in this watershed would presumably be limited to
predominantly the cultivated mucklands. Large seasonal concentrations of
waterfowl in Oak Orchard Swamp as a source of phosphorus was also considered.
Possible sinks of phosphorus include Oak Orchard Swamp, Glenwood Lake,
Waterport Pond and the stream bed. However, during certain times of the year
these areas could also serve as sources of phosphorus through internal
nutrient recycling. A determination of residence time as a function of
stream discharge can provide insight on the effects of physical, chemical and
biological processes on instream phosphorus flux.
STUDY SITES
Three primary stream sites (Figures la, 2) were intensively studied
throughout the 12-month project period (May 1984 through April 1985). Site 1
with a drainage area of 102 km2 was located at the Route 98 bridge in the
Town of Elba. About 80 percent of the Elba muck and all of the Manning muck
are drained at that point. Site 2 drained 388 km2 and was located at the
Harrison Road bridge just downstream of Oak Orchard Swamp (this site was
Peverly's (1982) Site 5). Elevation decreases only 6 m in the 27 km between
Site 1 and Site 2, a mean flow gradient of 1 in 4500. The outlet of
Waterport Pond was Site 3, located about 10 km from the mouth of the creek at
Point Breeze. Site 3 drained an area of 587 km2, 83 percent of the total
watershed. A study site was also located on the Erie Barge Canal where it
discharges into Oak Orchard Creek during low-flow months (Site 4). Two
additional stations were established on Waterport Pond: Site 5 at Kenyonville
Road, representing the upper arm of the impoundment; and Site 6 at Waterport
13
-------
Road, representing the main body of the impoundment. These two stations were
monitored from June through December 1984.
FLOW MEASUREMENTS
An enclosed monitoring shed was located at Site 1 and housed a pumping
system which provided stream water to a refrigerated Sigmaraotor automatic
sampler and a Westphalia industrial centrifuge. Backwater conditions were
expected at this site due to the extremely small slope of the stream section,
therefore two stilling wells with continuous punch recorders were installed
approximately 1 km apart above and below Site 1. In this way, the second
gauge could be used to adjust for effects of backwater.
At Site 2 a small shed housed an automatic sampler and a continuous
stage recorder. The bedrock stream bottom at this site led to installation
of a bubbler type stage recorder. An existing USGS wire-weight gauge
attached to the bridge was used to calibrate stream stage as measured by the
bubbler system.
Stages-discharge relationships were developed for both sites through
frequent manual stream gaugings using a Marsh-McBirney 201 velocity meter on
a cable and reel system. During periods of ice cover, gauging was performed
through holes cut in the ice, thereby ensuring an accurate continuous flow
record.
Site 1 was gauged each time a sample was collected, at least once
weekly. A total of 168 gaugings were performed at Site 1 from May 1984
through April 1985. Site 2 was gauged a minimum of once weekly and every
time there was a change in stage of 0.15 m (0.5 ft) or more. Total number of
gaugings at Site 2 for the aforementioned period was 99.
Discharge at Site 3 was determined through use of release records
provided by Niagara-Mohawk Corporation which operates a power-generating
station at the Waterport Dam. There was no way to ascertain accuracy of the
Niagara-Mohawk records because raw data from which average daily flows were
computed were not made available to NYSDEC. Values were assumed accurate to
two significant digits, however.
Incremental flow between Site 2 and Site 3 was estimated by the
difference in flow between the two sites during months when Erie Barge Canal
discharge was not supplementing Oak Orchard Creek (December tlirough April).
It was found to average about 26 percent of the Site 2 flow. Erie Barge
Canal additions during May through November were estimated by subtracting the
sum of the Site 2 flow and the estimated incremental flow from the Site 3
flow. Flows from the three monitored wastewater treatment plants, Elba,
Oakfield and Medina, were estimated from filed data at NYSDEC.
A time-of-travel survey was conducted under high-flow conditions to
estimate residence time of water in Oak Orchard Swamp and travel time in the
creek downstream of the swamp. The fluorescent dye JRhodamine WT was injected
into the creek at Site 1 and traced through the system to Site 2. Portable
14
-------
automatic samplers were placed at intermediate locations to monitor progress
of the dye, and the refrigerated automatic sampler in the monitoring shed at
Site 2 was used to collect samples there. Concentrations of dye in samples
were determined using a Sequoia-Turner Model III fluorometer and a
calibration curve.
FIELD DATA AND SAMPLE COLLECTION
From May 1, 1984 to April 30, 1985, the designated water year for the
study, Oak Orchard Creek was monitored continuously and sampled both on
routine and storm event bases. During that time, 489 samples were collected
and analyzed, with an additional 89 samples analyzed as part of the external
quality assurance program. Of the 489 samples, 79 were collected during
seven runoff events.
Stream water grab samples were collected by lowering a weighted sampling
bucket from the road bridge. Occasionally during high flow events, automatic
samplers were used to supplement grab sampling. The frequency of sampling
during runoff periods was between one and three times per day at each of the
three primary sampling sites.
Sampling of the Waterport Pond outlet was accomplished by lowering the
weighted bucket from a stone pier just upstream of where water enters the
Niagara-Mohawk penstock. The two lentic sampling locations on Waterport
Pond, Sites 5 and 6, were sampled from bridges using a 2-L PVC Kemmerer
bottle. Epilimnion and hypolimnion composite samples were collected
separately with reference to the location of the thermocline. Using the
bridge bucket, Erie Barge Canal water was sampled from the penstock just
above where it discharges to Oak Orchard Creek.
The three wastewater treatment facilities, Elba, Oakfield and Medina,
were sampled on an occasional basis. Plant staff provided 24-h composite
effluent samples when requested.
Air and water temperature were recorded at the time of each stream
sampling. At Sites 5 and 6, dissolved oxygen and water temperature profiles
were determined using a Yellow Springs Instruments Model G4 dissolved oxygen
meter. In addition, chlorophyll a. concentrations in the epilimnion and
hypolimnion were measured at these sites using an acetone
extraction-fluorometric determination procedure (APHA et al., 1980).
Precipitation was measured through use of a precipitation collector
installed at the Oak Orchard Wildlife Management Area maintenance center.
These data were supplemented with records from the Batavia and Albion weather
stations.
Bulk suspended sediment was collected at Site 1 during two runoff events
by pumping stream water through a Westphalia industrial centrifuge. The
general use of continuous centrifugation for the collection of suspended
sediment in fluvial systems is described by Ongley (1982).
15
-------
Bottom sediments were collected at several stream sites using a Ponar
grab lowered from a road bridge. In addition, through access with a rowboat,
a number of sediment cores were collected from Oak Orchard Swamp.
SAMPLE PROCESSING AND ANALYSES
Samples were processed in a small field laboratory at the Oak Orchard
Wildlife Management Area maintenance center. Processing usually was
performed within a few hours of collection and never after more than 24 h.
Until processing could begin, samples were stored at 4°C.
Processing of water samples generally involved establishing a sample
record, filtering sample aliquots, dispensing filtrate and whole aliquots to
clean, labeled sample containers, adding preservative, and freezing.
Filtration through a pre^washed 0.45-u membrane filter defined dissolved
chemical fractions.
Alkalinity in mg/L CaCO_ was determined at the field lab by sulfuric
acid titration to pH 4.5 (USEPA, 1979) , and pH was measured on a
Sargent-Welch pH meter. Suspended solids concentration was determined at the
field lab by filtration through a pre-washed, dried and tared glass fiber
filter followed by drying at 103°C and reweighing (USEPA, 1979).
Total phosphorus (TP), dissolved phosphorus (DP) and dissolved molybdate
reactive phosphorus (DMRP) were determined by the New York State Department
of Health (DOH) lab in Albany, New York, using the ascorbic acid-molybdate
complex Method 365.2 (USEPA, 1979). DMRP was considered to represent the
phosphorus fraction immediately available for biological utilization.
TP and DP samples were preserved with 5N H^SO. and frozen, while DMRP
samples were frozen unacidified. Because unpublisned work by Bouldin et al.
suggested that freezing of natural water samples could result in a loss of
DMRP through calcite precipitation, an experiment was conducted at the
beginning of the project to determine the appropriateness of this
preservation method. A single, Site 1 water sample was filtered and split
into two subsamples. One subsample was spiked with phosphate solution to
raise the DMRP concentration by about 1.5 mg/L. Five aliquots of each
subsample were analyzed forthwith, while ten aliquots of each subsample were
frozen and stored in plastic containers. At 7 and 14 d, five aliquots of
each were thawed and analyzed. Coefficients of variation for all sets of
five replicates were 3 percent or less indicating a high degree of precision
within replicates. There were significant differences (p < .05) between mean
concentrations of the unfrozen versus the two frozen batches for both the
unspiked and spiked groups. Percent differences between the frozen and
unfrozen groups were very small, however. Significant differences were
attributed to differences in the degree of analytical accuracy on any given
day rather than a systematic loss of DMRP through freezing. Although the
alkalinity of Oak Orchard Creek is generally in the range of 100 to 300 mg/L
as CaCO.,, the freezing of DMRP samples did not appear to play a role in
phosphorus loss.
16
-------
Sediment concentrations of 0.1 N sodium hydroxide-extractable
phosphorus, citrate dithionite bicarbonate-extractable phosphorus, and
hydrochloric acid-extractable phosphorus were measured at DOH using the
sequential extraction method originally presented by Chang and Jackson (1957)
and later modified by Williams et al. (1971a, 1971b) and Logan (1978) .
Sediment particle size analysis was performed at Cornell University's Soil
Laboratory according to the modified pipette method of Dower and Olson
(1980) . The percents of ten size fractions, including five sand, three silt,
and two clay sizes, were reported. Surface area per unit sediment mass
(cm2/g) was calculated for each sediment sample using the following equation
adapted from Pavlou (1980):
10 N. A .
i=i p V
s
where a = correction factor = 1 x 10 2
N. = percent of the i particle size fraction in the sample
p = density of the sediment (assumed 2 g/cm3)
r = radius of the particle as a sphere = \ midpoint of the
size fraction range
A = 4 II r 2 = area of the particle as a sphere
o o
V = 4/3 H r 3 = volume of the particle as a sphere
o 5
The surface area of a particular sediment is a general indicator of its
sorbing capacity for dissolved chemical substances. Actual adsorption or
desorption depends, among other things, on the dissolved and sediment
concentrations of the substance, the partition coefficient, and the
concentration of sediment in the water.
Pesticide analyses of water and sediment were performed according to
Method 610 (USEPA, 1982a) and Method 8080 (USEPA, 1982b) , respectively.
QUALITY ASSURANCE
The quality assurance program for the project is detailed in a separate
document (Longabucco, 1984) . Briefly, analytical quality control was
administered internally by DOH and externally by NYSDEC. Internal quality
control consisted of reference sample, duplicate sample, chemical recovery
and performance programs. The external program involved the submittal of
duplicate, field-spiked (spiking the whole sample prior to processing) , and
lab-spiked (spiking the individual processed aliquots) samples at the rate of
one set every twenty samples. In addition, blind distilled water blanks were
submitted on an occasional basis. Suspended solids were analyzed in
duplicate at the field lab for every twentieth sample. A summary of the
external quality assurance program is presented in Table 1.
17
-------
TABLE 1. EXTERNAL QUALITY ASSURANCE PROGRAM SUMMARY FOR ANALYSES OF WATER SAMPLES.
oo
PARAMETER n
DMRP 10
10
DP 4
6
TP 11
9
DUPLICATES
Mean % Diff.
6
11
8
11
8
10
.3
.1
.7
.9
.5
.6
(H)*
(L)
(H)
(L)
(H)
(L)
RECOVERY (field)
n Mean %
8
10
3
6
9
9
93
70
68
73
77
83
.8
.0
.3
.9
.2
.3
(H)
(L)
(H)
(L)
(H)
(L)
RECOVERY (lab)
n Mean %
10
10
3
6
11
9
95
86
85
99
112
99
.5
.0
.0
.6
.2
.4
(H)
(L)
(H)
(L)
(H)
(L)
* Sample phosphorus concentrations of £0.10 mg/L for DMRP and DP and <0.15 mg/L
for TP were considered low level (L) ; concentrations above those values were
considered high level (II).
-------
SECTION 5
RESULTS
HYDROLOGY
Oak Orchard Creek hydrological data for the period May 1, 1984 to April
30, 1985 are given in Table 2. Average monthly precipitation was calculated
from measurements made at the Albion and Batavia NOAA meterological stations
and at the Oak Orchard maintenance center. During the 12 months, 81.4 cm of
precipitation fell on the watershed. VJhen comparing the Batavia station's
precipitation record of 82.98 cm with its 30-yr average of 86.28 cm, the
study period is shown to be slightly drier than normal on an annual basis.
However, individual months deviated more from the average than the total
annual amount would indicate. The months of January, February, March and
April, when a major portion of the runoff usually occurs, were drier than the
Batavia 30-yr average for those months by the respective percentages of 36,
40, 2 and 65; however, May experienced 85 percent more precipitation than the
average. July, October and November were drier by 40, 60 and 56 percents,
respectively, while August, September and December were wetter by 34, 64 and
20 percents, respectively. Precipitation at the more northern Albion station
was generally greater than at the Batavia station.
Exclusive of Erie Barge Canal inputs, runoff via Oak Orchard Creek
measured at the outlet of Waterport Pond accounted for 31. cm or 38 percent
of the average annual precipitation of 81.4 cm. Predictably, late winter and
early spring months accounted for much of the annual runoff: 55 percent
occurred from February through April. In contrast, only 2 percent occurred
during the summer months of July, August and September, even though
precipitation during these summer months was about 9 cm greater.
Runoff differed substantially among the three monitored sites. Total
runoff for the watershed section beginning at the headwaters to Site 1 was
24. on or 30 percent of the total precipitation. In the second watershed
section, Site 1 to Site 2, runoff was 41. cm or 51 percent of the annual
precipitation. Nearly 12 cm more snow and rainfall were recorded from
December through March in the mid-portion of the watershed, which contains
more of the drainage area of this section, than in the southern portion. The
last monitored section of the watershed, Site 2 to Site 3, had the least
runoff, only 19. cm or 24 percent of total annual precipitation.
Ratios of calculated monthly runoff to average monthly precipitation
(runoff coefficients) for each monitored watershed section are expressed as
percentages in Table 2. Percents greater than 100 can largely be explained
by the melting of snow deposited in preceding months. At each site, ratios
19
-------
TABLE 2. PRECIPITATION, RUNOFF AND FLOW FOR OAK ORCHARD CREEK (MAY 1984 THROUGH APRIL 1985) .
Respective drainage areas are 102 kma, 388 kma, 587 km*.
Month
May
June
July
August
September
October
Novenfcer
December
January
February
March
April
TOTALS
*Average
Precipitation
(cm)
13.1
7.16
3.48
8.48
12.8
3.05
4.45
7.29
6.02
5.46
6.73
3.40
81.42
** Incremental
Runoff
(cm)
4.9
0.63
0.08
0.07
0.21
0.03
0.25
1.1
1.6
5.4
6.9
3.2
24.37
Runoff
Coeff.
(%)
37
9
2
1
2
1
6
15
26
100
102
93
Total
Flow
(106 m')
5.0
0.64
0.08
0.07
0.21
0.03
0.25
1.1
1.6
5.6
7.0
3.2
24.78
Incremental
Runoff
(cm)
6.0
2.3
0.50
0.30
0.23
0.63
0.94
1.6
6.0
3.7
14.
5.9
41.90
Runoff
Coeff.
(%)
45
29
14
4
2
21
21
22
99
68
201
174
Total
Flow
(106 m>)
22.
6.6
1.5
0.92
0.86
1.8
2.9
5.8
19.
16.
46.
20.
143.38
Incremental
Runoff
(cm)
4.5
0.86
0.19
0.12
0.11
0.24
0.38
0.39
3.9
-0.53tt
7.9
1.2
19.26
Runoff
Coeff.
(%)
32
10
5
1
1
6
7
19
65
-
117
36
tTotal
Flow
(106 m')
31.
8.3
1.9
1.2
1.1
2.3
3.7
6.5
26.
15.
61.
23.
181.00
Barge Canal
Estimated
Flow
(106 nv>)
14.
15.
14.
14.
14.
15.
16.
2.0
-
-
-
-
104.0
Total Discharge
from
Waterport Pond
(106 m')
45.
23.
16.
15.
15.
17.
20.
8.5
26.
15.
61.
23.
284.5
* Watershed precipitation data from two NOAA meteorological stations in Batavia and Albion and from project field lab located in Shelby, New York.
** "Incremental runoff" is the runoff for intermediate sections of the watershed, i.e., headwaters-Site 1, Site 1-Site 2, Site 2-Site 3. "Total
flow" is total measured flow at each site.
t Exclusive of Erie Barge Canal flow.
tt The negative incremental runoff value is the result of a time-lag in flow during a large end-of-the-month storm event.
-------
ranged from a low of 1 percent in August to greater than 100 percent in
March.
Runoff expressed as total monthly flow volume is shown for the three
sites in Figure 3. Release of water, impounded in the refuge pools prior to
the start of our study period in May 1984, accounted for about 20 percent
(5.6 x 106 in3) of the flow measured at Site 2 during the months of May
through July 1984. Additional releases from November 1984 through January
1985 of water impounded at various times of the year amounted to about 7.0 x
106 m3 (Whittemore, 1985). However, the bulk of the flow was recorded in the
months of January through April 1985, after the pools had been lowered and,
essentially, a flow-through system existed.
Mean annual discharge rates during the study period at Site 1, Site 2
and Site 3 were 0.79 m3/s, 4.5 m3/s and 5.8 m3/s, respectively. Maximum
instantaneous rates recorded during event flows at Sites 1, 2 and 3 were 21
m3/s, 37 m3/s and 59 m3/s, respectively.
The estimated monthly flow contributed to Oak Orchard Creek from the
Erie Barge Canal is also shown in Table 2. This diversion of water is meant
to supplement low Oak Orchard summer and autumn flows in order to maintain
power production at the two Niagara-Mohawk hydroelectric facilities. The
7-month input from the Barge Canal accounted for 36 percent of the annual
discharge at Waterport Pond during the study period, a significant
contribution.
Shown in Figure 4 is a curve constructed to estimate time-of-travel in
Oak Orchard Creek at different flow rates by extrapolating actual
measurements using velocity/flow relations derived from Boning (1974). At
the mean annual discharge rate for Site 2 of 4.5 m3/s, time-of-travel from
Site 1, a distance of 27 km, is estimated to be in the range of 220-340 h.
At the maximum recorded instantaneous rate of 37 m3/s, time-of-travel drops
to about 110-125 h, indicating that even during high flows, there is still a
considerable amount of residence time in Oak Orchard Swamp. Time-of-travel
downstream of the swamp to Site 2, a distance of 6 km, is less than 10 h
under high-flow conditions, however. During summer low-flow periods, there
are times when there is virtually no flow between Sites 1 and 2 due to the
extremely small slope in this stream section. In fact, under strong westerly
winds, visual observations were often made of surface water flowing upstream
at Site 1.
WATER CHEMISTRY
pH and Alkalinity
Nearly every water sample collected was analyzed for pH and alkalinity.
Table 3 lists the mean, range, and number of samples by site for each
parameter. There is a considerable range in pH over the year at most sites.
Generally, the lower pH values are associated with winter and early spring,
and the higher values with summer and early fall. pH values of 8 and above
in Waterport Pond (Sites 3, 5, 6) coincided with high summer phytoplankton
21
-------
•o 10
x
n
e „
SITE 1
MJJASONDJFMA
x
n
E
50
40
30
2O
K>
SITE 2
MJJASONDJFMA
60
50
20
1O
SITE 3
MJJASONDJFMA
1984 1985
FIGURE 3. Runoff at Sites 1, 2 and 3 for the study period.
22
-------
400
300
100
actual measurements
in this range
10 15 20 25
Site 2 Flow (mVs)
30
35
40
FIGURE 4. Estimated relationship of Site 2 discharge and time-of-travel
in Oak Orchard Creek between Sites 1 and 2.
growth which tends to remove C0? from the water and shift pH upward.
Similarly, alkalinity was variable during the year. Values greater than
200 itig/L CaCO_ were measured at Sites 1 and 2. The lowest alkalinities were
recorded during high flow periods in February and March, most likely a result
of dilution from snowmelt.
Phosphorus
A summary of the results of phosphorus analyses during the study period
is given in Table 4. The highest concentrations were observed at Site 1
which drains the surrounding cultivated mucklands. DMKP often accounted for
a major fraction of the total phosphorus measured at this site.
Phosphorus concentrations were generally reduced at Site 2 after the
creek leaves Oak Orchard Swamp, but can still be considered elevated when
compared to concentrations reported in other rural New York streams. At Site
2, also, a large portion of measured TP was DMRP.
Site 3 at the Waterport Pond outlet had somewhat lower concentrations
than the two previous stream sites. This is likely due in part to dilution
with flows from other tributaries and the Erie Barge Canal when it is
23
-------
TABLE 3. pH AND ALKALINITY MEANS AND RANGES OF WATER SAMPLES COLLECTED
DURING STUDY PERIOD.
SITE n
1 101
2 86
3 57
4 31
5e* 20
5h 19
6e 20
6h 20
Mean
7.1
6.8
7.2
7.4
7.7
7.6
7.7
7.3
Range
6.0
5.9
6.1
6.8
7.0
7.1
7.0
6.8
- 8.2
- 8.0
- 8.6
- 8.2
- 8.2
- 8.0
- 8.5
- 7.8
n
99
86
56
30
18
17
20
20
- ALKALI!
Mean
190
170
140
130
140
150
140
170
7-rrmr
Range
75 -
72 -
70 -
110 -
110 -
110 -
110 -
160 -
300
250
190
180
180
190
180
190
* "e" and "h" refer to the epilimnion and hypoliinnion of Waterport Pond
during stratification.
Note: Sites 5 and 6 were sampled from June 1984 through December 1984;
Site 4 from May 1984 through November 1984.
discharging to Oak Orchard Creek. A smaller fraction of the TP appears to be
DMRP at this site, in general, only about half.
Site 4 on the Erie Barge Canal had the lowest levels of phosphorus
observed in the system, and, on the average, only about 30 percent of the
total phosphorus was DMRP.
The two bridge stations on Waterport Pond, Sites 5 and 6, had phosphorus
concentrations comparable to those observed at Site 3, except for Site 6, .
The hypolimnetic waters at this location experienced increasingly higher
phosphorus concentrations as the summer wore on, resulting in an elevated
mean. Figures 5 and 6 show the changes in phosphorus concentration at Sites
5 and 6 during their sampling period. The considerable increase in
hypoliionetic concentration at Site 6, and to a lesser extent at Site 5 from
mid-July to mid-August, may reflect a release of phosphorus from bottom
sediments to the overlying waters.
24
-------
TABLE 4. TOTAL PHOSPHORUS AND DISSOLVED MOLYBDATE REACTIVE PHOSPHORUS
CONCENTRATION (mg/L) MEANS AND RANGES FOR SAMPLE SITES
DURING STUDY PERIOD.
SITE n
1 100
2 87
3 57
4 31
5 * 20
e
5. 20
n
6 20
e
6h 20
j.j.
Mean
0.59
(0.72)t
0.22
(0.19)
0.12
(0.13)
0.082
(0.081)
0.12
0.16
0.11
0.26
Range
0.19
0.10
0.058
0.031
0.077
0.079
0.072
0.075
- 1.6
- 0.55
- 0.26
- 0.12
- 0.18
- 0.28
- 0.16
- 0.72
n
95
87
56
31
19
19
20
19
Mean Range
0.37
(0.54)
0.14
(0.13)
0.062
(0.069)
0.027
(0.027)
0.070
0.088
0.062
0.19
0.098
0.027
0.013
0.009
0.031
0.039
0.021
0.039
- 1.4
- 0.46
- 0.16
- 0.054
- 0.12
- 0.18
- 0.10
- 0.60
t Flow^weighted means are in parentheses.
* "e" and "h" refer to the epilimnion and hypolimnion of Waterport Pond
during stratification.
Note: Sites 5 and 6 were sampled from June 1984 through December 1984;
Site 4 from May 1984 to November 1984.
Sewage effluent is discharged to Oak Orchard Creek and its tributaries
by three municipal wastewater treatment plants. Based on nine
weekly-composite samples, the mean effluent phosphorus concentration at the
Elba STP was 3.2 mg/L. For the Oakfield STP, the mean concentration was 1.3
mg/L, based on ten weekly-composite samples. The Medina plant averaged 1.7
mg/L based on eleven weekly-composite samples.
The U.S. Gypsum paperboard plant was not thought to be a significant
contributor of phosphorus to the creek because the majority of its annual
flow consists of relatively unpolluted groundwater in the form of mine
25
-------
SITE 5
Epilimnion
075
EOSO
o.
CHLg
JUKI JUL AUG SEP OCT NOV DEC
50,§
O
o
25—
1*050
O.
Hypolimnion
CHLg
JUN JUL AUG SEP OCT NOV DEC
SOE
O
O
20^
FIGURE 5. Time course of total phosphorus, dissolved oxygen and chloro-
phyll a concentrations at Site 5.
^050
TP
Epilimnion
DO
\CHLg
SITE 6
JUN JUL AUG SEP OCT NOV DEC
=ol
S
SO-
"i
20 S
15"
•
,§050
Hypolimnion
15 S
10 o
5 i
5
JUN JUL AUG SEP OCT NOV DEC
FIGURE 6. Time course of total phosphorus, dissolved oxygen and chloro-
phyll a concentrations at Site 6.
26
-------
drainage. The plant effluent was not sampled directly; however, four samples
for DMKP analysis were collected in the receiving stream approximately 100 m
downstream from the plant outfall during August 1984 as part of a special
study. The stream flow consisted almost entirely of plant discharge (includ-
ing mine drainage) at this time of the year, and samples were considered to
be fairly representative. The mean DMKP value of the four samples was 0.031
mg/L; the highest value was 0.050 mg/L. The only other data available were
reported by the company on a 1981 permit renewal application. At that time,
the results of 46 phosphorus analyses of effluent from the paperboard
wastewater treatment plant indicated a mean value of 0.23 mg/L with a maximum
value of 0.44 mg/L. It is unknown whether these values are reported as
phosphate (PpO,-) or as elemental phosphorus (P) .
Figure 7 illustrates the time course of total phosphorus concentrations
during the 1-yr study at the three primary sampling locations. There appears
to be a dampening of concentration variability over time as distance from the
jnucklands increases. At Site 1, extreme TP highs of 1.5 mg/L or more were
recorded during August and again in late February-early March. The lowest
concentrations observed at Site 1, about 0.20 mg/L, occurred during winter
non-event periods.
There was less variability over time in TP concentration at Site 2 which
exhibited a range of only 0.45 mg/L. Interestingly, the highest concentra-
tions were recorded during the summer and early autumn months, whereas the
lowest concentrations were observed during late winter-early spring,
non-event periods.
Total phosphorus concentrations at Site 3 displayed relatively little
variation during the study year, ranging from 0.058 mg/L to a one-time high
of 0.26 mg/L. The highest concentrations were observed during winter and
early spring. The lowest concentrations of both TP and DMKP in water leaving
Waterport Pond were recorded in July and early August.
Runoff events occurring during the winter-spring period resulted in-
significant changes in concentration and loading of phosphorus. The largest
event, in terms of flow, during the study period occurred from February 22 to
about March 4, 1985 and is analyzed in detail. With more than 130 cm of snow
on the ground, 2.90 cm of rain coupled with a wanning trend resulted in rapid
snowmelt and considerable runoff. The creek reached its highest recorded
level in the swamp as measured by a staff gauge at Knowlesville Road
(Whittemore, 1985).
The time courses of creek discharge, concentrations of phosphorus and
suspended sediment, and mass loading rates at Site 1 during this event are
illustrated in Figure 8. Suspended sediment concentration followed the rise
in creek flow and both peaked at about the same time. As TP concentration
rose with increasing suspended sediment concentration, DMKP decreased.
However, as flow lessened both TP and DMKP concentration began to rise,
peaking about 10 d after the event began. As is evident in Figure 8, by that
time almost all of the phosphorus was DMKP.
The greatest loading rate of phosphorus occurred at the peak of the
27
-------
16
14
1.2
10
0.8
06
O
04
02
-------
SITE
10
o
04
22 23 24 25 26 27 28 1 2 3 4
Feb Mar
22 23 24 25 26 27 28 1
Feb
234
Mar
•300
•a
o
o
J30
P
>
"10
0
22 23 24 25 26 27 28 1 2 34
Feb Mar
10,000
in
TJ
•o
V
1000
100
V)
22 23 24 25 26 27 28 1 2 3 4
Feb Mar
FIGURE 8. Changes in flow, phosphorus and suspended sediment at Site I
during the February 22-March 4 runoff event.
29
-------
hydrograph, but elevated loading rates were sustained for a considerable time
thereafter due to the increase in phosphorus concentration. The suspended
sediment loading rate rose several orders of magnitude to peak at the time of
peak discharge. During the 11-d event period (February 22 through March 4),
it is estimated that 5100 kg of total phosphorus and 500 tonnes of sediment
were transported through Oak Orchard Creek at Site 1. Approximately 78
percent of the phosphorus load was DMKP.
Characteristics of the same event at Site 2 are presented in Figure 9.
The hydrograph exhibited a smoother, less rapid rise than at Site 1, and it
remained elevated for a longer period. Concentrations of TP rose slightly
with the small increase in suspended sediment concentration, then continued
to rise as flow decreased, similar to the pattern observed at Site 1. The
changes in suspended sediment concentration were relatively insignificant
compared to the large increase measured at Site 1. Loading rates of TP and
DMRP peaked about 2 d after the peak flow rate, whereas the highest suspended
sediment loading rate occurred ahead of peak flow.
For the purposes of better defining transport and delivery of nutrients
and sediment from Site 1 to Site 2 during a single event, a comparison of
mass loads as measured at each location is needed. However, before the
hydrographs could return to pre-event levels for this particular event, more
rainfall and runoff on March 4 caused a rise again. As a way of estimating
the total load at each site that would have been produced by this event had
conditions remained stable, the hydrographs were extrapolated along the slope
of the falling limbs according to the hydrograph separation method of Chow
(1964) until pre-event flow rates were reached. Average daily flows were
computed for the extended time period and multiplied by assumed concentrations
to generate average daily loads. In this manner, the questions of
interference from subsequent events and time-lag between the two sites are
addressed.
Employing this method, the Site 1 hydrograph would have continued
falling for 6 d more before reaching the pre-event flow, while the Site 2
hydrograph would have continued for another 13 d. Total phosphorus and
suspended sediment loads estimated for the entire single event at Site 1
would then be 5700 kg and 500 tonnes, respectively, compared to 11-d actual
measurements of 5100 kg and 500 tonnes. At Site 2, the total phosphorus load
would have been 6900 kg and the sediment load 220 tonnes, compared to 14-d
actual measurements of 5400 kg and 180 tonnes. Clearly indicated is a large
net loss of sediment to Oak Orchard Swamp which lies between Sites 1 and 2.
In contrast, there appears to be little attenuation of phosphorus between the
two sites. The additional 1200 kg of phosphorus at Site 2 presumably is
attributable to drainage downstream of Site 1.
The time course of phosphorus loading in kg/d is shown for the three
study sites in Figure 10. At Site 1, the loading rate approached zero almost
continuously from June through December. High-flow periods are indicated by
the peaks in May, December-January, and February-April. The highest loading
rate was measured at-nearly 900 kg/d during the February 22 to March 4 event.
At Site 2, the extreme highs in loading rate seen at Site 1 were not as
30
-------
SITE 2
100
o
10
23 24 25 26 27 28 1 2 3 4 5 6 7 8
Feb Mar
-, °25
1*020
o 015
010
005
TP
20-J
o>
10 "
o
5 ci
23 24 25 26 27 28 1 2 3 4 5 6 7 8
Feb Mar
1500
25
§15
23 24 25 26 27 28 1 2 34 5 6 78
Feb Mar
3
1000
o
V)
-g 500
•o
23 24 25 26 27 28 1 2 345 678
Feb Mar
FIGURE 9. Changes in flow, phosphorus and suspended sediment at Site 2 during the February 22-
March 4 runoff event.
-------
evident. Certainly, event periods are distinguishable, but overall, loading
rates were less variable during the study period. Unlike Site 1, there was
always some flow at Site 2 producing some loading.
Daily loading rates at Site 3 were about as variable as at Site 1.
Event periods are clearly discernible. There is a close correlation (r =
.96) between daily flow rate and phosphorus loading rate due to the rela-
tively small variation in phosphorus concentration. Flows during the summer
months were sustained by the Erie Barge Canal diversions. Though this water
had a low phosphorus concentration when compared to Oak Orchard Creek, it
provided the bulk of the flow (83 percent) and phosphorus load (62 percent)
to Waterport Pond for the period June through November.
Monthly loads of phosphorus at each of the three primary study sites are
compared in Figure 11. The annual total phosphorus load at Site 1 was 18,000
kg which is equivalent to 1.7 kg/ha on s unit watershed basis. The months of
February and March accounted for 55 percent of this annual load. Seventy-
five percent of the total phosphorus loaded at Site 1 was as DMRP, while
about 16 percent was associated with suspended particulate matter. The
remaining 9 percent was dissolved but unreactive.
The total load measured at Site 2 was 27,000 kg, a 52 percent increase
from Site 1. Contrastingly, watershed drainage area increased 280 percent
between the two sites. Annual loading for the area between the two sites was
reduced to 0.32 kg/ha. Major phosphorus inputs in this watershed section
include the remaining 600 ha of cultivated Elba muck and the Oakfield
municipal wastewater discharge via Tributary 138-28-3d. The fraction of the
total phosphorus load measured as DMRP dropped to about 67 percent at Site 2.
Particulate phosphorus accounted for 22 percent of the load, while the
dissolved unreactive fraction made up the remaining 11 percent. Fifty-nine
percent of the total phosphorus was loaded during the months of March, April
and May.
The same three nonths accounted for 50 percent of the load leaving
Waterport Pond (Site 3), although the other 50 percent was more evenly
distributed over the remaining months than at the previous two sites.
Between Sites 2 and 3, a drainage area increase of 51 percent, the phosphorus
load increased 37 percent. Dissolved molybdate reactive phosphorus was
reduced to 54 percent (20,000 kg) of the total load, while dissolved
unreactive phosphorus and particulate phosphorus increased to 19 and 27
percents, respectively. The annual loading rate in the section between Sites
2 and 3 is equivalent to 0.50 kg/ha of total phosphorus on a unit area basis.
Monthly loads of total phosphorus were always greater at Site 3 than at
Site 2, except for April 1985. This was not the case for DMRP. During the
months of July through January, the monthly loads of this form were greater
at Site 3 than at Site 2, while during February through June they were less.
The results of a synoptic survey on May 13, 1985, of flow and phosphorus
concentration in Oak Orchard Creek and its tributaries under steady-state,
low-flow conditions revealed no previously unsuspected sources or sinks of
phosphorus. Loading rates determined for the tributaries and at a number of
32
-------
SITE 1
li i ii n
MJJASONDJFMA
\ 600
o>
JC
^* 500
400
= 300
o
v>
O
£
200
100
O 0
SITE 2
MJJASONDJFMA
9OO
-« 800
-o
;§" 70°
"§ 600
3
500
c/>
I 400
£ 3°°
•5 200
10O
0
SITE 3
MJJASONDJFMA
1984 1985
FIGURE 10. Phosphorus loading rates at Sites 1, 2 and 3 during study period.
33
-------
SITE 1
MAY
3400 | 2«00
(19%)
JUN-NOV
620
(4 %)
DEC
770
(4%)
FEB
3900
(22%)
TP LOAD = (8,000 Kg
DMRP LOAD = !4,OOOKg
SITE 3
APR
2900
(8%)
MAY
5300
(14%)
SITE 2
OCT
1400
(4%)
DEC
NOV 880
2200 (2%)
(6%)
TP LOAD = 37,OOOKg
DMRP LOAD = 20,OOOKg
APR
3500
(13%)
MAR
8800
(33%)
MAR
10,000
(28%)
JUN
2200
(8%
FEB
2200
(8%)
(12%)
TP LOAD = 27,OOOKg
DMRP LOAD= 18,COOKg
DURP
JAM
4100
(11%)
DMRP
FIGURE 11. Monthly loads of phosphorus at Sites 1, 2 and 3. "PP" is particu-
late P, "DURP" is dissolved unreactive P, and "DMRP" is dissolved
molybdate reactive P.
34
-------
locations on Oak Orchard Creek provided information on the sources and
delivery of phosphorus through the system. On that particular day, both TP
and DMKP instantaneous loading rates in Oak Orchard increased until Site 2,
then decreased between Site 2 and Site 3. Continuous sampling and monitoring
throughout the study year have already shown a consistent loading increase
from Site 1 to Site 3. The decrease in loading rate apparent during the
synoptic survey reflected a single day's sampling conducted under conditions
of very low stream flow before Erie Barge Canal additions began.
Temperature, Dissolved Oxygen and Chlorophyll a
Temperature, dissolved oxygen and chlorophyll a^ concentrations were
monitored at Sites 5 and 6 from June to December 1984. Temperature and
dissolved oxygen isopleths are shown in Figures 12 and 13. At both stations
a period of hypolimnetic anoxia was apparent; however, it was more extensive
and longer-lasting at the deeper, more stratified Site 6. Isothermal
conditions were reached at Site 5 by late October and at Site 6 by
mid-November, indicating the completion of fall overturn.
Temperature (°C)
JUN JUL «O6 SEP OCT NOV DEC
7
Dissolved Oxygen (ppm)
•WN
JUL AUG SEP OCT NOV DEC
FIGURE 12. Temperature and dissolved oxygen isopleths for Site 5.
35
-------
0
2
4
I6
fe
10
12
16
Temperature (°C)
m~~r
JUL
AUG
SEP
OCT
NOV
DEC
Dissolved Oxygen (ppm)
OCT
NOV
DEC
FIGURE 13. Temperature and dissolved oxygen isopleths for Site 6.
Chlorophyll a. reached peak concentrations of about 30 ug/L in late
August and again in late September at Site 6 (Figure 6). The highest value
observed at Site 5 was 16 ug/L in mid-June ^Figure 5). Three grab water
samples collected at the Waterport Pond outlet (Site 3) on May 3, 20, 31,
1985, had respective chlorophyll a concentrations of 43, 42 and 10 ug/L.
The higher values may indicate that in Waterport Pond spring diatom blooms
are more significant in terms of phytoplankton populations than the summer
algal species. Except for the two aforementioned peaks and the May 1985
values, observed chlorophyll a. concentrations did not exceed 15 ug/L. Based
on Mikol's (1982) trophic classification of New York lakes which categorizes
eutrophic lakes as those having chlorophyll a concentrations greater than 12
ug/L, Waterport Pond can be characterized as moderately eutrophic.
36
-------
PESTICIDES
The pesticides of main concern were DDT (dichloro-diphenyl-trichloro-
ethane) and its metabolites because of past heavy usage in the study area and
the chemicals' long-term persistence in the environment. In 1974, sediment
grab samples were collected at two stream locations within the mucklands and
analyzed for fourteen pesticides by the USEPA Rochester Field Office.
Samples collected from the Manning Tributary just downstream from cultivated
fields contained low levels of EDDT (27 ug/kg dry weight). The sample
collected from the Oak Orchard Creek main stem at a point draining both the
Elba and Manning mucks contained substantially higher levels of the ZDDT (595
ug/kg dry weight) (SCS, 1975).
For this study, pesticide analyses were performed on three soil samples
(one from the Manning muck, two from the Elba muck), three suspended sediment
samples collected at Site 1 during events, three creek bottom sediment
samples collected at Site 1 and in Oak Orchard Swamp, and two event water
samples. Results of the analyses are given in Table 5.
Clearly, the highest concentrations were found in the cultivated muck
soil samples. At these sites much more DDT was present than either of the
two weathered forms, DDE and ODD. As expected because of its low solubility,
none of the pesticide forms were detected in water samples.
In suspended sediment, the greatest concentrations were of DDT, although
DDE was present in smaller amounts, while ODD was detected only once. There
was considerable difference between the two years; the Site 1 suspended
sediment sample collected on February 24, 1985 contained concentrations about
a magnitude lower than the previous year's two samples.
A grab sample of surface bed sediment from Site 1 contained small
amounts of all three compounds, with ODD found in the greatest concentration.
Surface and subsurface bed sediment collected in Oak Orchard Swamp (about
midway between Sites 1 and 2) contained very small amounts of DDE and ODD,
with no detectable amounts of DDT.
There was no presence of DDT or its metabolites above the detection
limit in either of the two water samples collected during event flows.
Unless in fairly high concentrations, with the method used it is often
difficult to detect in an aqueous solution substances with a strong affinity
for sorbing to particulate matter, such as DDT.
37
-------
TABLE 5. CONCENTRATIONS OF DDT AND METABOLITES IN MUCK SOILS
AND IN OAK ORCHARD CREEK SEDIMENTS AND WATER.
SITE
Manning Muck
Elba Muck 1
Elba Muck 2
Site 1 (3/22/84)
Site 1 (3/23/84)
Site 1
Oak Orchard Swamp
Oak Orchard Swamp
Site 1 (2/24/85)
Site 1 (2/24/85)
Site 2 (2/24/85)
SAMPLE TYPE
Field soil
Field soil
Field soil
Susp. Sed.
Susp. Sed.
Surface Bed Sed.
Surface Bed Sed.
Subsurf. Bed Sed.
Susp. Sed.
Water
Water
COMPOUND CONCEINiTRATION (ppb)
p,p'DDT p,p'DDE p,p'DDD
5700
5500
3400
740
1100
13
ND
ND
100
ND
ND
ND
ND
860
320
560
26
4
7
30
ND
ND
400
ND
920
ND
ND
70
4
5
40
ND
ND
ND = Not detected
NOTE: Method detection limits were at least 10 ppb for water
and 16 ppb for soils.
38
-------
SECTION 6
DISCUSSION
PHOSPHORUS, SEDIMENT AND PESTICIDE LOSSES FROM CULTIVATED MUCKLANDS
Past research has shown that artificially-drained, cultivated mucklands
are often significant sources of phosphorus, particularly the dissolved
reactive form. Studies of the Elba muck indicate that up to 30 kg/ha of DMRP
can be leached from the soil and lost to drainage waters. Transport of this
phosphorus to nearby bodies of water can lead to nutrient enrichment and
advanced eutrophication.
One of the main objectives of this study was to assess the importance of
the mucklands in the total phosphorus loading from Oak Orchard Creek to Lake
Ontario by extending the scope of Peverly's (1982) work. More intensive
monitoring and evaluation was undertaken in this study, although the results
generally support many of the conclusions drawn by Peverly regarding the
mucklands as a significant contributor of phosphorus.
Peverly's study years of 1977 to 1979 were somewhat wetter, having on
the average about 25 cm/yr more precipitation than our study period. Total
flow at Peverly's muckland stream site (see Figure 2) (drainage area of 56.7
km2) both study years was about 40 x 106 m3, compared to our flow of 25 x 106
m3 at Site 1 2.4 km downstream (drainage area of 102 km2). It is interesting
to note, however, that during his first study year (1977-1978), Peverly
calculated an event period (January through March) flow of 13.4 x 106 m3
which is very close to our flow estimate of 14. x 106 m3 at Site 1 for the
same months in 1985. Similarly, his estimated total annual DMRP load was
about 14 tonnes, as was ours. In contrast, during his second study year
Peverly calculated a January through March flow of about 30 x 106 m3, more
than twice the first year's, with a resultant annual DMRP load of 43 tonnes,
more than three times the first year's. These data support the contention
that flow volume during the critical months of January tlirough March is more
significant in determining annual loading frcm the mucklands than is total
annual precipitation or total annual flow volume.
Peverly (1982) calculated areal loading rates for each site by dividing
the difference in loads between successive sites by the watershed drainage
area between sites. For his upstream site (see Figure 2) (drainage area =
19.4 km2), which drained mineral soils, the DMRP areal loading rates for the
two study years ranged from 0.82 to 0.84 kg/ha. Peverly's loading rates for
the downstream section (drainage area =56.7 minus 19.4 km2) which, without
the upstream drainage, represented muckland input almost exclusively, ranged
from 3.36 kg/ha in 1978 to 11.1 kg/ha in 1979, the greater load year. Our
Site 1 (total drainage area = 102 km2) included a larger muckland area and
39
-------
received flow from the Manning muck as well as additional upland area.
Employing Peverly's areal DMRP loading rate for Oak Orchard mineral soil
upland of 0.83 kg/ha and multiplying by the approximately 7450 ha of upland,
predominantly dairy farms, drained by Site 1, yields a DMRP load of 6200 kg
associated with this source. Subtracting this amount frcm the total annual
DMRP load of 14,000 kg measured at Site 1 gives a load of 7800 kg associated
with the 2750 ha of muckland, or an areal rate of 2.8 kg/ha which is similar
to Peverly's first year estimate of 3.36.
In a study of nutrient losses in Delaware County, New York, Brown et al.
(1983) estimated the areal DMRP loading rate for a small dairy farm
sub-watershed to be 0.34 kg/ha, a somewhat lower value than Peverly's 0.83
kg/ha. Employing this rate at Site 1 yields an upland load of 2500 kg, a
muckland load of 12,000 kg and a muckland areal rate of 4.4 kg/ha. In either
case, the mucklands appear to account for between 56 and 86 percent of the
DMKP load measured at Site 1, but represent only 27 percent of the drainage
area.
Dissolved phosphorus was determined to account for 84 percent of the
total load delivered at Site 1 during the study year. This is quite a large
fraction when compared to amounts measured in other New York watersheds.
Brown et al. (1983) report that DP comprised 46 percent of the total
phosphorus load measured in the West Branch Delaware River and 40 percent of
the load measured in a tributary within the WBDR basin. Bouldin et al.
(1975) estimate dissolved phosphorus accounted for about 30 percent of the
total phosphorus load measured in Fall Creek (Tompkins County, New York)
during their 20-month study period. The lack of significant soil erosion
from overland flow and the predominance of leaching processes associated with
subsurface drainage in the mucklands likely contribute to the large dissolved
phosphorus load observed at Site 1.
Particulate phosphorus comprised 2900 kg (16 percent) of the total
annual load at Site 1. The suspended sediment load for the year was
estimated to be 1100 tonnes. The average sediment-phosphorus concentration
(particulate phosphorus load divided by sediment load) was calculated to be
2670 mg/kg. Actual measurements by serial extraction of total
sediment-phosphorus in suspended sediment collected at Site 1 during events,
ranged frcm 1360 to 1650 mg/kg. The concentration of serially-extracted
phosphorus in surficial bed sediment collected at five locations up to 15 km
downstream of Site 1 averaged 847 mg/kg and ranged from 588 to 1340 mg/kg.
Bed sediment collected by Peverly at his muckland site contained, on the
average, 800 mg/kg of phosphorus (Peverly, 1985) based on a dry ashing and
hydrofluoric acid digestion method..
It has been shown that sediment-phosphorus concentration is directly
related to sediment surface area (Brown et al., 1983) which is largely a
function of percent clay. Therefore, concentrations lower than the average
sediment-phosphorus concentration of 2670 mg/kg are expected in suspended
sediment collected during high flow periods and in bed sediments because a
greater percentage of sand- and silt-sized particles generally comprise these
sediments resulting in smaller surface areas per unit sediment mass (see
Equation 1). During low-flow months, when it is expected that only fine,
40
-------
predominantly organic material was being transported, average monthly
sediment-phosphorus concentrations at Site 1, determined by the ratio of
monthly particulate phosphorus load to sediment load, were quite a bit higher
than the average annual concentration, ranging from 3000 to 15,000 mg/kg.
In Figure 14, a high correlation (r = .98) between phosphorus
concentrations in suspended and bed sediments and surface area is illustrated
for samples collected in the non-calcareous West Branch Delaware River (WBDR)
watershed (Brown et al., 1983). Plotting Oak Orchard samples on this graph
for comparison reveals that Site 1 event suspended sediments and some bed
sediments contain about as much phosphorus as WBDR suspended sediments having
considerably greater surface areas. Oak Orchard bed sediments contain at
least twice as much phosphorus as WBDR bed sediments having the same surface
areas. With respect to the WBDR sediments, it appears then that sediments
from Oak Orchard Creek are considerably enriched with phosphorus.
Relatively little of the suspended sediment transported past Site 1 is
thought to derive from the mucklands themselves because of the erosion
2000
1500
o>
x
o
o 1000
O
0
E
500
o o
A A
WBDR SS •
WBDR BS A
SITE 1 SS O
OAK ORCH. BS
Surface Ar*a
3 4
(cm2/g x 10 )
FIGURE 14. Relationship between phosphorus concentration in sediments
and sediment surface area (r=.98, p<.01). WBDR data adapted
from Brown et al., 1983.
41
-------
control and drainage structures installed in the system, although at times,
wind erosion can be a factor in removing soil from the fields. Likely, the
major source is "clean" mineral soil eroded from sloping upstream areas.
Upon entering the deep, channelized stream sections in the mucklands, much of
this sediment probably settles out and is loosely deposited on the stream
bed. Resuspension and transport can occur during high-velocity event periods
allowing mixing in the creek with phosphorus-rich muckland drainage waters.
Subsequent adsorption of phosphorus by the sediments would likely occur under-
time se conditions until an equilibrium between dissolved and sediment
phosphorus is reached. The adsorbed phosphorus may become permanently
mineralized by combining with calcium, aluminum and iron, or remain in a
labile form, only to be desorbed under different conditions of instream
phosphorus, oxygen and sediment concentrations. Sediments, then, which
derived from upland, mineral soils and contained relatively little
phosphorus, eventually, after perhaps a number of settlings and
resuspensions, leave the muckland stream section considerably enriched as
suggested by the sediment data presented here.
Instream phosphorus concentrations rose considerably during the month of
August at Site 1 (Figure 7) . Similar high summer concentrations were
observed by Peverly (1985) at his muckland stream site in 1978 and 1979.
This suggests a release of phosphorus from the sediments as water becomes
anoxic near the bottom due to a virtual absence of flow. Also, any drainage
from the mucklands would receive very little dilution from upstream flow,
presumably minimal, during this period.
At Site 1, instreain phosphorus concentrations were observed to vary over
the course of runoff events (Figure 8) . Phosphorus concentration (C ) can be
considered as the sum of the mass loading rates (W.) from various sources
divided by the sum of hydraulic loading rates (Q.) of those sources:
n n x
C = ( Z W.) / ( I Q.) . [2]
r i=l x i=l x
The instantaneous phosphorus loading rate W. can be described as Q.C.
where C. is the phosphorus concentration in the i source. In simplest:
terms, instream phosphorus concentration during a runoff event can then be
viewed as the result of contributions from combined surface and subsurface
runoff (Q ) and groundwater runoff (Q ) sources:
s g
Substituting the relationship of Q = Q + Q into Equation [3] yields, after
rearrangement of terms:
Ct = [(Cs " V ~ ] + Cg ' [4]
Qt
From this equation it can be seen that a chemical with a greater
concentration in groundwater than in combined surface and subsurface runoff
(C > C ) will increase in stream concentration as the proportion of surface
ana subsurface runoff in total discharge decreases.
42
-------
In fact, this very phenomenon was observed at Site 1 during the large
February-March 1985 event. Using the hydrograph separation technique of
Chow (1964), ratios of Q /Q during hydrograph rise and recession were
determined and regressed Igainst instream phosphorus concentrations (Figure
15) revealing an inverse function. Instream phosphorus concentrations rose
as the hydrograph receded or as the ratio of surface to total discharge
decreased. A similar relationship was observed in the West Branch Delaware
River for reactive silica (Brown et al., 1983), a substance which is
generally slowly leached from soil and delivered in groundwater flow.
Conversely, WBDR instream phosphorus concentrations during events lessened as
the proportion of surface runoff decreased. In terms of event loading, our
results indicate a greater significance associated with delayed subsurface
and groundwater runoff than with surface runoff in Oak Orchard Creek at Site
1 due to phosphorus leaching from cultivated muckland soils.
Most of the phosphorus loading at Site 1 occurred during the few
critical winter-spring months of intensive snowmelt and rainfall. Comparison
to results of Peverly (1982) suggests that annual loads could be considerably
higher than that measured in this study, depending on the ground condition,
snow cover and rainfall in those critical months. The cultivated mucklands
not only represent a significant source of dissolved phosphorus per unit
drainage area, but appear to influence sediment phosphorus enrichment through
various instream physical and chemical processes as well.
The results of this study support the use of management practices that
limit fertilizer-phosphorus additions and prevent surplus phosphorus build-up
in the muck soils. Fertilizer application rates of nearly 50 kg/ha P are
currently used on the Elba muck (Cogger and Duxbury, 1984). In many cases,
there is virtually no plant response to these additions due to the abundant
reserve of available phosphorus already present in the soil (see Volume 2).
A fertilizer management program for the mucks would rely heavily on soil
testing which provides an accurate indicator for determining most plant
nutrient requirements. It is believed that, over time, leaching losses and,
hence, transport to the creek, would gradually lessen as reduced amounts of
phosphorus are added to the soils. Specific plant response studies and
A single hydrograph separation, as was used here, breaks the runoff into
its groundwater component and its combined surface and subsurface component.
Chow's method works best with an ideal situation, what he terms "virgin
flow", i.e. stream flow unaffected by artificial works of man on or in the
stream or watershed. The mucklands, with their tile drain systems and stream
channelization projects, present a less than ideal situation. We suggest
that the bulk of drainage from the mucklands behaves more like groundwater
runoff, percolating down to deeper depths and discharging at a slower rate
through tile drains, rather than like subsurface runoff which mostly
infiltrates the upper soil horizons and moves fairly rapidly in a lateral
direction toward the stream as shallow, perched groundwater above the main
groundwater level. Therefore, in this analysis, Q is seen as representing
surface runoff plus a small amount of shallow subsurface runoff, while, for
the muckland areas at least, Q represents a true groundwater component plus
a delayed subsurface input. "
43
-------
S I TE 1
1.50
1.25
1.00
U>
E
o
z
o
o
.75
.50
.25
1.25
^ 1.00
Ol
E
.75
00
r=-.79
p
-------
STREAM TRANSPORT OF PHOSPHORUS AND SEDIMENT THROUGH OAK ORCHARD SWAMP
The results of this study strongly implicate the cultivated inucklands as
the major factor in controlling phosphorus loading at Site 2. An analysis of
the February-March 1985 event at this site shows the concentration of phos-
phorus increasing as the ratio of combined surface and subsurface to total
runoff decreases (Figure 16), the same relationship observed at Site 1 (see
Figure 15). Because these two sites exhibit similar concentration/flow
responses, there is evidence that during the high-flow, late winter-early
spring period, at least, phosphorus is conservatively transported through the
swamp with little loss to bed sediments or through aquatic plant uptake.
To some extent, the swamp undoubtably mediates phosphorus and sediment
loading at Site 2, acting as either a sink or a source, depending on flow
conditions and season. Although we did not specifically monitor to determine
effects from the swamp, certain inferences may be made based on the data we
did collect and work done by others.
SITE 2
.25
.20
O)
o .15
o
O
.10
.05
.30
.25
.20
Dl
.15
o
O
.10
.05
r=-0.79
p
-------
Annual net increases in both total phosphorus and DMRP loads of 52 and
34 percents, respectively, were measured from Site 1 to Site 2 providing
additional evidence of conservative transport in this stream section.
Peverly (1982) observed a net loss in DMRP and nitrogen in his second study
year only, when annual flow at Site 2 was about one-half that in his previous
study year. However, he did not address the impoundment of water in the
refuge during late spring which may have accounted for some of the flow and
load deficit when released in late fall after his study period had ended.
An annual net loss in sediment load was apparent in this study, reduced
from 1100 tonnes at Site 1 to 860 tonnes at Site 2. Most of this loss
occurred at the end of February 1985 during the large runoff event when
extremely high concentrations of suspended sediment were measured at Site 1
but were not evident at Site 2. A certain amount of settling is expected,
especially of the larger particles suspended and transported by high-velocity
flows, considering the low creek gradient and lengthy time-of-travel through
the swamp. Employing the average sediment-phosphorus concentration derived
for Site 1 suspended sediments, the particulate phosphorus loss associated
with sedimentation during the study period would be about 600 kg.
Instream phosphorus concentration generally decreased from Site 1 to
Site 2 presumably through dilution by incremental flow entering downstream of
Site 1, rather than through retention by the swamp. Even though the flow
balance calculation indicates a large water volume contributed by incremental
flow, the phosphorus load associated with it is not considered to be particu-
larly large. Allowing conservative transport of the Site 1 total phosphorus
load, the load at Site 2 from all other sources is 9200 kg (27,000 kg minus
18,000 kg). A separation of incremental loading into low-flow months, June
to November 1984, and high-flow months December 1984 to April 1985 (May 1984
was excluded because, for the purposes of this analysis, it belongs in the
high-flow period of the previous year) gives some clue as to possible
mechanisms operating in the swamp which affect loads measured at Site 2.
During the high-flow months given, the net increase in phosphorus load
between Site 1 and 2 was 4800 kg, while the increase in flow was 88. x 106 m3
at a time when water was not being impounded in the refuge. Flow-weighted,
average total phosphorus concentration (load divided by flow) in this
incremental flow would then be 0.054 mg/L, which is within the range expected
in stream runoff from a rural watershed. In the low-flow months, load
increased by 3900 kg between sites but flow by only 13. x 106 m3, resulting
in a flow-weighted, average total phosphorus concentration of 0.30 mg/L.
Generally, streams whose phosphorus sources are solely nonpoint-derived
exhibit a pattern of high phosphorus concentrations associated with
high-flow, surface runoff events and low concentrations with low (base) flow.
Streams which receive loads predominantly from constant point source
discharges of phosphorus tend to display an inverse relationship between
concentration and flow due to effects of dilution. Loads from the Elba and
Oakfield STPs, if conservatively transported to Site 2, would account for an
average phosphorus concentration in incremental flow of 0.005 mg/L during
high-flow months and 0.030 mg/L in low-flow months. Thus, the incremental
flow load appears predominantly nonpoint source-controlled and one would not
expect to observe concentrations rising during low-flow periods unless
46
-------
another source, insignificant during high flows, was present and acting like
a point source.
Internal loading through phosphorus release from bottom sediments and
decomposing plant material, may explain the point source-like increase in
incremental flow phosphorus concentration observed during low-flow months.
In simple terms, sediments in contact with oxygenated overlying waters may
either adsorb or desorb phosphorus depending on the concentration gradient
between labile sediment-phosphorus and dissolved phosphorus (MacAllister and
Logan, 1978; Logan, 1982). In anoxic waters with low redox potential,
insoluble ferric phosphate in the sediments is reduced, liberating dissolved
phosphorus and changing the equilibrium between sorbed and interstitial water
phosphorus (Meals and Cassell, 1978). Daily phosphorus release rates
reflecting different sediment composition, past nutrient loading, present
water body trophic state, and oxygen concentration in the sediment-water
interface have been reported to range from <0.5 to >100 mg/m2, with the
higher rates generally reflecting anaerobic conditions (Fillos and Swanson,
1975; Frevert, 1980; Funk et al., 1980; Mawson et al., 1983).
Although actual measurements of dissolved oxygen in Oak Orchard Swamp
were not made in this study, Peverly (1985) noted that in summer months creek
bed sediments were anaerobic and warm, and Whittemore (1986) indicated that
anoxic conditions probably exist in the drawn-down refuge pools in low-flow
months. A daily phosphorus release rate of 1 to 2 mg/m2 from the estimated
total bottom surface area of stream bed plus drawn-down pools of 9.5 km2
would account for much of the load measured at Site 2 which was not
associated with incremental flow during the 6-month low-flow period. This
range of release rates is similar to some of the lower estimates reported by
the previously cited investigators.
Return of phosphorus to the system through plant senescence and
decomposition may contribute to the load measured at Site 2 in low-flow
months. Some wetland studies have documented a large, end-of-the-growing-
season phosphorus release following decomposition (Klopatek, 1975). Peverly
(1985) estimated nutrient accumulation and release by aquatic macrophytes
rooted in the main stream bed of Oak Orchard Swamp. The potential contri-
bution of phosphorus to the creek from instream plant senescence was
determined to be about 28 percent of the creek load in autumn; however, this
amounted to only 1 to 2 percent of the total annual load, the bulk of which
is delivered during spring thaw. Peverly did not assess the phosphorus
contribution from refuge pool vegetation, a certain amount of which would be
expected to reach Site 2 and add to the load. Overall, the annual effect
from wetland plants in this system is thought to be small in light of the
sizable load delivered from the mucklands. Seasonally, the load always
increased between Sites 1 and 2 supporting the contention that the swamp is
not permanently removing phosphorus on a large-scale basis and may actually,
in fact, be a source of the nutrient at certain times of the year.
A consideration when evaluating water quality in the Oak Orchard Creek
watershed is the possible effects large gatherings of waterfowl in the
refuges could have on nutrient loading. Previous studies investigating
impacts on water quality were conducted in Montezuma NWR (Have, 1973),
47
-------
located in central New York, and in Bosque del Apache NWR (Brierly et al.,
1975a; 1975b) in New Mexico. At Bosque del Apache, total nitrogen and
dissolved phosphorus concentrations were, respectively, 15 and 39 percents
higher in the waterfowl pond area than in inflowing water. Total phosphorus
concentrations in the pond were actually 43 percent lower than in inflowing
water, however. Water leaving the refuge had concentrations of nutrients
similar to or lower than those in the incoming water, on the average.
At Montezuma, concentrations of nitrogen and phosphorus at the refuge
outlet were generally higher than in the two inflowing creeks from August
through November. Nutrient levels did not seem to be affected by excreta
from bird populations, but rather reflected conditions in the marsh. High
concentrations of plant and animal life in the refuge corresponded to
increased concentrations of phosphorus. It was suggested that continuous
availability of phosphorus for microbial growth was accomplished by various
forms of mechanical mixing of the bottom sediments, with subsequent
phosphorus release, through the actions of carp, foraging waterfowl and
macroinvertebrates. Phosphorus levels were at their lowest during periods of
ice cover when microbial activity and mechanical mixing had the least effect.
As water temperature decreased with approaching winter, total nitrogen and
phosphorus declined because of reductions in microbial populations. Nitrogen
and phosphorus levels were of the same magnitude at the outlet as in the two
creeks until May. Apparently, the presence of birds on the refuge during
winter-spring months had no significant effect on nutrient concentrations in
the water.
An attempt was made to calculate the amount of phosphorus that would be
excreted by waterfowl visiting Iroquois NWR and Oak Orchard WMA in order to
determine the relative significance of this source to annual loading in Oak
Orchard Creek. Chandler (1986), Iroquois Refuge Manager, provided estimates
of average waterfowl populations and number of weeks these populations are
sustained in Iroquois during certain times of the year. Table 6 summarizes
this information and gives estimates of the phosphorus associated with the
visiting birds. Oak Orchard WMA supports populations approximately half of
those at Iroquois (Chandler, 1986). The total phosphorus excreted by
waterfowl utilizing the two refuges would then be about 5 tonnes.
Not all of a bird's time will be spent in the refuge or even in the
watershed, however. Geese frequently feed in upland areas, particularly
cornfields. Chandler (1986) estimates that during migration periods geese
and ducks spend 35-40 percent and 75 percent, respectively, of their time in
the refuge. This would reduce the amount of phosphorus excreted in the
wetlands to about 3 tonnes. At this point, it is impossible to assess how
much of this quantity might be delivered to Oak Orchard Creek. Some of the
loading from swamp bottom sediments may represent this source. When compared
to the phosphorus load derived from the mucklands, the potential contribution
from visiting waterfowl is relatively insignificant and certainly not
controllable.
A brief analysis of the U.S. Gypsum plant data indicates a small load
associated with this source. Using the highest DMRP concentration of 0.050
mg/L that we measured in the tributary receiving wastewater (including mine
48
-------
TABLE 6. ESTIMATES OF PHOSPHORUS IN EXCRETA PRODUCED BY VISITING
WATERFOWL AT IROQUOIS NATIONAL WILDLIFE REFUGE.
Goose Avg. Duck Avg. Total
Time of Year Population
Fall Migration 7000
Spring Migration 30,000
Simmer 900
adults
500
juveniles
Weeks
7
7
16
6
*kg P
377
1615
111
23
Population
10,000
12,000
2,500
adults
3,500
juveniles
Weeks
9
9
16
6
kg P
433
519
192
101
kg P
810
2134
303
124
Total 3371
* Porcella et al. (1974) give a value of 0.35 kg P/bird produced annually by
domestic ducks which generally are considerably larger than wild ducks.
Therefore, a value of 0.25 kg P/bird was assumed for wild ducks and 0.40 kg
P/bird for wild geese.
drainage) and the average flow of 30,000 m3/d reported for the study period
yields a DMRP load of 500 kg annually. This would account for less than 3
percent of the DMRP load measured at Site 2. Employing the average effluent
phosphorus value reported on the plant's permit application renewal, assuming
it was reported as P not P2°5' and the avera9e plant discharge (excluding
mine drainage) of 4000 m3/a, would yield an annual TP load of 340 kg. This
compares well with the DMRP load computed from the stream sampling. The
estimated TP load would account for less than 2 percent of the load measured
at Site 2.
LOADING TO LAKE ONTARIO
After leaving Waterport Pond, Oak Orchard Creek continues for another 10
km until it enters Lake Ontario at Point Breeze. We did not specifically
investigate the effects of this final stream segment on the phosphorus load
leaving Waterport Pond. There was little evidence of significant phosphorus
attenuation by the stream bed in the monitored sections. Therefore,
conservative transport from Waterport Dam to Lake Ontario was assumed.
For our study year, May 1984 through April 1985, the calculated TP load
at Waterport outlet was 37. tonnes with DMRP accounting for 20. tonnes
49
-------
(Figure 11). Twenty-eight percent of the TP load and 35 percent of the DMRP
load exited Waterport Pond during the month of March alone. In contrast,
about the same percentages were exported over the 7-month period from June
through December.
Though the TP load gained 10. tonnes between Sites 2 and 3, the DMRP
load increased by only 2. tonnes. From Tables 7a and 7b, which present mass
balances for TP and DMKP in Waterport Pond, it is apparent that the Erie
Barge Canal is a significant source of the particulate and unreactive
dissolved phosphorus forms, contributing a combined total of about 5600 kg
annually. This fact accounts for some of the large gain observed in total
phosphorus load between Sites ? and 3.
The single synoptic survey conducted under low-flow conditions before
Erie Barge Canal additions began suggested a decrease in instantaneous
phosphorus loading rate between Site 2 and 3. It is possible that flow
measurement inaccuracies or time-lags between sites could explain the
decrease. A net decrease in phosphorus in the spring of the year before flow
augmentation begins may actually be a real phenomenon resulting from algal
uptake and particle settling in Glenwood Lake and Waterport Pond. There is
evidence that removal of DMRP occurs in Waterport Pond to some extent. In
Table 7b, it can be seen that for the months of May, June, February, March
and April, the monthly loads of DMRP leaving Waterport Pond were less than
those measured upstream at Site 2. The difference between the annual DMRP
load calculated by mass balance to be entering Waterport Pond and the actual
amount measured at the outlet of 6700 kg (27,000 minus 20,000), if assumed to
make up the difference observed in calculated and actual annual total phos-
phorus loads, suggests that any loss in the impoundment is of predominantly
this immediately bioavailable fraction with little effect on the remaining
phosphorus forms.
Further evidence of DMRP removal is presented in Figure 17.
Concentrations of DMRP are plotted against time (June to December) for the
three Waterport sampling sites: Site 5 being the uplake location; Site 6
being mid-lake; and Site 3 being the outlet (only epilimnetic data were
plotted for Sites 5 and 6). Clearly, a concentration gradient existed with
increasing distance down-lake from June, when sampling began, until early
September. This type of relationship was not apparent when the unreactive
(TP minus DMRP) concentration was plotted in the same manner. Dilution with
incremental flow containing lower levels of phosphorus does not explain the
DMRP differences observed between these three sites. From the flow volumes
presented in Table 2, it can be seen that incremental flow downstream of Site
2 accounted for only 4 percent of the total discharge at Waterport outlet for
the period June to September. Even if this incremental flow had a zero
phosphorus concentration, it could not effect the extent of gradient shown
here.
The Site 3 spike in mid-November (Figure 17) possibly reflects mixing of
epilimnetic waters with hypolimnetic waters high in phosphorus from bottom
sediment release (see Figure 6) during fall overturn. This internal loading
may account for the actual DMRP monthly load leaving Waterport Pond in
November being greater than the load calculated to be entering the
50
-------
Ol
E
o
c
o
CL
QC
.120
.090
.060
.030
0
-Q Site 5 «pi
• • Site 6 epl
*-.-* Site 3
J J A S O N D
Month
FIGURE 17. Time course of DMRP concentration at Sites 5, 6 and 3.
impoundment (see Table 7b). From the mass balance calculation, the overall
annual net effect of Waterport Pond on DMRP during the study period appears
to be removal of about 25 percent of the load entering via Oak Orchard Creek.
The sediment load leaving Waterport Pond during the study period was
calculated to be 2500 tonnes. Particulate phosphorus accounted for 10 tonnes
(27 percent) of the total phosphorus load. The ratio of particulate
phosphorus load to sediment load, or average sediment-phosphorus
concentration, of about 4000 mg/kg still reflects significant enrichment,
though what portion might be potentially available for algal growth, is
unknown. A large amount of sediment is contributed by the Erie Barge Canal
which, during the 7 months of supplemental input, loaded about 1900 tonnes of
sediment to Oak Orchard Creek.
From Tables 7a and 7b, it can be seen that the bulk of the phosphorus
load to Waterport Pond is derived from Site 2 whose load is directly related
to losses upstream from cultivated mucklands. The conservative transport of
muckland phosphorus could account for as much as 39 percent of the TP load
and 72 percent of the DMRP load leaving Waterport Pond. Opportunities for
reductions from other sources are limited. When the Medina STP is in full
compliance with the 1.0 mg/L effluent phosphorus limit, the annual TP load
from this source is expected to decrease by about 2.4 tonnes, thereby
-------
TABLE 7a. MASS BALANCE OF TOTAL PHOSPHORUS LOAD (kg) ENTERING WATEKPORT POND ON A MONTHLY BASIS COMPARED TO ACTUAL
MEASURED LOAD AT SITE 3 FOR THE STUDY PERIOD.
WliNFPH
**CALCULATED
„ , C'riJ-TTVT' n-rrrm -> T /-*TVT-\
Site 2
May
June
July
August
September
October
November
December
January
February
March
April
TOTALS
(%)
3
2
3
2
8
3
27
,900
,200
630
360
350
490
490
790
,300
,200
,800
,500
,000
(63)
* Incremental
Flow
290
86
19
12
11
24
38
75
240
220
590
260
1,800
(4)
Erie Barge Canal
1
1
1
1
1
1
8
,300
,200
,400
,300
,100
820
,000
120
-
-
-
—
,200
(19)
Medina
STP
500
480
500
500
480
500
480
500
500
450
500
480
5,900
(14)
Total (%)
6,
4,
2,
2,
1,
1,
2,
1,
4,
2,
9,
4,
43,
000
000
500
200
900
800
000
500
000
900
900
200
000
(14)
(9)
(6)
(5)
(5)
(4)
(5)
(3)
(9)
(7)
(23)
(10)
(100)
ACTUAL
Total
5,300
3,000
1,600
1,500
1,400
1,400
2,200
880
4,100
2,400
10,000
2,900
37,000
(100)
(%)
(14)
(8)
(4)
(4)
(4)
(4)
(6)
(2)
(11)
(7)
(28)
(8)
* Loads based on flow estimates of [0.26 x Site 2 flow] and average TP concentration of 0.050 mg/L.
** Theoretical load assuming conservative transport.
-------
TABLE 7b. MASS BALANCE OF DISSOLVED MOLYBDATE REACTIVE PHOSPHORUS LOAD (kg) ENTERING WATERPORT
POND ON A MONTHLY BASIS COMPARED TO ACTUAL MEASURED LOAD AT SITE 3 FOR THE STUDY PERIOD.
un
**CALCULATED
Site 2
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
TOTALS
3,
1,
1,
1,
7,
1,
18,
000
800
420
250
240
270
330
500
900
100
100
200
000
(%) (67)
*Incremental
Flow
120
34
8
5
5
10
15
30
97
87
240
110
760
(3)
Erie Barge Canal
350
420
380
390
460
320
410
49
-
-
-
-
2,800
(11)
tMedina
STP
420
410
420
420
410
420
410
420
420
380
420
410
5,000
(19)
ACTUAL
CTHTT? "3 T r\An
Total (%)
3
2
1
1
1
1
1
1
2
1
7
1
27
,900
,600
,200
,100
,100
,000
,200
,000
,400
,500
,800
,700
,000
(15)
(10)
(5)
(4)
(4)
(4)
(4)
(4)
(9)
(6)
(29)
(6)
(100)
Total (%)
2,
1,
1,
2,
7,
20,
300
600
800
610
690
890
400
560
300
970
000
760
000
(12)
(8)
(4)
(3)
(4)
(4)
(7)
(3)
(11)
(5)
(35)
(4)
(100)
* Loads based on flow estimates of [0.26 x Site 2 flow] and average DMRP concentration of 0.020 mg/L.
t Assume 85 percent of STP discharge is DMRP.
** Theoretical load assuming conservative transport.
-------
reducing the total calculated input to Waterport Pond by only 6 percent.
Even though the Erie Barge Canal contributes 19 percent of the TP load, which
is relatively significant, it is a largely uncontrollable source by the very
nature of its use and origin. The key to substantially reducing the nonpoint
source load from Oak Orchard Creek to Lake Ontario appears to be large-scale
control of phosphorus leaching losses from cultivated mucklands during late
winter-early spring, high-flow periods.
54
-------
EEFERENCES
American Public Health Association, American Water Works Association, and
Water Pollution Control Federation. 1980. Standard Methods for the
Examination of Water and Wastewater. Atner. Public Health Assoc.,
Washington, D.C.
Boning, C.W. 1974. Generalization of stream travel rates and dispersion
characteristics from time-of-travel measurements. J. Res. U.S.G.S.
2 (4)-.495-499.
Brierly, J.A., D.K. Brandvold, and C.J. Popp. 1975a. Waterfowl refuge
effect on water quality: I. Bacterial populations. J. Water Poll.
Control Fed. 47(7):1892-1900.
1975b. Waterfowl refuge effect on water quality: II. Chemical
and physical parameters. J. Water Poll. Control Fed. 48(4):680-687.
Brown, M.P., M.R. Rafferty, and P. Longabucco. 1983. Phosphorus Transport
in the West Branch of the Delaware River Watershed. Volume 3 of Nonpoint
Source Control of Phosphorus - A Watershed Evaluation. Final report to
the United State Environmental Protection Agency, Region II, New York.
New York State Department of Environmental Conservation. 112 p.
Bouldin, D.R., A.H. Johnson, and D.A. Lauer. 1975. Transport in streams.
Pages 61-120 in Nitrogen and Phosphorus; Food Production, Waste, the
Environment, K.S. Porter (ed.). Ann Arbor Sci., Ann Arbor, MI.
Chandler, E. 1986. Personal Cortrnunication. Ircquois NWR.
Chang, S.C. and M.L. Jackson. 1957. Fractionation of soil phosphorus. Soil
Sci. 84: 133-144.
Chow, V.T. 1964. Runoff. Pages 14-1 to 14-54 in Handbook of Applied
Hydrology, V.T. Chew (ed.). McGraw-Hill Book Co., New York, NY.
Cogger, C. and J.M. Duxbury. 1984. Factors affecting phosphorus losses from
cultivated organic soils. J. Env. Qual. 13(1):111-114.
Dower, M.M. and K.R. Olson. 1980. Revised Pipet Method of Particle Size
Analysis. Soil Characterization Laboratory, Dept. of Agronomy, Cornell
University, Ithaca, NY.
55
-------
Duxbury, J.M. and J.H. Peverly. 1978. Nitrogen and phosphorus losses from
organic soils. J. Env. Qual. 7(4):566-570.
Erickson, A.E. and B.C. Ellis. 1971. The nutrient content of drainage water
from agricultural land. Mich. State Univ. Res. Bull. No. 31.
Fillos, J. and W.R. Swanson. 1975. The release rate of nutrients from river
and lake sediments. J. Water Poll. Control Fed. 47 (5)-.1032-1043.
Frevert, T. 1980. Dissolved oxygen dependent phosphorus release from
profundal sediments of Lake Constance. Hydrobiologia 74:17-28.
Funk, W.H., H.L. Gibbons, Jr., and G.C. Bailey. 1980. Lake assessment in
preparation for a multiphase restoration treatment. In "Restoration of
Lakes and Inland Waters", Proc. International Symposium.
EPA-440/5-81-010, USEPA, Office of Water Regulations and Standards,
Washington, D.C.
Hanway, J.J. and J.M. Laflen. 1974. Plant nutrient losses from tile outlet
terraces. J. Env. Qual. 3:351-356.
Have, M.R. 1973. Effects of migratory waterfowl on water quality at the
Montezuma National Wildlife Refuge, Seneca County, New York. J. Res.
Geol. Surv. 1(6):725-734.
Hortenstine, C.C. and R.B. Forbes. 1972. Concentrations of nitrogen,
phosphorus, potassium, and total soluble salts in soil solution samples
from fertilized and unfertilized Histosols. J. Env. Qual. 1:446-449.
Klausner, S. 1986. Personal communication. Cornell Univ.
Klopatek, J.M. 1975. The role of emergent macrophytes in mineral cycling in
a freshwater marsh. Pages 367-393 in Mineral Cycling in Southeastern
Ecosystems, F.G. Howell, J.B. Gentry, M.H. Smith (eds.). U.S. ERDA.
Logan, T.J. 1982. Mechanisms for release of sediment-bound phosphate to
water and the effects of agricultural land management on fluvial
transport of particulate and dissolved phosphate. Hydrobiologia
92:519-530.
1978. Chemical Extraction as an Index of Bioavailability of
Phosphate in Lake Erie Suspended Sediments. Lake Erie Wastewater
Management Study. U.S. Army Corps of Engineers, Buffalo, NY. 49 p.
Longabucco, P. 1984. Quality Assurance Plan for the Agricultural Nonpoint
Source Control of Phosphorus in the New York State Lake Ontario Basin
Project. New York State Department of Environmental Conservation,
Albany, NY. 98 p.
56
-------
Lucas, R.E. 1982. Organic Soils (Histosols): Formation, Distribution,
Physical and Chemical Properties and Management for Crop Productions.
Research Report 435, Farm Science. Michigan State University, East
Lansing, MI. 77 p.
Mawson, S.J., H.L. Gibbons, Jr., W.H. Funk, and K.E. Hartz. 1983.
Phosphorus flux rates in lake sediments. J. Water Poll. Control Fed.
55 (8):1105-1110.
McCallister, D.L. and T.J. Logan. 1978. Phosphate adsorption-desorption
characteristics of soils and bottom sediments in the Maumee River basin
of Ohio. J. Env. Dual. 7(1):87-92.
Meals, D.W. and E.A. Cassell. 1977. Phosphorus Interactions in the .Aquatic
Environment. Completion Report, Task 1.4, Eutrophication Task Force:
Lake Champlain Basin Level B Study. Univ. Vermont, Burlington, VT.
Miller, M.H. 1979. Contribution of nitrogen and phosphorus to subsurface
drainage water from intensively cropped mineral and organic soils in
Ontario. J. Env. Qual. 8:42-48.
New York State Department of Environmental Conservation. 1981. Final Report
- New York Angler Survey, 1976-1977. Bureau of Fisheries, New York State
Department of Environmental Conservation, Albany, NY.
. 1975. Water Quality Management Plan for Planning Area 03-01,
Lake Ontario West. New York State Department of Environmental
Conservation, Albany, NY.
. Discharge monitoring report files. Bureau of Wastewater
Facilities Operations, New York State Department of Environmental
Conservation, Albany, NY.
Nicholls, K.H. and H.R. MacCrimmon. 1974. Nutrients in subsurface and
runoff waters of the Holland marsh, Ontario. J. Env. Qual. 3(l):31-35.
Pavlou, S.P. 1980. Thermodynamic aspects of equilibrium sorption of
persistent organic molecules in the sediment-seawater interface: a
framework for predicting distributions in the aquatic, environment. Pages
323-333 in Contaminants and Sediments, Volume 2, R.A. Baker (ed.). Ann
Arbor Science, Ann Arbor, MI.
Peverly, J.H. 1985. Element accumulation and release by macrophytes in a
wetland stream. J. Env. Qual. 14(1):137-143.
1982. Stream transport of nutrients through a wetland. J. Env.
Qual. 11(1):38-43.
Porcella, D.B., et al. 1974. Comprehensive management of phosphorus water
pollution. EPA 600/5-74-010, USEPA, Washington, D.C.
57
-------
Soil Conservation Service. 1975. Plan and Environmental Impact Statement -
Oak Orchard Creek Watershed, Genesee and Orleans Counties, New York.
USDA SCS, Syracuse, New York.
. 1974. Erosion and Sediment Inventory. USDA SCS, Syracuse, NY.
106 p.
Starowitz, C. 1984. Personal communication. Genesee-Orleans Vegetable
Growers Cooperative Assoc., Inc.
Turkow, D.P., W. Dickinson, and L. Pratt. 1982. Oak Orchard Creek and West
Branch of Sandy Creek - Stream Survey. Orleans County Health Department,
Albion, NY.
United States Department of Interior Fish and Wildlife Service. 1964.
Iroquois National Wildlife Refuge, New York: Master Development Plan.
USFWS Region 5, Boston, MA.
United States Environmental Protection Agency. 1982a. Test Methods for
Organic Chemical Analysis of Municipal and Industrial Wastewater, July
1982. EPA-600/4-82-057, USEPA, July 1982. Washington, D.C.
. 1982b. Test Methods for Evaluating Solid Waste, July 1982.
SW-846 2nd Ed., USEPA. SW-846, 2nd ed. Washington, D.C.
1979. Methods for Chemical Analysis of Water and Wastes.
Washington, D.C. EPA 600/4-790-020. USEPA,
U.S. Gypsum Staff. 1985. Personal communication.
Whittemore, R. 1986. Personal communication. Iroquois NWR.
1985. Annual water management program report for Iroquois
National Wildlife Refuge. USFWS Region 5, Boston, MA.
Williams, J.D.H., J.R. Syers, D.E. Armstrong, and R.F. Harris. 1971a.
Characterization of inorganic phosphate in noncalcareous lake sediments.
Soil Sci. Soc. Amer. Proc. 35:356-561.
. 197Ib. Fractionation of inorganic phosphate in calcareous lake
sediments. Soil Sci. Soc. Amer. Proc. 35:250-255.
U.S. Government Printing Office : 1991 - 281-724/43568
58
-------
1. REPORT NO.
EPA-905/9-91-006A
4. TITLE AND SUBTITLE
Agricultural Nonpoint Sour
Lake Ontario Basin
Volume I- Deliver;
Mucklands in the C
TECHNICAL REPORT DATA
2.
se Control of Phosphorus in the New York State
f of Phosphorus to Lake Ontario from Cultivated
Dak Orchard Creek.
7. AUTHOR(S)
Patricia Longabucco
Michael R. Rafferty
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Bureau of Technical Services and Research
Division of Water
New York State Department of Environmental Consulting
50 Wolf Road
Albany, New York 12233-3502
12. SPONSORING AGENCY NAME AND ADDRESS
Great Lakes National Program Office
U.S. Environmental Protection Agency
230 South Dearborn Street
Chicago, Illinois 60604
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
1987
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
A42B2A
11. CONTRACT/GRANT NO.
R005725
13. TYPE OF REPORT AND PERIOD COVERED
Final-1984-1986
14. SPONSORING AGENCY CODE
GLNPO
15. SUPPLEMENTARY NOTES
Ralph Christensen, USEPA Project Officer
John Lowrey, Technical Assistant
16. ABSTRACT
Cultivated mucklands in western New York State were investigated as a nonpoint source of phosphorus to Lake Ontario. The 70,500-ha Oak Orchard Creek
watershed, which drains to Lake Ontario, was selected for the study area. It is located in Genesee and Orleans Counties, New York, and contains 3250 ha of heavily
fertilized muck cropland on which predominantly vegetable crops are grown. The creek was monitored at several sites from May 1984 through April 1985 to determine
the role of the mucklands in annual phosphorus loading to the lake.
At an upstream site which drained approximately 10,200 ha, including the majority of the muck cropland, the creek load was 18,000 kg of total phosphorus with 75
percent of it as dissolved reactive phosphorus. Two-thirds of the annual load was delivered in the 3-month, high-flow period of February through April. Runoff during
the late winter-early spring period appears to be the most important hydrologic factor in governing annual phosphorus loading from the mucklands, greater than either
total precipitation or total runoff for the year.
Control of phosphorus losses from the mucklands would appear to offer the most significant and cost-effective opportunities for loading reductions from this
watershed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTIONS b. IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field Group
Muckland
Organic Soil
Sediment
Phosphorus
Nutrient Transport
18. DISTRIBUTION STATEME
Document available to the ]
Technical Information Serv
Springfield, VA 22161
NT 19. SECURITY CLASS (This Report) 21. NO. OF PAGES
)ublic through the National None 59
cc NTTS
""'" iivJ 20. SECURITY CLASS (This page) 22. PRICE
None
EPA Form 2220-1 (Rev. 1-91) PREVIOUS EDITION IS OBSOLETE
------- |