MONITORING A WETLAND WASTEWATER TREATMENT SYSTEM
AT CANNON BEACH,OREGON
BY
KENNETH T. FRANKLIN and ROBERT E. FRENKEL
100
50
METERS
CONTROL
A Report Prepared for the U.S. Environmental Protection Agency, Region 10
Under Grant No. X-000328-01-0
Department of Geography
Oregon State University
Corvallls, Oregon 97331
April 1987
-------
sv
MONITORING A WETLAND WASTEWATER TREATMENT SYSTEM
AT CANNON BEACH, OREGON
BY
KENNETH T. FRANKLIN and ROBERT E. FRENKEL
Property 01 U.S. Environmental
«tH8fttA§eaeyLH«ry OMP-104
MAY 1 5 2000
> ft* toWW, talfe, WA 31101
A Report Prepared for the U.S. Environmental Protection Agency. Region 10
Under Grant No. X-000328-01-0
Department of Geography
Oregon State University
Corvalll8, Oregon 87331
April 1987
U.S. EPA LIBRARY REGION 10 MATERIALS
RXDD
GEAMB
-------
ACKNOWLEDGEMENTS
This research was supported in part by the Environmental
Protection Agency, Region 10 under Grant No. X-000328-01-0. This
report would never have been completed without the help and support
of the Oregon Division of State Lands, especially Ken Bierly and
Joyce Wescott.
Field work was made possible with the volunteer assistance of
Irv Jones, Neal Maine, Ed Johnson, Dan Elek, and the indomitable
super subaquatic crew of GGS 539. Special thanks to Doug O'Neil who
devised a unique method of collecting Daphnia. Thoughtful criticism
and advice were provided by Ralph Rogers, Don Thompson, Dave Bella,
and Keith Muckleston. Thanks to Liz Frenkel for providing an
efficient courier service. Last but not least, special thanks to
Jean Franklin for editorial and technical assistance as well as
moral support.
i
-------
PREFACE
In June of 1984, the City of Cannon Beach initiated tertiary
treatment of chlorinated wastewater with a natural forested
wetland. Monitoring of this wetland was a condition of the Corps of
Engineers 404 permit. In the spring and summer of 1986, we
remonitored the wetland using the sampling system established in
1984. We developed a new method of vegetation sampling for future
trends employing nested frequency plots, and analyzed water balance
and water quality. This report summarizes the 1986 monitoring and
research. We developed specific recommendations for future
monitoring and management of this wetland wastewater treatment
facility.
Work was carried out by groups of Oregon State University,
Department of Geography graduate students during the spring and
summer. With the assistance of the Environmental Protection Agency,
Region 10, through Grant X-000328-01-0, analysis and report
preparation was initiated in late September.
We hope this report will assist in the future planning of
natural wetland wastewater treatment systems in the Pacific
Northwest and in the continued monitoring and management of the
Cannon Beach facility.
ii
-------
EXECUTIVE SUMMARY
The City of Cannon Beach uses a seven hectare red alder/
slough sedge/twinberry palustrine wetland, divided into two cells in
series, to treat chlorinated effluent from a four-cell aerated/
facultative lagoon sewage treatment system. Operation began June,
1984. The system has proven to be an effective means of meeting
summer wastewater discharge limitations.
The Corps of Engineers 404 permit requires biological
monitoring of this system to evaluate future wetland treatment
system proposals in the region and for improved wastewater wetland
management. Vegetation was sampled in 1984 by permanent plots prior
to facility initiation and repeated in 1986. Generally, herb and
shrub cover changed little since 1984. In channelized and deeply
flooded areas herb cover decreased. Slough sedge cover increased
slightly in the shallowly flooded eastern section of cell 2.
By 1986, a complex pattern of flooding stress is exhibited by
defoliated, sparsely leaved, and some dead red alder trees in deeper
water areas. The hummocky topography complicates prediction of
flooding stress effects.
Field research in spring and summer, 1986 used a nested
frequency method to sample herbaceous vegetation. A total of 155
2 2
nested frequency sampling plots (0.25 m nested within 1.0m)
provides baseline data for vegetation trend analyses. Nested
iii
-------
frequency vegetation cover agrees with permanent plot cover;
however, nested frequency yields substantially more plots per
man-hour than the permanent plot method.
Mean wetland influent and effluent flow rates were 0.40 and
0.06 MGD respectively in 1986. An independent water budget estimate
suggests ground water infiltration is at least 65-85% of the water
loss.
The Cannon Beach system has met ODEQ water quality discharge
standards for three years, reducing B0Dc and TSS concentrations
J
by 40% and 85% respectively. Phytoplankton constitute most of the
influent suspended solids. Within 30-40 meters into treatment cell
1, the phytoplankton concentration is greatly reduced; suspended
solids thereafter are mostly composed of plant detritus and humic
materials. Settling and resuspension of suspended solids varies
complexly throughout the facility.
Tree loss is not likely to adversly affect the water quality
treatment; however, long-term nutrient retention may diminish with
tree death. With an aging system, passerine bird habitat will
decrease in the western section, but waterfowl habitat will
increase. Slough sedge cover is likely to remain relatively stable
or increase. Monitoring needs to be improved and coordinated with
management.
iv
-------
Monitoring Recommendations
Two levels of monitoring are recommended;
Level 1
Plant operators should:
a. Record changes made in discharge header wastewater
distribution and weir levels and weather conditions;
b. Measure hydroperiod on a bi-weekly basis while
taking required water samples.
c. Determine area of inundation as a function of
influent flow rate.
Repeat the nested frequency sampling method every two
years.
Repeat the permanent plot sampling method every two
years. The influent should be shut off for a one week
period during the vegetation sampling survey.
Researchers should integrate hydroperiod information
with vegetation survey data and draw inferences with
caution.
Repeat basal diameter tree survey in 1989 and thereafter
at 3 year intervals.
Level 2
1. In addition to level 1 monitoring, measure hydroperiod
on a yearly basis, at least bi-weekly.
2. Conduct a detailed topographic survey.
v
-------
3. Develop a mathematical model to determine the area of
inundation as a function of water level, and water level
as a function of influent flow rates, rainfall and
weather conditions. Experiments should be conducted to
determine variability in water level with changes in
water dispersal.
4. Place a network of shallow wells throughout the wetland
to investigate ground water infiltration.
5. Monitor influent and effluent nitrogen and phosphorus
bi-weekly. Report 'removal efficiencies' with caution
because dissolved materials may leave by ground water
infiltration.
Management Recommendations
It is unlikely that BOD and TSS loading of Ecola Creek
through wetland discharge will exceed ODEQ limitations in the near
future. If limitations are exceeded, then operators will probably
need to lower water levels and decrease the water turnover rate. A
last, and highly unlikely, alternative should this occur would be to
make the flow in the wetland more tortuous. A willingness to
experiment based on clearly defined objectives, monitor changes, and
incorporate observations into management decisions is a sound
approach to managing the Cannon Beach wastewater wetland treatment
system.
vi
-------
CONTENTS
ACKNOWLEDGEMENTS i
PREFACE ii
EXECUTIVE SUMMARY iii
INTRODUCTION 1
WETLAND WASTEWATER TREATMENT SYSTEM 5
Setting 5
Design 8
RESEARCH OBJECTIVES 12
VEGETATION ECOLOGY 14
Introduction 14
1984/1986 VEGETATION SURVEY COMPARISONS 20
Methods and Materials 20
Results and Discussion 22
1986 MONITORING METHODS 29
Introduction 29
Methods and Materials 33
Results and Oiscussion 36
Recommendations for Resampling 44
WATER BUDGET ESTIMATE 46
Introduction 46
Methods 47
Water Budget Model 47
Discussion 55
WATER QUALITY 59
SUMMARY AND CONCLUSIONS 68
Monitoring Recommendations 71
Management Recommendations 74
LITERATURE CITED 75
APPENDIX A. Tree transect data 79
APPENDIX B. Nested frequency sampling notes 83
APPENDIX C. Species list 84
vii
-------
INTRODUCTION
Wetlands have been, and continue to be, subject to all sorts
of abuses by humans: they are filled, drained, and used to collect
inadvertent discharges of polluted water. Through a combination of
clearing, draining, and flood control projects, 30-50% of the
original wetland area of the continental U.S. has been lost (Office
of Technology Assessment 1984). Since the 1970's there has been
greater public recognition that wetlands deserve protection, as they
often provide valuable ecological services, or functions, to the
human population and to the biosphere as a whole (Mitsch and
Gosselink 1986). These functions include cleansing polluted waters
(Godfrey et al. 1985, Nichols 1983), preventing floods (Novitzki
1979, Verry and Boelter 1979), protecting shorelines (Knutson et al.
1981), and providing important habitats for flora and fauna
(Williams and Dodd 1979, Odum et al. 1984).
The recognition that wetlands can improve water quality under
certain situations has called attention to the planned use of both
natural and artificial wetland treatment systems as cost-effective
alternatives for upgrading sewage treatment facilities and treating
urban and agricultural runoff. Several reviews and conferences that
have dealt with the potential and limitations of wetland treatment
systems include Hammer and Kadlec (1983), the U.S. Environmental
Protection Agency (1983), the U.S. Environmental Protection Agency
and U.S. Fish & Wildlife Service (1984), and Godfrey et al. (1985).
While there are hundreds of wetland treatment facilities of various
-------
sizes and designs located in the southeastern and midwestern part of
the U.S., comparatively few exist in the Pacific Northwest.
The City of Cannon Beach, on the northern Oregon coast, is
among the first municipalities in the Pacific Northwest to use a
natural wetland in combination with conventional technology to treat
domestic wastewater. A seven hectare red alder/slough sedge/
twinberry palustrine wetland, divided into two cells in series,
treats chlorinated effluent from a four-cell aerated/facultative
lagoon system (Figure 1). Operation began June, 1984.
101
100
100
METERS
Cefl 2
Cefl 1
AB - Aeration Basin
1,2,3 - Facultative lagoon*
S - Sludge dloposal pits
C - Chlorine contact chamber
WOP - Winter outfall pipe
Cell l.Cell 2 - Wetland treatment cello
Figure 1. Vicinity and general schematic of the Cannon Beach
wetland wastewater treatment facility.
2
-------
Wetland treatment occurs Hay 1 through November 1 to meet
the more stringent summer wastewater discharge standards established
by the Oregon Department of Environmental Quality (ODEQ) (Table 1).
These limitations are 10 mg/1 B0Dc (five-day biochemical oxygen
demand) and 10 mg/1 TSS (total suspended solids), hereinafter
referred to as 10/10 limitations.
Table 1. Ecola Creek discharge requirements (KCM 1981).
Monthly Average Effluent Concentrations (mg/1)
Period BOD,. TSS
Pre-pro.ject (before 6/1/84)
9/20 - 5/19 30 50
5/20 - 9/19a
Present (after 6/1/84)
11/1 - 5/31b 30 50
5/1 - 10/31C 10 10
3 Wastewater held in lagoons in summer; discharge allowed only
with ODEQ written permission.
b Direct discharge of lagoon sewage effluent into Ecola Creek.
Q
Discharge of lagoon sewage effluent into Ecola Creek via wetland
3
-------
During the 1970's, the existing three-cell, five hectare
lagoon treatment system was approaching its design capacity.
Required to hold sewage in the lagoons May-September by ODEQ, the
city was being forced into upgrading its sewage facilities because
the lagoon's storage capacity was projected to be inadequate for
summer holding (KCM 1981, Thompson and Minor 1986). The problem was
caused by a large influx of tourists in summer, when the permanent
population1 of 1200 swells to as much as ten times that number at the
height of the visitor season.
OOEQ's concern was for the potential public health problems
associated with low summertime creek flows and high effluent
discharges. The concern was that the creek would receive effluent
and polluted waters would pose a hazard to the large numbers of
tourists on the beach. Over a period of nearly eight years
(1975-1982) several alternative proposals for upgrading the
treatment system were considered and negotiated between the City's
Sewer Advisory Board, state and federal resource and regulatory
agencies.
Treatment alternatives ranged from conventional technologies
such as chemical treatment, a package activated sludge plant and
slow sand filtration to an aquaculture system, and an artificial
marsh treatment system (Thompson and Minor 1986). By 1980, the city
favored treatment by the artificial marsh, but the concept met with
disapproval by some state and federal agencies, which cited issues
4
-------
of land use regulation and excessive wildlife habitat alteration
(KCM 1981, Thompson and Minor 1986). In 1982, after intense
negotiation, the present, and least expensive scheme — a natural
wetlands wastewater treatment system combined with improvements to
the existing lagoons — was adopted by the City Council and approved
by all appropriate agencies.
WETLAND WASTEWATER TREATMENT SYSTEM
Setting
The wetland, located in Clatsop County adjacent to the Oregon
Coast Highway (U.S 101), is composed of several different
vegetation cover types resulting from historical patterns of land
use (Figure 2). The eastern portion of the treatment cells are
dominated by a Sitka spruce (Picea sitchensis) and red alder (Alnus
rubra) overstory, a remnant of a once extensive Sitka spruce
forest. The understory includes red elderberry (Sambucus racemosa),
sword fern (Polystichum munitum) and other typical coastal species.
The largest trees are located in this area (some Sitka spruce are
over five feet in diameter).
Tree cover is less dense towards the western portion of the
cells, a consequence of logging and highway construction. In low
areas, treeless patches and stagnant pools are dominated by slough
5
-------
so
METERS
Remnant of old-qrowth Sitka
Ef-i'J spruce forest (hiqhest elevation)
ffn Low tree density, primarily
red alder (low elevetion area
Moderate tree density, red alder
overstory (intermediate elevation
-^1 Moderate tree density, red alder
with Sitka spruce ir'derstory
(intermediate elevation)
Veqetation buffer, mostly
dense, young Sitka sprucp
Figure 2. Vegetation cover types in the wastewater wetland in 1986.
6
-------
sedge (Carex obnupta). skunk cabbage (Lvslchlturo americanum). and
twinberry (Lonicera involucrata). The dominant understory species
throughout the wetland is slough sedge.
Present flooding of the treatment cells with secondary
effluent has so far had a minor influence on wetland vegetation
composition and structure compared to historical direct and indirect
anthropogenic disturbances. These include (1) historic filling of
land for the town of Cannon Beach which restricted drainage of the
wetland area; (2) clearcut logging in the 20th century, most
recently in the mid-1950s; and (3) construction of Highway 101 which
severed the previous wetland configuration and altered the hydrology.
The nearest climatological station is in the city of Seaside,
eight miles to the north. Average annual precipitation (essentially
all rainfall) is approximately 70-80 inches, about 75% of which
falls during November-March. Summer months can have periods of
relatively dry weather, but fog is very common and probably reduces
evapotranspiration rates compared to more inland locations.
Two soil types are mapped by the Soil Conservation Service
(1982) in the wetland: Nehalem silt loam and the Coquille-Clatsop
complex. Nehalem silt loam is found in the areas of higher
elevation near the creek, a natural levee. This is a well-drained
to moderately well-drained soil formed in the mixed alluvium of the
Ecola Creek floodplain. No surface sewage lagoon effluent is in
7
-------
taetefcS
w9
///Veoe^8^
3.
Fi9ure n,oes
,scM"»e Pl?
of tw ^T11
o<€ota^on „ to *caAe*
at\c r"ePreS t
-------
contact with these soils. The Coqui1le-Clatsop complex is the
dominant soil type in the treatment cells; these are poorly-drained
soils composed of a very dark grayish brown silt loam surface layer
(33 cm thick), underlain by a dark grayish brown silty clay loam and
silty clay to 150 cm. Four bore holes were drilled for a foundation
study by Kelly/Strazer Associates (1983) indicating the wetland
soils "consist predominantly of soft organic silts with peat layers
which occasionally grade to medium stiff below about 4 feet".
Design
The primary objective of the treatment system is to meet the
10/10 limitations as shown in Table 1. A secondary objective is to
minimize disturbance of wildlife habitat, a habitat of importance
especially for winter elk herds. During the winter a herd of
Roosevelt elk, averaging 18-20 animals, wanders over the lower Ecola
Creek watershed, including the project area (KCM 1981).
The treatment cells are simple in design; the clay-lined,
partially rip-rapped, dikes forming the cell boundaries take
advantage of the highway barrier to the west and the natural levee
of Ecola Creek to the east (Figure 3). During construction,
structural disturbance to the swamp was limited to occasional tree
removal and placement of fill material for the dikes. A six-port
discharge header on the south dike, producing a small stream of
discharged wastewater localized at each pipe orifice, allows some
spatial control over effluent dispersal. Hydraulic control is also
9
-------
too
50
METERS
Primary channels
Extent of area of inundation
Expansion and contraction of the
area of inundation^ f(influent
flow rates, rainfall,
and evapotranspiration)
Figure 4. Generalized diagram of channelization and water dispersal
for the treatment cells in 1986. Weirs 3 and 4 closed to overflow.
Data is based on staff gauge observations on September 14-15 with
the mean influent flow rate equal to 0.310 MGD.
10
-------
obtained by four weirs 1n the middle dike. For the 1986 operational
season, weirs 3 and 4 were closed to discharge; weir 5 was set at a
level that allowed overflow, and weir 6 had very infrequent overflow
due to its higher elevational position. Weir levels are not
currently calibrated to wetland water levels and the operational
hydroperiod.
Figure 4 shows the generalized water dispersal pattern for
the 1986 operational season. The channelization pattern and total
area of inundation is of course a function of influent flow rates.
Several primary channels carry higher velocity flows in the eastern
section of the cells and slower velocity within deeper water areas
in the lower elevational western section of the cells.
A ditch located outside of cell 1, adjacent to the south
dike, is connected to Ecola Creek, allowing flushing by creek waters
and sheet flow from the control section during heavy rains (Figure
3). This design helps maintain pre-project hydrologic conditions
during the winter season. Spillways are also located in the dikes
to help discharge high water flows. Two weirs at the north end of
cell 2 allow discharge of effluent to Ecola Creek along a natural
channel. Ouring the winter, tidal influence extends to this channel
causing some backflow into cell 2 (Elek, personal communication
1987). A current meter that measures average daily flow rates is
located at this discharge structure.
An approximate 10 m wide vegetation buffer, primarily a dense
stand of young Sitka spruce, is located along the western boundary
II
-------
of the treatment cells and provides an effective noise and visual
buffer from heavy highway traffic. Fencing has been limited to
chainlink gates at the dike entrances to the highway, allowing
passage of curious pedestrians, fisherman, deer, and other animals.
RESEARCH OBJECTIVES
An important Section 404 Corps of Engineers permit condition
for the Cannon Beach project is the following: "A biological
monitoring plan will be devised to document the effects of the
system's operation on floral assemblage. From the data collected, a
management plan will follow which allows for optimal water level
maintenance to promote desirable/beneficial vegetation assemblages
within the treatment facility" (Corps of Engineers 1983). The
precursor to this permit condition involves the need for data to
help make informed decisions on future proposals to use wetlands for
tertiary treatment. In the absence of data and background, approval
of the Cannon Beach project was based on a need for a data base and
experience for wetland wastewater treatment in the Pacific Northwest.
Treatment plant operators currently monitor BOD,., TSS,
chlorine residualj and fecal coliform on a bi-weekly basis for
facultative lagoon and wetland influent and effluent; hydrological
monitoring is limited to automatic recordings of lagoon and wetland
influent and effluent flow rates. Despite capital costs of over
$1.5 million for the project, no funds were allocated for any
12
-------
biological monitoring. Although a requirement of the Corps of
Engineers' permit conditions, the only biological monitoring
conducted prior to 1986 was a volunteer effort by Corps of Engineers
(COE), Oregon Department of Fish & Wildlife (ODFW), and U.S. Fish &
Wildlife (USFWS) staff to establish a baseline survey for vegetation
trend assessment just prior to system operation (Rogers 1984).
A research program reported upon here, funded by the
Environmental Protection Agency, Region 10, conducted April through
September, 1986 sought to repeat the 1984 vegetation survey and
develop additional ecological and hydrological monitoring methods
for improved wastewater wetland management. Because of the
uncertainty associated with funding future monitoring efforts,
inexpensive yet effective monitoring procedures are desired.
Three objectives of the research study addressed the
following questions: (1) how have water and to a lesser extent
nutrient additions since 1984 affected vegetation composition and
structure?; (2) is the hydrological system operating as predicted
and how might it vary as the wastewater wetland ages?, and; (3) how
is the facility functioning with respect to improved wastewater
treatment? These questions will each be dealt with separately in
the following sections of the paper.
13
-------
Western section of cell 1
VI
s.
Ol
*->
Ol
E
OJ
QJ
>
a;
-------
VEGETATION ECOLOGY
Introduction
The primary disturbance associated with effluent discharge
into the wetland is plant flooding stress. Frequent winter flooding
and occasional flooding during the growing season is normal for the
site. An estimated hydroperiod for the site (Figure 5), however,
shows flooding now occurs during the growing season, the period
during which plants are most susceptible to metabolic stress
associated with anaerobiosis in flooded soils (Bedinger 1979, Gill
1970). The operational hydroperiod is determined by influent flow
rates and weather conditions (Figures 6 and 7).
The degree of flooding stress imposed upon the plant
community is a complex phenomena dependent on multiple factors. For
example, Van der Valk (1981) describes a model based on the
interaction of plant life history traits — life span, propagule
longevity, and propagule establishment requirements — and the
timing of water drawdowns and flooding events; the operational
hydroperiod places selection pressure on those species capable of
reproducing and growing in the new conditions. At the Cannon Beach
site, however, little autecological data exists for even the
dominant species, prohibiting the construction of such a model
(except perhaps for red alder). As shown by the profiles of
Transects C and 0 (Figure 8) and the water dispersal pattern (Figure
4), predicting flooding stress and subsequent successional change is
15
-------
Effluent discharge
Ito Ecola Creak
via lagoon outfall
(11/1)
Sewage lagoon influent
Sewage lagoon effluent
Wetlan
-------
A larqe rain event caused loa jam
to smash current meter. Overflow
also occurred at spillway 5 hut is
not recorded.
Wetland Influent
oo
.0 -
Wetland Effluent
cc
Oc too#'
-H
July
1986
Figure 7. Wetland influent and effluent flow rates and rainfall for
the 1986 operational season. Rainfall is recorded at the Cannon
Beach City Hall.
17
-------
West
Transect D
East
UiS. 101
Transect C (partial)
U.S. "101
? . 2i° i 4i° (meters)
Horizontal Scale
Figure 8. Topographic profiles for transects D and C (partial).
61 Water
¦W«ttendslnftuen»=0.331 MOO
18
-------
a complex spatial and temporal problem given different operational
hydroperiods throughout the treatment cells.
During the growing season, vegetation plays a role in
renovation of the sewage lagoon effluent by nutrient uptake
(temporary and long-term), serving as attachment sites for microbes,
acting as physical filters for suspended solids, and reducing solar
radiation by the canopy cover, which in turn reduces algal
productivity. Determining a mass balance for BOD, TSS, and various
nutrients for the Cannon Beach treatment cells is a very complex and
difficult undertaking far beyond the scope of the present research
study. Despite the important role of the soil and microbial
community in pollutant removal and transformation, ecological
monitoring described in this report is limited to the assessment of
vegetation change induced by treatment operation. How biotic
fluctuations caused by flooding will influence pollutant removal and
transformation can only be assessed by future studies.
In the following section two methods of vegetation sampling
and analysis for trend assessment are described: (1) Rogers (1984)
sampling procedure conducted in 1984 and repeated in 1986
(hereinafter called the permanent plot method); and (2) a procedure
using nested frequency quadrats conducted in 1986.
19
-------
50
100
METERS
Location of I m * l m herbaceous plot
variable = X cover.
Tree inventory transects; spp
and basal diameters recorded for
trees greater than 10 cm "dbh"
Figure 9. Permanent plot sampling design (Rogers 1984).
20
-------
1984/1986 VEGETATION SURVEY COMPARISONS
Methods and Materials
A volunteer group of COE, OOFW, and USFWS staff established a
baseline vegetation survey May 31 and June 1, 1984 by using a method
of estimating herbaceous and shrub cover along a stratified series
of transects (Rogers 1984) (Figure 9). At each transect point shown
in Figure 9, a one square meter herbaceous plot was randomly located
within 10 meters of the transect point and marked by two 1/2 inch X
2 m PVC pipes stuck in the soil. A total of 22 herbaceous plots
were sampled along six transects. A 1 m x 10 m shrub plot was also
centered at right angles to the transect at each transect point,
without any markings (22 total). For each herbaceous and shrub
plot, species were identified and their percent cover (rooted and
not rooted within the plot) estimated to the nearest 5%. Percent
cover was also estimated for bare ground and standing water.
In addition to the herb and shrub plots, tree sampling belt
transects were established along transects A, C, and F. In a band
five meters north and south of the transect line all trees greater
than 10 cm in diameter were recorded by species and diameter. Trees
were numbered serially eastward along the transects with 1 in. X 3
in. aluminum foil tags secured with galvanized nails.
21
-------
15
10-
5-
o-
10 -
5 -
o«=
lo —
t '5 '
° 10-
5 ~
o —
Atfl
Rusp
Lyam
tor
bo-
So
w
_0_H2Z3
3 o-
2.0-
/o
Caob
~ 1984
E11986
/or
c 5
0
O
Q.
Oesa
-rem
1
* * *
15
10
5
0
Adpe
fS
IC
s
0
Pomu
C «< I 1
C»H 2
C«ll 1
c«ll 2
Fiqure 10. Herbaceous sampling results for dominant species
usinq the permanent plot method, 1984 vs. 1986. Species acronyms
are given in Appendix C.
22
-------
Results and Discussion
Herbs and Shrubs
Given the nature of the method, statistical significance of
changes from 1984 to 1986 could not be calculated. Therefore,
assuming changes of at least 10% are "significant", the key results
depicted in Figures 10 and 11 are as follows:
(1) All three ferns (Athyrium fi1ix-femina. Adiantum
pedatum. and Polystichum munitum) show no "significant"
changes in 1986 relative to 1984 for cell 1 (ca. 4%). A
reduction in the abundance of these species, however, is
expected in the areas of channelization and deeper water
(Figure 4). In cell 2, however, P. munitum increased by ca.
10% with A. filix-femina and A. pedatum showing no
"significant" changes.
(2) A pre-project prediction concerning Carex obnupta was
that it would likely survive summertime flooding (Demgen,
1983). This appears to be the case for three years
operation. Figure 10 indicates essentially no reduction in
C. obnupta coverage in cell 1 for 1986 relative to 1984.
Cell 2 shows a 20% increase in C. obnupta cover.
(3) From Figure 11, none of the shrubs show a "significant"
change, except in cell 2 where Lonicera involucrata has 28%
23
-------
/5r
IO
5
Gash
' < • «
# T //
o K-"-l
O
O
iS
10
5
o
3 1.8
Loin I T
•/ ' /
~ 1984
H 1986
c
o
o
k_
©
Q.
/6
to
5
Rusp
« *v *
/ - .
* * ' r-4
* • •
«•
* «
. ^ ,...
1 1' ¦¦¦»!
control
call 1
cell 2
Figure 11. Shrub sampling results for dominant species
using the permanent plot method, 1984 vs. 1986. Species
acronyms are given in Appendix C.
24
-------
less cover in 1986 than 1984. Qualitative observations in
1986 confirm a higher degree of flooding stress (defoliated
branches) in the western section of cell 2 than elsewhere in
the wetland.
One must be very cautious when interpreting the results
depicted in Figures 10 and 11 for herbaceous and shrub changes since
1984, for the following reasons:
(1) Different survey crews were involved (1984 vs. 1986) and
there was no calibration of cover estimates within and
between survey crews. Subjectivity is greatly increased when
different survey crews are involved, especially combined with
the cold and very wet weather conditions of April 26, 1986.
(2) Sampling for herbaceous and shrub cover occurred May 30,
1984 vs. April 26, 1986; phenological differences are
probable but not assessed.
(3) Precise relocation of
consuming despite staking,
leading to imprecise cover
marked by light blue paint on
to relocate.
plots was difficult and time
Shrub plots were not staked,
estimates. Transect points,
trees, were sometimes difficult
25
-------
(4) Estimating percent cover of shrubs in plots greater than
one square meter can be highly inaccurate. This was
especially true for multiple stemmed shrub and vine species,
such as R. spectabilis and L. involucrata;
(5) Except for estimating cover on an overall treatment cell
basis the number of sample plots and their spatial location
is inadequate to sample the microhabitat complexity and total
acreage involved. For increased accuracy and resolution,
more sample plots are required (e.g., comparing changes west
to east, lower elevation areas to higher elevations, etc.).
(6) Whether a plant was rooted or non-rooted within a
sampling plot was not specified at the outset. Lack of this
information resulted in considerable uncertainty in cover
estimates, especially for A. rubra and shrub species.
Trees
A new baseline data set was collected June-July 1986 and is
listed and described in Appendix A. Unfortunately, during the
original tree transect survey of 1984, care was not taken to
standardize the vertical position of diameter measurements;
therefore, no meaningful comparison of basal growth can be made
between the 1986 and 1984 measurements. According to Rogers
(personal communication) diameters were measured at "eye height";
26
-------
however, this height varies with the height of the investigator and,
particularly, the side of the tree at which the measurement is
taken. This height can vary from a few inches to four feet or more.
From the 1986 data set importance values are derived for each
tree species and are shown in Table 2A. The importance value is the
sum of relative frequency and relative dominance for each species
and allows a comparison of overstory sructure and composition for
the control and treatment cell sections of the wetland. Transect A
(control) is more similar to transect F for Alnus rubra and Picea
sitchensis than to transect C. The difference is accounted for by a
larger importance value for P. sitchensis along transect C (because
of a larger mean basal area) and the presence of Pyrus fusca along
transects C and F and not along A. The essential point is that
there are differences in canopy structure and composition for the
control and treatment cell sections that impart different, and
difficult to assess, influences on successional change in each
respective section of the wetland — exclusive of wastewater
flooding disturbance. Table 2C also shows results of
double-sampling the tree transects in 1986 in order to provide
descriptive statistical information when the survey is repeated.
An additional problem with this method is the assumption that
basal growth accurately assesses flooding stress and can predict
survival of trees; basal diameter growth can stop yet the trees may
continue to survive. Studies by Mitsch and Rust (1984) indicate a
27
-------
Table 2. Tree transect survey results for 1986. See Appendix A for
data list and Appendix C for species acronyms.
A. Relative Frequency (RF). Relative Dominance (RO), and Importance Value (IV) for tree
transects.
Transect A Transect C Transect F
RF
RD
IV
RF
R0
IV
RF
RD
IV
Alru
88.4
85.9
114.3
76.0
69.1
145.1
84.1
91.6
175.7
P1s1
11 .6
14.1
25.7
14.0
28.6
42.6
8.7
4.6
13.3
Pyfu
0.0
0.0
0.0
10.0
2.3
12.3
7.2
3.8
11.0
RF = (frequency of given spp)(sum of frequencies of all spp)-1 x 100
RO = (sum of basal area of given spp)(sum of basal area of all spp)-1 x 100
IV = RF + RD
B. Basal Areas (BA) for the tree transects. Sampling areas: Transect A = 0.218 ha.
Transect C = 0.203 ha, and Transect F = 0.170 ha.
Basal Area (m2/ha)
Transect A N Transect C N Transect F N_
Alru 21.9 61 11.0 38 16.5 50
P1s1 3.6 8 4.6 7 0.8 6
Pyfu 0.0 0 0.4 5 0.7 5
C. Resampling results 1n 1986 to determine a statistical confidence.
Transect A Transect C Transect F
N 26 10 14
d(cm) 0.019 0.040 0.214
<5 ±0.017 ± 0.035 ±0.190
d = mean difference between sampling values.
er« standard deviation based on rf- ; note when resampling: 95% C.l. of a single
measurement, x, 1s x ~ 1.74d.
28
-------
complex nonlinear relationship between growth and flooding stress
and such a nonlinear relationship should be accounted for when this
baseline data set is resampled and analyzed.
Flooding stress, consisting of defoliated trees, sparsely
leaved tress, and some dead trees, is observed in 1986 for red alder
and twinberry in deeper water towards the western section of the
treatment cells and adjacent to the dikes where water ponds (Figure
12). Determining the extent and stablity of the observed flooding
stress is a complex spatial and temporal problem highly dependent on
i
the operational hydroperiod which has not yet been recorded.
Observations along the dikes appear to indicate ca. 25% of the
treatment cell area is under significant stress after three years
operation (Figure 12). But as Figure 13 also indicates, high
hummocks within the wetland contain trees of relatively good vigor,
complicating any prediction.
1986 MONITORING METHODS
Introduction
In response to the problems discovered with the permanent
plot method for sampling herbaceous vegetation, a method was
developed using nested frequency quadrats following a similar
approach by Hironaka (1985) and Smith et al. (1986) in rangeland
vegetation. The use of species frequency as a variable in
vegetation sampling is a common and well-established method for
29
-------
Boundary of acute red alder flooding
stress as observed from the dikes.
Fiqure .1?. Generalized boundary of acute red alder flooding
stress observed from the wetland boundary dikes, 1986. Stressed red
alder consisted of defoliated trees, sparsely leaved trees, and some
dead trees.
30
-------
CMt
Transect D
Transect C (partial)
Water
H VH.VC— — AJru—
HOfi7ont«l
QuaWattve Vtoor Classes
For Alnua cubu (Atru)
VA= healthy vigor (fully leafed out)
VB=intermediate vigor
VC=poor vigor (leaves few. or discolored, or smaO
RP=Reference point for profile elevations
Oate: 9/13, 14/86 Wetland Influent=0.331 MOD
Figure 13. Qualitative vigor zones for red alder along
transects D and C (partial). Note that hummocks within
the acute zone of flooding stress (Figure 12 — determined
by observations made solely from the dikes) contain
trees of healthy vigor.
31
-------
sampling herbaceous vegetation 1n rangeland and other vegetation
types.
Frequency is the number of times a species is recorded 1n a
set of plots or points and is usually expressed as a percent.
Interpretation of frequency does present problems, however.
Frequency depends on plot size and shape as well as the spatial
distribution of species in the sample area. When plants are not
dispersed in a regular or random pattern, frequency has no fixed
relationship to density, abundance or cover (Mueller-Dombois and
Ellenberg 1974). On the other hand, frequency data is easily
collected and is objective (either a plant is or is not in a plot).
Nested frequency methods establish a set of different plot sizes
such that in any resampling the same plot size is used in
reassessment of the trend for a given species.
Nested frequency sampling has not been utilized extensively
in wetland environments. The method is based on the presence or
absence of a species in a nested series of quadrats. For vegetation
trend analyses, nested frequency is considered to have greater
objectivity, repeatablity, and rapidity of use than repeat cover or
yield methods (Hironaka 1985, Smith et al. 1986). The results are
also more stable relative to seasonality than cover methods. One of
the major attractions of nested frequency for trend assessment is
that it does not require relocation of transects or plots. It
essentially assesses the entire stand of vegetation and therefore
can establish trend for the stand. A fixed plot method, on the
other hand, can only assess trend for the plot.
32
-------
Because species frequency varies as a function of quadrat
size, the use of nested quadrats increases the probability of
choosing an appropriate quadrat size (Hironaka 1985). Frequency
values alone, however, do not allow for spatial pattern description,
nor for accurate assessment of cover or biomass, and therefore cover
estimates often accompany frequency measurements.
Methods and Materials
For convenience and to evenly distribute a sampling network
throughout the wetland, nested frequency plots were taken along
seven belt transects of indeterminate width. Ideally, nested plots
would be randomly located throughout the wetland, but statistically
random location would have been impossible. Theoretically, in
vegetation reassessment, relocation of transects and plots is not
required. However, because of the heterogeneity of the vegetation
in cells 1 and 2, approximate transect relocation would be highly
desirable in reassessment.
Transects A through F generally corresponded with Rogers
(1984) transect locations (Figure 14). Transect G was added to more
completely sample cell 1. Two transects were established in the
control area and cell 2, and three in cell 1. Nested frequency
sampling was accomplished along all transects in a "semi-random"
manner.
After local trial two plot sizes were chosen: a 50 cm x 50
cm nested within a 100 cm x,100 cm plot. Four types of data were
33
-------
100
50
METERS
Herbaceous sampling plots located
semi-randomly within these bands
!«/•; t.*»I • v: .i, .V
Figure 14. Nested frequency/microenvironmental sampling method
(1986). Sampling transects A-C and D-F corresspond to the permanent
plot method. Transect G was added in 1986.
34
-------
collected at each plot: frequency, cover, water depth, and a
qualitative aquatic index used to gain some qualitative measure of
wetland character and degree of inundation. The plots were made in
the field by using four 1.5 m x 3/4 inch PVC pipes as a frame (the
pipes doubling as walking sticks — an important consideration in
extremely difficult terrain and given the necessity of minimizing
hand-held objects). A starting point was located at either end of
the transect and a magnetic bearing of 73 or 253 degrees gave the
general heading.
At approximately 10 m intervals along the transect one team
member would stand with either a telescopic stadia rod or two
connected PVC pipes (total length of 3 meters) and await the bearing
and distance of the next plot location which was called out by the
other team member. The distance and direction was chosen by using a
random number table photocopied onto a 3x5 index card. Distances
varied from 0 to 9 meters and random bearings were selected from a
180 degree "forward pointing" semicircle, i.e., no backtracking.
Once the angle and distance coordinates were chosen, the plot was
established with the PVC extensions each marked at 50 cm intervals.
Recorded separately at each 50 cm x 50 cm and 100 cm x 100 cm plot
location were species presence or absence and an estimate of the
percent cover in the 100 x 100 cm plot based on the Braun-Blanquet
scale (less than 5%, 5-25%, 25-50%, 50-75%, greater than 75%).
35
-------
Figure 15. Structural classification of the microenvironment
associated with each nested frequency sampling plot.
36
-------
A qualitative aquatic index value was recorded at each 100 cm
x 100 cm plot location. Figure 15 schematically illustrates the
classification used to assign aquatic index values. This index
was: aquatic (A)= greater than 50% plot area saturated or
inundated; intermediate aquatic (IA)= 25-50% plot area saturated or
inundated; and upland or non-aquatic (NA)= less than 25% of plot
area saturated or inundated. The purpose of this index was to gain
some qualitative measure of wetland character and degree of
inundation. When a plot was inundated, the water depth was
recorded. Aquatic index values are of course a function of influent
flow rates and weather conditions.
Results and Discussion
Spatial heterogeneity in plant cover for the treatment cells
and the control is revealed by results of the nested frequency
sampling survey (Tables 3-6). For example, A. filix-femina and C.
obnupta have relatively low frequencies along transect C relative to
the other transects (Table 3). There is also a distinct difference
in frequency values for A. fi 1 ix-femina on an east vs. west basis
for cells 1 and 2 relative to the control (Table 6). The data
suggest the adverse effects of the operational hydroperiod in the
western section of cells 1 and 2 on this fern. However, for C.
obnupta the east vs. west difference in frequency values is not
apparant. For C. obnupta these spatial differences imply that (1)
wastewater flooding has had a greater impact adjacent to the
37
-------
Table 3. Frequency along each transect of dominant plant species in
nested 50 cm x 50 cm and 100 cm x 100 cm plots. A species that
occurs in the 50 cm x 50 cm plot automatically also occurs in the
100 cm x 100 cm plot. Species acronyms are given in Appendix C.
Frequency (X) 1n
50 cm x 50 cm plot 100 cm x 100 cm plot
Life
Transect
Transect
Species
Form
A
B
C
G
0
£
F
A
e
C
G
0
E
F
Atf 1
fern
26
26
13
41
29
24
38
50
61
23
71
50
53
50
Pomu
fern
23
9
7
12
21
18
6
27
30
13
18
29
18
16
Caob
sedge
50
52
17
65
50
76
88
68
70
23
82
79
88
100
Lem1
aquatic
0
0
7
24
36
0
6
0
0
7
24
36
0
6
Loin
shrub
9
4
7
0
7
12
3
23
4
7
18
29
18
13
Lyam
herb
27
22
13
0
7
0
0
64
48
23
6
7
6
3
Oesa
herb
36
35
23
6
29
18
22
73
52
47
18
36
29
28
Rusp
shrub
18
9
13
29
43
18
9
32
30
23
47
43
35
9
Table 4. Frequency by cell of dominant plant species in nested
50 cm x 50 cm and 100 cm x 100 cm plots. Species acronyms are given
in Appendix C.
Frequency (X) In
50 cm x 50 cm plot 100 cm x 100 cm plot
Control Cell 1 Cell 2 Control Cell 1 Cell 2
Species Transects Transects Transects Transects Transects Transects
A i B C, G. & 0 E i F A & 9 C, 6, & D E «. F
Atf 1
24
25
33
56
39
78
Pomu
16
11
10
29
18
16
Caob
51
38
84
69
52
96
lemi
0
18
4
0
18
4
Loin
7
5
6
11
15
14
Lyam
24
8
0
56
15
4
Oesa
36
20
20
62
36
29
Rusp
13
25
12
31
34
18
-------
Table 5. Comparison of percent cover values: Permanent Plot Method
(PP) (1986) vs. 1986 Nested Frequency (NF) Sampling Method. Species
acronyms are given in Appendix C.
Control Cell 1 Cell 2
X Cover X Cover X Cover X Cover X Cover X Cover
Species PP Method NF Plots PP Method NF Plots PP Method NF Plots
Atf 1
6.9
3.1
0.7
4.6
0.8
4.1
Caob
20.6
19.5
21.4
22.0
63.3
55.7
Lem1
0.0
0.0
0.0
8.3
0.0
1.3
Loin
4.0
0.3
1.0
2.2
5.5
4.5
Lyam
11.6
9.6
3.6
2.3
1.7
0.10
Oesa
2.3
2.1
1 .6
2.3
1.7
1 .0
Pomu
4.0
3.2
1.0
3.0
23.0
2.4
Rusp
1.3
4.5
9.7
10.0
0.0
1.7
Table 6. 1986 frequency results of dominant plant species separated
on an east vs. west basis. The number to the left of the slash is
the frequency in the 50 cm x 50 cm plot size, and the number to the
right of the slash is the frequency in the 100 cm x 100 cm plot size.
Frequency 1n 50cm x 50cm plot size (X)/frequency 1n lm x lm plot size (X)
Control Cell 1 Cell 2
Species East 1/2 West 1/2 East 1/2 West 1/2 East 1/2 West 1/2
N=22 N-23 N=31 N=30 N=24 N=25
Atf 1
27/59
22/52
26/58
23/30
54/79
12/24
Caob
64/73
39/65
45/58
29/47
92/96
76/96
Lem1
0/0
0/0
3/3
33/33
0/0
0/8
Loin
0/5
13/22
6/20
3/10
0/0
12/28
Lyam
32/68
17/43
10/20
6/10
0/8
0/0
Oesa
50/73
22/52
26/53
13/20
38/54
4/4
Pomu
14/27
17/30
16/23
7/13
17/25
4/8
Rusp
5/27
22/35
29/32
20/37
21/25
4/12
39
-------
Table 7. 1986 aquatic index/rooting environment frequencies. Cell
2 data is incomplete for the aquatic index and is omitted. Results
are based on N = 45, control; N = 61, cell 1; and, N = 49, cell 2.
Frequency (*)
Aquatic Index Rooting Environment
rooted 1n rooted 1n
Location NA IA A log/upland muck N
Control 22 55 22 44 56 45
Cell 1 12 31 57 36 67 61
Cell 2 — — — 19 81 49
NA= nonaquatlc or upland; sample plot less than 25% Inundated
IA= Intermediate; sample plot 25-50* inundated
A= aquatic; sample plot greater than 50* inundated
Table 8. Dominant species frequencies in the three aquatic index classes.
Frequency
(X)
Life
Cell 1
Control
Species
Form
A
IA
NA
A
IA
NA
Atf 1
fern
25
58
8
16
48
36
Caob
sedge
41
41
18
21
62
17
Lem1
aquatic
82
12
0
0
0
0
Loin
shrub
22
56
22
33
50
17
lyam
herb
67
33
0
28
60
12
Oesa
herb
18
50
32
21
58
21
Pomu
fern
0
54
46
0
38
62
Rusp
shrub
14
57
29
7
64
29
40
-------
discharge header than elsewhere in the wetland, or (2) C. obnupta
had a lower frequency along transect C prior to the 1984 initiation
of effluent discharge into the wetland. Figure 10 (permanent plot
method) lends support to the second hypothesis, indicating C.
obnupta had a lower cover in cell 1 relative to the control and cell
2 prior to the initial discharge of wastewater (though not shown,
transects C and 0 were approximately the same in C. obnupta cover
for 1984, ca. 20%)
Another distinct spatial pattern is found for Lemna minor.
It is a floating aquatic, absent from the control section, but
present in areas of deeper water. Note the increase in frequency
values in cell 1 from the discharge header (transect C) to the
middle dike (transect D). In cell 2 it is absent along transect D
and is present along transect F (Table 3). Among the factors that
determine the distribution of this species are water availablity,
dissolved nutrient supply and canopy cover (shading). It is
expected that this species will increase in cover as the canopy
thins out (reduction in red alder density via flooding stress).
Explaining the distribution of L. minor and then attempting to
predict future change is further complicated by the presence of
dabbling ducks. Ducks, which feed on L. minor and influence its
cover locally, were found in the western sections of the treatment
cells throughout the summer of 1986.
Prior to sampling in 1986, Oenanthe sarmentosa. the water
parsley, was predicted to have a higher frequency in cell 1 relative
41
-------
to the control. This prediction was based on the observation of 0.
sarmentosa growing in large floating mats in the highly enriched
aeration basin, east of U.S. 101, which receives the incoming raw
sewage. Table 4 shows this prediction has not yet been borne out;
the frequency value of 0. sarmentosa for cell 1 is approximately
half of the control value. There is also a distinct east vs. west
difference shown in Table 6 — frequency values are greater in the
higher and less frequently inundated eastern section of the control
and cells 1 and 2. Cover data yields a somewhat different pattern:
Figure 10 shows 0. sarmentosa cover to be initially less in cell 1
relative to the control and cell 2 for 1984 and the cover values for
0. sarmentosa calculated from the nested frequency data indicate
nearly equal cover values in 1986. The cover data indicates an
increase in 0. sarmentosa for cell 1 relative to the control (the
small values for cover, however, make this interpretation difficult
to defend).
Surprising results are shown in Table 5, where the permanent
plot method is compared to the nested frequency method for percent
cover of the dominant plant species. Percent cover values for C.
obnupta (control and cell 1), L. involucrata (cells 1 and 2), L.
americanum (control and cell 1), 0. sarmentosa (control and cell 1),
P. munitum (control), and R. spectabilis (cell 1) are extremely
close for both methods. Sampling occured nearly three months later
with the nested frequency method and the survey was conducted by
several different individuals than the permanent plot survey crew of
42
-------
April 26, 1986. The problem with the permanent plot method,
however, is the small sample size that, although quite consistent
with the nested frequency method on an overall basis, is much too
small to make spatial separations in the data for reliable
description and trend assessment (Table 6).
Tables 7 and 8 show results of the aquatic index. Data were
incomplete for cell 2 and are therefore omitted. It is not
surprising that cell 1 has a larger frequency for A (aquatic) than
in the control given summer flooding in cell 1 for three years. Of
more importance, however, is the relation of the aquatic index to
dominant species shown in Table 8. For example, species with
relatively lower frequencies in the aquatic zone (A) — Athyrium
filix-femina (Atfi), Lonicera involucrata (Loin), Polystichum
munitum (Pomu), and Rubus spectabilis (Rusp) — are likely to be
adversely affected by an increased hydroperiod relative to the
current operational hydroperiod. Note that Carex obnupta and
Lysichitum americanum have larger frequency values for the aquatic
zone in cell 1 relative to the control, indicating possible
expansion into these areas. Future studies should assess this
observation.
Table 7 also shows the frequency of the rooting environment
(either rooted in muck, or rooted in a log or on an upland hummock)
for the treatment cells and control based on 155 nested frequency
plots. The data indicate more upland area and woody debris in the
control than in the two treatment cells. Cell 2 has substantially
43
-------
more 'muck area' than the control and cell 1. This spatial
difference in substrate influences the type of species that can
survive and colonize a given area of the wetland under selection
pressure of different operational hydroperiods. Spatial variablity
in the substrate (and thus plant cover) and quantity of woody debris
may also influence settling and resuspension rates of suspended
solids throughout the wetland.
Recommendations for Resampling
The nested frequency method described herein is a relatively
rapid means of sampling vegetation under the very difficult
conditions of the Cannon Beach wetland. For a two person team, one
can expect to sample ca. 10 plots per hour given the method
described. Sampling should occur just after the peak of the growing
season (August - September) and should be coordinated with the
treatment plant operators so that wetland influent is shut-off at
least five to seven days prior to sampling. It is recommended that
collapsible but convenient quadrat frames be used rather than the
PVC pipes. However, the importance of minimizing hand-held objects
should not be underestimated.
There is no need when resampling to establish flagging on
transects A through F. A hand-held sighting compass is sufficient
to maintain the desired bearing. The use of a telescopic stadia rod
for establishing each quadrat location (as described in the methods
section) is highly recommended.
44
-------
After resampling the wetland using the nested frequency
method, an Analysis of Variance Test can be used to determine (1)
if there are significant differences between the species means for a
given plot size and (2) if more than two years data are available,
establish the basis for performing the least significant differences
(LSO) test.
A warning for any future researchers that repeat the
vegetation sampling methods described in this report: interpret the
results with care. Because so little autecological information is
available for the species located in the Cannon Beach area (under
normal conditions much less being flooded with secondary effluent)
one can draw erroneous conclusions from changes in plant cover and
frequencies. For example, the operational hydroperiod and enriched
wastewater not only affects species physiologically by reducing
oxygen availablity to the roots, but can also transport propagules
not usually transported by water during the summer (as in the
control). Neither the permanent plot method nor the nested
frequency sampling method will yield any information on the
significance of this phenomenon. Other complications include the
effects of relatively dry and early spring weather (e.g., 1987) that
enables some plant species to establish significant growth prior to
wastewater discharge compared to wetter and colder springtimes —
how does this affect plant survival, reproduction, and ultimately
plant distribution and succession?
45
-------
WATER BUDGET ESTIMATE
Introduction
Hydrology is perhaps the key determinant influencing the type
of wetland found in a given area (Mitsch and Gosselink 1986). The
water regime is also of prime importance in determining pollutant
inputs and outputs of a wetland. Therefore, understanding
site-specific hydrological characteristics of the Cannon Beach
wetland is necessary to make informed management decisions as the
wetland wastewater treatment system ages and in the event of
undesirable changes in water quality and wildlife habitat. For
example, a key relationship for the site would be the area of
inundation as a function of the operational hydroperiod (water
depth, frequency and amplitude of flooding, and duration of
flooding), which is directly related to the influent flow rates.
These data could be used to control, in a more systematic fashion
than present, the desired wetland plant and animal community (i.e.,
fewer or more trees, greater production of Daphnia sp. and other
invertebrates, etc.) while still maintaining necessary wastewater
treatment.
Channelization and water dispersal and their relationship to
vegetation change in the treatment cells have been discussed in the
previous section. These hydrologic and biotic interactions also
influence pollutant removal, transformation, and transport processes
in the wetland.
46
-------
Seasonal and summer operational varlablity in the quantities
and rates of wetland water inflows and outflows determine the
storage status of the wetland. To infer pollutant inflows and
outflows from the water budget requires precise knowledge of water
budget components. This section considers a simple mathematical
model to estimate the water budget for the treatment cells.
Methods
The data collected for the water budget model are derived
from Cannon Beach water quality monitoring reports, literature
reviews, and limited field work. The only original data are water
depth measurements obtained by staff gauge observations.
Lack of funding prevented application of more sophisticated
techniques of study, but an attempt is made to demonstrate the
utility of simple, inexpensive methods of studying wetland hydrology.
Water Budget Model
Figure 16 shows measured hydrologic characteristics of the
Cannon Beach wastewater wetland for summer operation during 1985 and
1986. Two striking characteristics of Figure 16 are (1) the
significant water loss in the wetland, and (2), the earlier start of
effluent discharge into Ecola Creek for 1986 (51 days) relative to
that in 1985 (101 days). These two chacteristics are investigated
by a simple mathematical model conceptually depicted in Figure 17.
The fundamental concept used to calculate the water budget is
a water mass balance for the wetland water sheet; in words, the
47
-------
.6
Cell 2 empty
1-* Dry Weather
o
o
5.*
-
LU
¦- .3
<
CE
-
A
o
/\
/
\
".1
-
•
.0
May J#« Jul Aug Sep Oct
19 8 5
~ Wetland influent (from sewage lagoons)
H Wetland etftuenl (to Ecola Creek)
• O Rainfall recorded at Cannon Beach City Hall
Mir Jon Jul Aug Sep Oct
1986
*Plant operator notes.
Wetland effluent flow rate for 1985 is not measured by stream qauqe
but is. an estimate by the treatment plant operator.
Figure 16. Hydrology of the Cannon Beach wastewater
wetland -- 1985 vs. 1986.
48
-------
(Time rate of change of water mass in the wetland water sheet) =
(rate of water input Into wetland) - (rate of water output out of wetland)
-0
r
GUI Z
. <
l O
1
Ct
II
Q\
Q. +
Parameters
Measured Estimated
V= volume of water storaqe in wetland
£V= change in V over the balance period
wetland influent flow rate
Q0= wetland effluent flow rate
P= precipitation rate
S^= surface inflow rate (excluding 0 -j)
Gj- ground water inflow rate
C)0= ground water outflow rate
ET= evapotranspiration rate
^ - density of water
The uNi-hi of eac A
are mass/+ i/ne
Figure 17. A conceptual model of the water budget for the
Cannon Beach wastewater wetland.
49
-------
concept is (time rate of change of water mass within the water
sheet)= (rate of water input into wetland) - (rate of water output
out of wetland). Mathematically, the relationship is expressed by
the following equation
(AV)/At = (Qi ^ + S.^+ G. ^ ) + (ET^ + GQ£+ Qq£ )
Each term in equation 1 is defined in Figure 17. The density
term, is common to each term in equation 1 and cancels out
resulting in the following equation
AY/ At = (Q. + P + S. + G.) + (ET + G0 + Q0)
Also, Av = AL A(L) where AL is the average water level in
the treatment cells and A(L) is the area of inundation as a function
of L.
As a first order approximation assume surface inflows
excluding the wastewater influent () and ground water inflows
(G^) are negligible for the balance period. Heavy rainfall is
infrequent during the summer and this assumption appears valid. In
addition, if the effluent flow rate (Qq) is the same at the
beginning and end of the balance period, assume AL = 0, thus Av =
0. Under these assumptions equation 2 reduces to
0=Q. +P-ET-G -Q
l o o
(3)
-------
Measured quantities for equation 2 are Q.., P, and Qq. If
an evapotranspiration rate (ET) is assumed then the ground water
outflow rate can be estimated as the residual of equation 3:
Gq = Q. + P - ET - Qo (4)
Figure 18 shows the balance period used for this model from
July 16 to October 2, 1986; effluent flow rates (Qq) are 0.098
and 0.099 HGD respectively and influent flow rates are 0.441 and
0.486 MGD respectively.
For the measured va lues of Q., P, and Q the units of
i o
volume/time are converted to cm/time by the following formula:
(Xgal/day)(3.069x10_6ac-ft/gal)(12in/ft)(2.54cm/in)/l5 ac
= Y cm/day
From Figure 18, total values for Q.., P, and Qq for the
balance period are:
Q^= 188.8 cm added to the wetland
P= 17.8 cm measured by a local rain gauge
0 = 19.9 cm lost from the wetland by surface outflows
o
Given the aforementioned assumptions, all that is needed to
estimate ground water outflows (Gq) is an assumed mean ET rate for
the treatment cells over the balance period. Unfortunately no
51
-------
,TJV '<¦>
Oc-h Z
B^iAftc e
Feri cA
O = Ralntall
w«tl«nd ln(iu«nt
o.oiy
10 tO SO 10 20 «0
Stpltmtti *4* October »i
1986
Figure 18. The balance period used to estimate the water budget
for the Cannon Beach wastewater wetland.
52
-------
direct evapotranspiration measurements are available for Oregon
coastal zone swamp forests. In the absence of direct measurements
pan evaporation data from the Astoria experiment station (Table 9)
are used. Pan evaporation rates for Astoria are likely to be a poor
surrogate for evapotranspiration at the wastewater wetland because
the evaporation pan is more a giant slow response thermometer than
an index of actual evapotranspiration rates over heterogenous stands
of vegetation (William P. Lowry, personal communication 1987).
However, pan evaporation data can be used to estimate a range of
possible ET values to test the model's sensitivity to assumed ET
rates (Figure 19).
Table 9. Astoria pan evaporation data (Astoria Experiment Station
1963-1972).
Astoria Pan Evaporation Rate
Rate
Rate
max
Rate
No. of
Total
Month
cm/month
cm/month
cm/dav
# days
cm
July
12.0 1.4
13.4
0.43
16
6.9
Aug
10.6 1.3
11.9
0.38
31
11.9
Sept
7.6 1.4
9.0
0.30
30
9.0
Oct
4.4 0.9
5.3
0.17
2
0.3
Total=
28.1 cm
53
-------
(G0/(Qi ~ P)) * 100, where G0= Q-j + P-ET- Q_
• (188.8 ~ 17.8 - ET - 19.9) cm
• (186.7 - ET) cm
O
o
x
vo
vo
o
CM
X
o
CD
II
O
O
Q.
+
o
CD
100r
,ET« Astoria Pan Evap.
= 28.1 cm tota I
= 0.36 cm/day
ET= 2 x Astoria Pan Evap.
= 56.2 cm total
0.72 cm/day
J I 1—
304050
70 80
Estimated Total Evapotranspiration (cm)
(July 16 - October 2, 1986)
Figure 19. An estimate of ground water outflows (G0) expressed as
a percentage of total water inputs into the Cannon Beach wastewater
wetland (G0/(Q-j + P) x 100) for the balance period of July 16 to
October 2, 1986.
54
-------
Figure 19 graphs estimated ground water outflows as a
function of total estimated evapotranspiration for the balance
period. Note that the calculated ground water outflow is relatively
insensitive to large variations in assumed evapotranspiration
values. For example, given an estimated ET= 28.1 cm (0.36 cm/day),
which is the Astoria pan evaporation rate, Gq= 77% of Q.. + P.
Even with a value double the Astoria pan evaporation rate, ET= 56.2
cm (0.72 cm/day), Gq is still 63% of Q.. + P. This latter value
of ET is highly unlikely given that the maximum possible ET rate for
any forest is 0.6 cm/day (Waring and Schlesinger 1985). With the
generally foggy conditions at Cannon Beach during the summer the ET
value of 28.1 cm may even be too high, but only direct site
measurements will lay the question to rest. A hypothetical curve is
also shown in Figure 17 where Qq= 150 cm; in this case
proportional change in GQ is greater as a function of ET than the
actual curve and its magnitude is only 10-20% Q.. + P.
Discussion
Calculating components of a wetland water budget as a
residual of a mass balance equation, such as equation 1, may lead to
significant errors; the component that is derived may contain an
unknown error which may be larger than the value of the component
itself (Solokov and Chapman 1974). For long-term, regional studies
the storage term, Av, can be considered equal to zero, however, for
relatively short balance periods (e.g., a few months) and smaller
55
-------
areas, & V is generally not equal to zero and even more precise data
1s needed for the storage term (Solokov and Chapman 1974, Oooge
1975).
How valid is. the assumption for the Cannon Beach site that
AV equals zero for the balance period? Effluent flow rates for the
beginning and end of the balance period are essentially identical (a
difference of 0.001 MGD or 0.006 cm/day). The total amount of water
added to the wetland 5 days prior to July 16 (Q^ + P, with
S^=G|=0) equals 13.8 cm and the total amount of water added to
the wetland 5 days prior to October 1 (Q^ + P, with S..=G.=0)
equals 16.2 cm, only a 15% difference. Thus, the assumption that
the water level had an approximate net change of zero (AL=0) for
the balance period' appears reasonable.
The literature offers conflicting information regarding the
magnitude of evaporation from open water relative to
evapotranspiration from wetland vegetation. In a forested pond
cypress dome in Florida, Heimburg (1984) reported swamp
evapotranspiration as 80% of pan evaporation during the dry season
(spring and fall) and as low as 60% of pan evaporation during the
wet season (summer). Eisenlohr (1966) found 10% lower
evapotranspiration from vegetated prairie potholes than from
non-vegetated potholes in North Dakota. However, Hall et al.
(1972) estimated, through measurements and calculations, that a
stand of vegetation in a small New Hampshire wetland lost 80% more
water than did the open water in the wetland.
56
-------
Because of the conflicting measurements and difficulty of
directly measuring evapotranspiration from wetlands Linacre (1976)
concluded that neither the presence of wetland vegetation nor the
type of vegetation had major influences on evaporation rates, at
least during the active growing season. Another complication is
that the red alder in the western portion of the cells undergoing
flooding stress have reduced transpirational rates relative to the
unstressed red alder and large Sitka spruce in the eastern portion
of the cells.
If the initial assumption that Av= 0 over the balance period
is valid, then the water budget model indicates ground water
outflows account for 65-85% of the total water input into the
wetland treatment cells for the balance period of July 16 through
October 2, 1986 (Figure 19). A large proportion of the water loss
via 6q likely flows laterally east towards Ecola Creek. Lateral
flow is probably impeded by the highway barrier west of the
treatment cells. Only a more detailed study using shallow wells
will be able to determine this aspect of the water budget more
precisely.
Why wetland effluent (Qq) starts so much earlier and at a
greater magnitude in 1986 relative to 1985 is difficult to answer
given so little measured data. If the operator's notes are correct
one would at first assume that ET is much greater in 1985 vs. 1986
because of the dry weather noted. However, the previous analysis
57
-------
suggests Gq is the primary sink for wetland water inputs (Q.. +
P). The various possiblities that may account for this difference
include:
1. The :Gq rate was greater in 1985 than 1986, with ET
rates about the same for both time periods or perhaps
ground water storage capacity is becoming saturated.
2. Both G and ET were greater in 1985 relative to
o
1986.
3. The water budget model in this paper is incorrect
with G smaller than estimated and ET the most
o
significant sink for water in the wetland (thus ET was
much greater in 1985 than 1986).
4. The treatment plant operator was incorrect in his
estimate of both Qq timing and magnitude.
Whichever the case may be, the consumption of water by ET +
G in the wetland for 1986 is substantial. The implications of
o
the significant water consumption in the wetland via ET + Gq are
discussed in the next section.
58
-------
WATER QUALITY
Materials which enter and leave the wetland treatment cells
are associated with solids, dissolved in water, or are gases. Total
inputs and outputs of these materials are dependent on the dynamics
of the hydrologic regime (Kadlec 1985). For example, large flow
rates mean higher flow velocities, more aeration and greater
resuspension rates of settable and suspended solids in different
sections of the wetland. Figures 20 and 21 display BOD,, and
suspended solids data taken along two approximately perpendicular
transects in cell 1. The water is deeper in the western section of
cell 1 and with large amounts of woody debris appears to flow more
slowly in this section than in the middle and eastern section of
cell 1, which has more clearly defined channels. Variable flow
rates are the norm for the treatment cells, but even with a constant
influent flow rate extremely small gradients throughout the wetland
can cause spatial variations in flow velocities resulting in complex
spatial variablity in settling and resuspension rates of algal
detritus, flocculent material, plant litter, and inorganic
substances. Seasonal variations in ambient temperature and the
hydrologic regime (hydroperiod and water budget) not only directly
affect wetland physical and chemical treatment processes but
influence the vegetation cover and subsequent litter production, the
decomposer and invertebrate community, and avifauna and mammalian
59
-------
0 20 40 60 80 100 120 140
WCST Dlatance along transect C (maters)
1 ¦
160 180
EAST
E 30
(B)
Q
o
ffi
20
10
.4
.11 /
.7
/\ 9
\ y*
10
Influent
*3
8
Sampling date: 8/26/86
*2
Each sample point is a
single grsb sample
•6
20 40 60 80 100 120 140 160
Distance from dlacharge header (meters)
180
Sample locations
,io
U CELL
*7
'j*> 5 4 II } it
/"* * 1 • w w
_f 1_
Figure 20. Spatial variation of five-day biochemical oxygen demand
(BODc) in cell 1. (A) Grab samples taken along transect
C. (B) Grab samples taken along a transect perpendicular to
(A).
60
-------
(A)
70
60
50
o
E
40-
rvT
30
*-
20-
10-
Influent
0 20
WES T
40 60 80 100 120 140 160 180 200
(Distance along transect C (meters) EAST
(B)
70-
60-
50-
40-
« 30
20
10
2
Influent
i H ^ S/_
Sampling date: 8/26/86
Each sample point is
a single grab sample
8
8
10
20 40 60 80 100 120 140
Oistance from discharge header (melees)
160 180
Sample locations:
to
+—(¦
U CELL I
'*1
'~J-*"-J-* 1-
W t ,tt t 1, t
Figure 21. Spatial variation of total suspended solids (TSS)
in cell 1. (A) Grab samples taken along transect C. (B) Grab
samples taken along a transect perpendicular to (A).
61
-------
(A)
240 -
220-
200-
ieo -
160"
140 -
-Vs.
CT
r—
120-
c
V
CO
100-
o
o
80-
CO
60 -
40-
20 *
1984
INFLUENT
FAC. LAGOON EFFL."
Vtf EFFL.-
(B)
(C)
300 -
275*
250 *
225 -
200 -
175 -
0>
F
150 -
to
125-
n
O
100-
CD
7rj -
50-
25-
240 -
220 ¦
200*
180 -
160-
140 *
o
120-
c
V
100-
o
O
80-
CD
60 -
40"
20 -
0 -
INFLUENT
1985
FAC. LAGOON EFFL
W effl.
influent
1986
FAC. LAGOON £FF1 .*
Figure 22. Mean monthly biochemical oxygen demand data for the
Cannon Beach wastewater treatment system -- 1984 - 1986. Sampling
frequency is approximately biweekly. (Cannon Beach monitoring
reports and Thompson and Minor 1986). Data is incomplete for 1986.
62
-------
activity; all of these factors can strongly influence total inputs
and outputs of waterborne substances (Kadlec 1985, Guntenspergen and
Stearns 1985).
Complex spatial variations were observed for Daphnia sp. and
copepod population densities throughout the wetland during the
summer of 1986. Daphnia feed on phytoplankton and have been found
to clarify wastewaters in experimental treatment systems (Oinges
1976) but can also stimulate growth and productivity of some algae
species (Porter 1976). The Daphnia ranged from white to deep red in
color, indicating different dissolved oxygen concentrations
throughout the wetland (Oinges 1976). The importance of
phytoplankton grazing by invertebrates in suspended solids reduction
as compared to physical settling processes for the Cannon Beach
wetland is not assessed.
The Cannon Beach wastewater wetland treatment system has
consistently met the 10/10 limitations for three years operation
(Figure 22 and 23). Because of the significant water loss via Gq
+ ET the wetland has been a very effective particulate filter given
loading rates of BOD and TSS of the past three years operation.
Figures 24 and 25 show BOD and TSS loading rates of the wetland and
Ecola Creek for 1986.
Cannon Beach operators report 'removal efficiencies' for BOD
and TSS to ODEQ. Removal efficiencies are calculated by comparing
effluent concentrations and total loading (concentration x flow
rate) of BOD and TSS to that of the influent. These water quality
63
-------
(A)
E
c/?
CO
?40
r.ro
TOO
180
1*0
HO
i?0
100'
pn
pn
40
20
0
1984
influent
FAC. LAGOON EFFL-
W CFFL.-
(B)
CT
E
CO
240
220
200
ieo
1*0
mo
1 20
100
80
60'
40
20 ¦
0
FAC. LAGOON ETFL
W effl.
1985
CT
E
«w
w
CO
?40
r?o
200
ieo
if o
140 ¦
1?0-
100"'
pn
*0
«*
0'
INFLUENT
FAC. LAGOON EFFL.'
W iFFL
1986
U
O
Figure 23. Mean monthly total suspended solids data for the Cannon
Beach wastewater treatment system -- 1984 - 1986. Sampling
frequency is approximately biweekly. (Cannon Beach monitoring
reports and Thompson and Minor 1986). Data is incomplete for 1986.
64
-------
>.
a
O
o
o
OQ
CO
x>
80-
70-
60-
so- \
20- \
10- \
-k Wetland Influent
>Wetland Effluent
Data not available for Jl and Aug
Jl
1986
Figure 24. Mean monthly biochemical oxygen demand loaading rates
for the wetland treatment cells and Ecola Creek — 1986. Data is
not available for July and August.
soor
260-
200 -
160
¦« Wet I and Influent
„ Wetland Effluent
100-
1986
Figure 25. Mean monthly total suspended solids loading rates
for the wetland treatment cells and Ecola Creek -- 1986.
65
-------
X Orthophosphate, PO|"
o Total Phosphorus
Each sample point is
a single grab sample
Cell J
¦ f» i)
a (to)
Xdtj
Cell 2.
C 1-C2
overflow
Discharge
header
C2 outfall to
£cola Creek
Distance from discharge header (meters)
Figure 26. Orthophosphate and total phosphorus data for the wetland
treatment cells -- 1986. Laboratory analysis conducted at the
Cannon Beach wastewater treatment plant using a Hach test kit
(Model # PO-24).
66
-------
constituents are measured from composite grab samples taken on the
same day. The term 'removal efficiency' as reported can be highly
misleading. Comparing outflow concentrations to inflow
concentrations without accounting for the time lag of advective
transport of these materials through the wetland can lead to
erroneous conclusions, especially if determining nutrient or fecal
coliform 'removal efficiencies'.
Because of time and funding constraints nutrient sampling for
the 1986 operational season is limited to two sampling periods for
phosphorus (Figure 26). The data is so limited as to prevent
drawing conclusions about phosphorus 'removal efficiencies'.
Samples were unfiltered and not representative of the 'undisturbed'
wetland water sheet. It is very difficult to obtain representative
water samples in the wetland without stirring up detritus and muck,
so future phosphorus sampling should focus on filtered samples and
dissolved phosphorus species. Phosphorus removal tends to be poor
in bogs with low soil mineral content and removal tends to be
highest in mineral soils in terrestrial environments (greater than
90% removal) (Richardson 1985). No soil chemical analyses have been
conducted on Coquille soils, but the Cannon Beach wetland is
probably intermediate in the soil inorganic-organic matter
spectrum. Ecola Creek is not considered nutrient sensitive because
of its proximity to the ocean.
67
-------
SUMMARY AND CONCLUSIONS
The City of Cannon Beach uses a seven hectare red alder/
slough sedge/twinberry palustrine wetland, divided into two cells in
series, to treat chlorinated effluent from a four-cell aerated/
facultative lagoon sewage treatment system. Operation began June,
1984. The system has proven to be an effective means of meeting
summer wastewater discharge limitations (10 mg/l BOD,, and 10 mg/l
TSS).
The primary purpose for biological monitoring, established by
special conditions of the Corps of Engineers 404 permit, is to
provide data to assist evaluation of future wetland treatment system
proposals in the region. Research will also permit development of
monitoring procedures for improved wastewater wetland management. A
vegetation sampling scheme (referred to as the permanent plot
method) was devised and baseline data collection performed in 1984
by staff of the Corps of Engineers, Oregon Department of Fish &
Wildlife, and the U.S. Fish & Wildlife Service. This vegetation
survey was repeated in 1986. Results indicate herbaceous and shrub
vegetation changes since 1984 are relatively minor. However, a
decrease in herbaceous cover has occurred in channelized and deeply
flooded areas, especially towards the western section of the
treatment cells. The western section is at a lower elevation
relative to the eastern section and thus has deeper water levels for
given influent flow rates. The data also indicate an increase in
68
-------
slough sedge cover in the eastern section of cell 2 which has
shallow water levels (3-10 cm) and frequent drawdowns.
A complex pattern of flooding stress, consisting of
defoliated trees, sparsely leaved trees, and some dead trees, is
observed in 1986 for red alder, and to a lesser extent twinberry, in
deeper water towards the western section of the treatment cells and
adjacent to the dikes where water ponds. The hummocky topography
complicates prediction of the extent and stability of flooding
stress because high hummocks in the western section of the cells
contain trees and shrubs of relatively good vigor. Prediction is
further complicated by red alder being blown over in storms and
subsequently surviving by adventitious rooting and apical dominance
shifting to former branches.
Field research conducted April through September 1986
developed a method of herbaceous vegetation sampling using the
nested frequency method; one hundred fifty-five nested frequency
sampling plots (50 cm x 50 cm nested within 1 m x 1 m) yield a
baseline data set for future vegetation trend analyses. An
excellent agreement was found for percent cover of herbaceous and
shrub vegetation as determined by the nested frequency method (155
plots) and the permanent plot method (22 herbaceous plots and 22
shrub plots). The nested frequency method, however, yields
substantially more plots per man-hour and thus more sampling
resolution per man-hour than the permanent plot method.
69
-------
For 1986, wetland influent and effluent flow rates were 0.40
and 0.038 MGD respectively. An independent water budget estimate
suggests ground water infiltration is at least 65-85% of the water
loss.
For three years the Cannon Beach system has met ODEQ water
quality discharge standards (10/10 limitations). BOO,- and TSS
D
wetland influent concentrations were reduced by 40% and 85%
respectively. Phytoplankton constitute the bulk of the suspended
solids discharged into the wetland. Within thirty to forty meters
into the first treatment cell the concentration of phytoplankton is
greatly reduced; the suspended solids in the wetland water sheet
are thereafter predominantly composed of plant detritus and humic
materials. Settling and resuspension rates of suspended solids vary
in a complex manner throughout the treatment cells. The relative
importance of physical settling, physical filtration by vegetation
and woody debris, invertebrate grazing, and reduced solar radiation
by the canopy cover to phytoplankton removal is not quantitatively
assessed by this study.
The loss of tree species is not likely to adversly affect
water quality treatment, especially for BOD and suspended solids.
Bacterial metabolism is the primary removal mechanism for BOD and
colloidal solids; sedimentation is the primary removal mechanism
for settable solids (Tchobanoglous and Gulp 1980). Longer term
retention of nutrients by trees may diminish in a portion of the
cells with extensive tree death, but the effect on overall nutrient
removal and transformation will probably be minimal.
70
-------
As the treatment system ages, passerine bird habitat will
decrease in the western section of the cells as the tree density
decreases, but waterfowl habitat will increase. Slough sedge cover
is likely to remain relatively stable and perhaps increase in
shallow portions of the cells.
Monitoring needs to be improved and coordinated with
management decisions. For example, how should dike vegetation be
maintained? Given the ability to manipulate water levels and the
dispersal pattern, how should these be varied and for what purpose?
Monitoring Recommendations
Two levels of monitoring are described in this section. The
first level is least expensive and the minimum recommended level of
monitoring. The second level is more costly but yields more
information. Achieving the first level does not mean that the
second level should not also periodically be reached — many
components of the second level require a single effort followed by
less demanding measurements. Monitoring recommendations are limited
to the wetland treatment operation of the Cannon Beach wastewater
treatment facility.
Level 1
1. Plant operators should:
a. Maintain a wetlands observation record that
includes a record of any changes made in
71
-------
discharge header wastewater distribution and weir
levels (i.e., type of change, reasons for change,
date of change), and a record of weather conditions.
b. Measure the hydroperiod at three water sampling
stations located in the western portion of the
treatment cells on a bi-weekly basis while taking
required water samples. Either a stationary staff
gauge or a mobile measuring rod could be used, with
an emphasis on a repeatable and consistent method.
c. Determine the area of inundation as a function of
influent flow rates to gain some understanding of
the operating range of the system.
2. It is highly recommended that the nested frequency
sampling method be repeated every two years; the survey
should be repeated during August-September. A greater
number of sample plots per man-hour can be obtained with
the nested frequency method than with the permanent plot
method, resulting in greater sampling resolution per
man-hour.
3. At minimum, the permanent plot sampling method should be
repeated every two years. The influent should be shut
off for a one week period during the vegetation
72
-------
sampling survey. This survey need not take place 1n the
same year as the nested frequency sampling. Researchers
should integrate hydroperiod information with vegetation
survey data and draw inferences with caution.
4. Repeat the tree survey for basal diameter in the summer
of 1989 and thereafter at 3-year intervals.
Observations should be made on tree vigor for each
tagged tree as well as generally throughout the wetland.
Level 2
1. Level 1 monitoring in addition to a detailed topographic
survey. The topographic survey would best be carried
out when the wetland was not flooded, such as
immediately after the October shutoff date.
2. Develop a simple mathematical model to determine the
area of inundation as a function of the water level,
A(L), and the water level, L, as a function of influent
flow rates, rainfall and weather conditions, L=f(Q..,
P, El"). Experiments should be conducted to determine
A(L) and L variability with changes in water dispersal
via the discharge pipes and weirs.
73
-------
3. A network of shallow wells can be placed at dike
perimeters to investigate ground water recharge.
4. Monitor nitrogen and phosphorus on at least a bi-weekly
basis for wetland influent and effluent. Caution should
be exercised, however, when reporting 'removal
efficiencies' because dissolved materials may be carried
out via ground water infiltration — this is not the
same type of 'removal' as consumption at solid
interfaces.
Management Recommendations
It is unlikely that wetland BOO and TSS loading of Ecola
Creek through wetland discharge will exceed ODEQ limitations in the
near future. If limitations are exceeded on a consistent basis then
operators will probably need to lower mean water levels and decrease
the water turnover rate by manipulating the dispersal pattern and
weir levels to achieve maximum contact of wastewater with soils. A
last, and highly unlikely, alternative should this occur would be to
make the flow more tortuous by damming up some of the main channels
with soil material and establishing Carex obnupta or other emergent
species.
A willingness to experiment based on clearly defined
objectives, monitor changes, and incorporate observations into
management is a sound approach to managing Oregon's first wastewater
wetland treatment system.
74
-------
LITERATURE CITED
Bedinger, B. 1979. Relation between forest species and
flooding, p. 427-435, In Wetland functions and values: The
state of our understanding. Greeson, P.E., J.R. Clark, and
J.E. Clark (eds). American Water Resource Association,
Minneapolis, Minnesota.
Corps of Engineers. 1983. Permit reference no. 071-0YA-4-
004772. Portland District Army Corps of Engineers,
Regulatory Branch.
Demgen, F. 1982. Site visit report - Cannon Beach 12-13-82.
On file at the Cannon Beach wastewater treatment plant
office.
Dinges, R. 1976. A proposed integrated biological
wastewater treatment system, p. 225-230, In Biological
control of water pollution. J. Tourbier and R.W. Pierson
(eds.). Univ. of Pennnsylvania, Philadelphia.
Dooge, J. 1975. The water balance of bogs and fens,
p. 233-271, In Hydrology of marsh-ridden areas.
Proceedings of the Minsk Symposium. The Unesco Press,
Paris.
Eisenlohr, Jr., W.S. 1975. Hydrology of marshy ponds on the
Coteau Du Missouri, p. 305-311, In Hydrology of
marsh-ridden areas. Proceedings of the Minsk Symposium.
The UNESCO Press, Paris.
Gill, C.J. 1970. The flooding tolerance of woody species--A
review. Forestry Abstracts 31: 671-688.
Godfrey, P.J., E.R. Kaynor, S. Pelczarski, and J. Benforado
(eds.). 1985. Ecological considerations in wetland
treatment of municipal wastewaters. Van Nostrand Reinhold
Co., New York.
Guntenspergen, G.R. and F. Stearns. 1985. Ecological
perspectives on wetland systems, p. 69-95, In Ecological
considerations in wetland treatment of municipal
wastewaters. Godfrey, P.J. et al. (eds). Van Nostrand
Reinhold Co., New York.
Hall, F.R., R.J Rutherford, and G.L. Byers. 1972. The
influence of a New England wetland on water quantity and
quality. New Hampshire Water Resource Center Research
Report 4. Univ. of New Hampshire, Durham.
Hammer, D.E., and R.H. Kadlec. 1983. Design principles for
wetland treatment systems. U.S. Environmental Protection
Agency Report EPA-600/2-83-026. Ada, Oklahoma.
75
-------
Heimburg, K. 1984. Hydrology of north-central Florida
cypress domes, p. 72-82, In Cypress swamps. K.C. Ewel and
H.T. Odum (eds). Univ. of Florida Press, Gainsville.
Hironaka, M. 1985. Frequency approaches to monitor rangeland
vegetation, p. 84-86, In Proceedings 38th Annual Meeting of
the Society for Range Management. Salt Lake City, Utah,
February 11-15, 1985.
Hitchcock, C.L. and A. Cronquist. 1973. Flora of the Pacific
Northwest. Univ. of Washington Press, Seattle, WA.
Kadlec, R.H. 1985. Aging phenomena in wastewater wetlands,
p. 338-345, In Ecological considerations in wetland
treatment of municipal wastewaters. Godfrey, P.J. et al.
(eds.). Van Nostrand Reinhold Co., New York.
Kelly/Strazer Associates. 1983. Geotechnical investigation
and report, proposed wetlands/marsh treatment area and
related facilities, City of Cannon Beach. Kelly/Strazer
Associates, Geotechnical Consultants, Portland, OR.
Knutson, P., et al. 1981. National survey of planted salt
marshes: (vegetative stabilization and wave stress).
Wetlands 1: 129-157.
Kramer, Chin & Mayo, Inc. 1981. Facilities plan addendum no. 2:
Development & evaluation of wetlands/marsh wastewater
treatment system - City of Cannon Beach. Kramer, Chin &
Mayo, Inc. Salem, Oregon.
Linacre, E. 1976. Swamps, p. 329-347, In Vegetation and the
atmosphere, vol. 2, Case Studies. J.L. Monteith, ed.
Academic Press, London.
Mitsch, W.J. and J.G. Gosselink. 1986. Wetlands. Van Nostrand
Reinhold Co., New York.
Mitsch, W.J. and W.G. Rust. 1984. Tree growth responses to
flooding in a bottomland forest in northeastern Illinois.
Forest Science 30: 499-510.
Mueller-Dombois, 0. and H. Ellenberg. 1974. Aims and methods
of vegetation ecology. John Wiley & Sons, New York. p. 62
Nichols, D.S. 1983. Capacity of natural wetlands to remove
nutrients from wastewater. J. Water Pollution Control
Federation 55: 495-505.
76
-------
Odum, W.E., Smith III, T.J., Hoover, J.K., and C.C. Mclvor. 1984
The ecology of tidal freshwater marshes of the United
States east coast: A community profile. U.S. Fish and
Wildlife Service. FWS/OBS-87/17. Washington D.C.
Office of Technology Assessment. 1984. Wetlands: Their use
and regulation. U.S. Congress 0TA-0-206, Washington D.C.
Porter, K.G. 1979. Enhancement of algal growth and
productivity by grazing zooplankton. Science 192:
1332-1334.
Richardson, C.J. 1985. Mechanisms controlling phosphorus
retention capacity in freshwater wetlands. Science 228:
1424-1427.
Rogers, R.T. 1984. Field report: Biological monitoring plan,
Cannon Beach wetland sewage treatment facility. On file at
the Portland District Army Corps of Engineers Regulatory
Branch office, Portland, Oregon.
Smith, S.D., Bunting, S.C., and M. Hironaka. 1986. Sensitivity
of frequency plots for detecting vegetation change.
Northwest Science 60: 279-286.
Soil Conservation Service. 1982. Soil interpretation records
for the Coqui1le-Clatsop Complex and the Nehalem Silt Loam,
unpublished.
Tchobanoglous, G. and G.L. Culp. 1980. Wetland systems for
wastewater treatment: An engineering assessment, p. 13-42,
In Aquaculture Systems for wastewater treatment: An
engineering assessment. EPA 430/9-80-007, U.S. EPA,
Washington, D.C.
Thompson, D. and J. Minor. 1986. Wetlands/marsh treatment
system improves lagoon effluent quality. Presented at the
52nd Annual Pacific Northwest Pollution Control Association.
U.S. Environmental Protection Agency. 1983. Freshwater
wetlands for wastewater management Environmental impact
statement - Phase 1 report. EPA 904/9-83-107, U.S. EPA
Region IV. Atlanta, Georgia.
U.S. Environmental Protection Agency and the U.S. Fish and
Wildlife Service. 1984. The ecological impacts of
wastewater on wetlands. EPA-905/3-84-002, U.S. EPA Region
V. Chicago, Illinois.
Van der Valk, A.G. 1981. Succession in wetlands: A Gleasonian
approach. Ecology 62: 688-696.
77
-------
Verry, E.S. and D.H. Boelter. 1979. Peatland hydrology,
p. 389-402, In Wetland functions and values: The state of
our understanding. P.E. Greeson, J.R. Clark, and J.E.
Clark (eds.). American Water Resource Association,
Minneapolis, Minnesota.
Waring, R.H. and W.H. Schlesinger. 1985. Hydrology of forest
ecosystems, In Forest ecosystems: Concepts and
management. Academic Press, Inc.
Williams, J.D. and C.K. Dodd, Jr. 1979. Importance of wetlands
to endangered and threatened species, In Wetland functions
and values: The state of our understanding. P.E. Greeson,
J.R. Clark, and J.E. Clark, eds. American Water Resource
Association, Minneapolis, Minnesota.
78
-------
APPENDIX A
Tree Transect Data
7raNSech A
&fc»wi/£ff
Di*M
tLWi
Gxuws/^ff
V
a"
^afe-5
f
Ifl-I
i€xf feul
Al^u
¦/**
^1-5
It
2.
i\'}
n
u
W
\*
,3
11-\
if
HI
v}-H>
U
*/
{{,*
i>
f%
t-VZ
VI
IH^
ii
2&1
¦ .1
in
fi»«3
ii
*rt>
y>.9
VI
1
|t*3
u
57
2.1,.O
vl
n
it.Z.
7
A
<*»
ti
n*>
II
1
n,4
•v.
«u
A >
ii
£?
\V
(o
%5
A*i
ii
/5*.?
n
bf
38.6
(¦ i*i
it
fM
y
U
hZr-
Z.l.f
**
<4
n
Ib-H
ii
2.0
Z.o.1
\
V
II
bf
i ?.V
H
Zf
<•5.1
U>£H«A>4^
ii
(n*
/^7
it
(Jf-0
il
b£>
26./
it
zl:*-
ii
67
/r.4/
H
*-1
u
¥$
^V-5
Pl4.\
1Q\
II
61
2C.5"
Alt*
ZJt-
11
7^
H/'k
3J
II
i*r
•zx.£
P/5.'
-yz-
*$¦3
n
7 6
Atr*
ys
3Vf
ii
77
5^-
u
M
ZIA
ii
7f?
5f.7
\\
/*/./
ii
24- f
It
3*
n
to
Z/.?
l<
"VI
Z*.*
fiii
HI
2^-2
i\
"ft
/*.*.
hi i
T2-
2>^f
V
l(fO
Al n\
—•
Pi<, \
*f0
-21.1
a
ft
*ibO
filr»
H-l
MA
il
K
Ht.t
?
£">\
2l
ii
Ti
*lr*
<;1./
11
7-/.1
r
LoC«vho*J ®*f
Me w^s;
/ii .
NCX +
I
V'
i
See App^J'X
fof Q/Lft>dyrt
79
-------
7>1'$
n
K'6
3*>V
h
*7
2-/.2-
>i
it.(
't
VL.T)
hs>
iv
V\-(o
Alfn
Z2..-L
£
icSl
M0+C.5
80
-------
tfofe.5
PlSt
I k.-z-
rffcrt -to 101
*
|W.i
A
n
N-2.
H
V
ii.*.
n
5"
If, 7
Q
Cl/
n
h
Z-Jf
ii
n
M
h
W
P
*
lis
<
u,
Altu
u
Pin
t&
'if?
/ .
AlAu
//
2./-3
Vv
j
AlfH
/£
0)
CO
ft si
/s
/$.*{
V\J|
n
t±
/P-7
*-> ,
ii
IS
/I-"1-
ii
lb
ii
'7
/Z.Z
AIM
/a
Mru
/?
3«».2-
/
i>
—
LT3KZ2JI
n
fifa
«-
*
>i
/tc
— i
h
U?
/5.<»
il
M
2.^5
M
Z"Z-
f2 W
il
2*4
—
—
2S
—'
CArt't -f-itid
Altn\
21/
/<>-3
ii
z
?>0.c
ft-fM
V»
10-4
rflrn
-M-
3V<
^Alru
3
Al'ru
V2.2-
AUm
11
^o.c?
5®
%•»
Pi6i
5/
11.'
AI/m
5*
1*1.?
fVP*
1-2,-0
dt*
¦Z.-Z.O
rui
5T
U.O
Air*
!5-c>
Airti
^"7
«.3
\I
5*
25 Z-
U
5"?
£*>
—-¦
bl
/iT,*
fa
>/•?
Pfsi
A l^n
a
t>4
it
ib 2.
it
bi
^7-Z.
de*^
M
b$
deaA
n
frl
2a.|
ii
7o
z-*7
Pk/
7/
2.|.?
AIoh
1-z-
Z2.J?
AIM
7,3
zz-V
Pi'ii
nH
tar'+fikk
~Ar*
HA
i\
fiP,
•v/s
H«*rt
M
ftc-
M
4v
/ ( 5-5-
U
A£
/tf
J
81
-------
Tr4H$*c+ /-"
V
pi<\n
ten)
Mcrfc.5
Piy;
1
/f'-T
/C7#
n
•Z
/7-?
v
*/
3
Z4f./
—5~~
n
V
"]§
4'
H
HP'I
//
I
H-1
a'
Aim
n
//•2,
V
>s- .
.4
Alrn
k
3?.£
^ >
f
Mh
1
—
w
lA\d
!>yfti
10
AlftA
n
/5"-*
\
/
A!**
fl~
'7-7
AlfiH
1}
/z.z
'7
z.*. 1
mn
If
16-to
phi
ic.f,
Alru
Ik
/M
ii
n
»i
m
/f-Y
a
il
u>.e
ii
***-
,3*
tc.i
flii
H-1
Air*
"i5"
3?
sX'
3£
P$.c
wn
31
ziw
3*
ii
34
<1
<^".3
fi<,i
11
m-v.
film
Hi
ik-i
A'na
W3
GtMVS/^ff
?5
Pi<\H
a. ft)
Mofc^
¦ 4^M
&>•$
n^H
45"
inn.
Jl^'A
%
y>.'\
ii
V7
ii
Z0.1
n
V?
2fv1
fm
it. i
tfU
.57
Ib.Z.
n
K2n
3®.?
u
s\
l\
ii
(pO
y>'*>
i1
*!
CAt'ifiuA
ii
fa
to. 3
f.si
Alt' 14
*¥
II
6S
^"*2-
M
tn£>
II
hi
11
ti
k.7-t
h
£>9
rt
W
if
If
%•/
n
V•
M.3
fi
h\
>i
Vi
3*-,7
ii
Is
ii
76
It.O
t\
m
So-1
M
It
U
7?
MeK+ f"r
eaifd'ikf.
82
-------
APPENDIX B
Nested Frequency Method Notes
A partial field data sheet is provided to help the next survey crew.
Example calculations are also shown below. Refer to the 1986 Monitoring
Methods section — Methods and Materials (p.33) — for details concerning
field methods.
I nr/U-.oM; A
w«r
\
z_
s
t
5"
y
1
s
<>PP
A
hA
A
A
IA
p
X
oesa
-1
-2.
+ i
-I
or 5
o+l
-ri
Po««
9-Z.
o+l
+7-
A+l;
o-Z
- 1
+ 1
-1
<
»
4
P
Q
0
0
O
| .0
o
O
P
KEY
Rooted within quadrat. 1n muck:
+ presence 1rt 50cm x 50cm plot, presence In In x I n plot
- absence 1n 50cm x 50cm plot, presence 1n lm x lm plot
Rooted within quadrat, on log or upland hummock:
0+ presence in 50cm x 50cm plot
0- absence in 50cm x 50cm plot, presence 1n lm2 plot
AQ1 = Aquatic Index:
A - aquatic, 50% plot inundated
IA - intermediate, 25-50% plot Inundated
NA - upland or nonaquatic, 25% plot Inundated
Numbers following +, -, or 0+, 0- are cover classes:
1 = <5%
2 = 5%< x < 25%
3 = 25% < x < 50%
4 = 50%< x < 75%
5 = >75%
h = depth of water relative to ground surface
Example frequency calculation:
Oesa: frequency in ,25m2 = 2/9 x 100 - 22%
frequency 1n 1.0m2 = 5/9 x 100 = 55%
Example aquatic index/spp frequency calculation:
Atf1: NA
IA
A
2/5 x 100 = 40%
2/5 x 100 - 40%
1/5 x 100 = 20%
83
-------
APPENDIX C
Species acronym list for the Cannon Beach wastewater wetland, 1986.
Taxonomy is based on Hitchcock and Cronquist (1973).
Scientific Name
Common Name
Acci
Acer circinatum
vine maple
Adpe
Adiantum pedatum
maidenhair fern
Alru
Alnus rubra
red alder
Atfi
Athvrium filix-femina
lady fern
Blsp
Blechnum spicant
deer fern
Cal sp.
Calichtriche sp.
water-starwort
Cade
Carex dewevana
sedge
Caob
Carex obnupta
slough sedge
Car sp.
Cardamine sd.
bitter cress
Cial
Circaea alpina
enchanter's nightshade
Ep sp.
Epilobium sp.
willow weed
Erar
Erechtites arquta
fi reweed
Gash
Gaultheria shallon
salal
Gatr
Gallium trifidum
bedstraw
Hy sp.
Hvdrophvllum sp.
waterleaf
Lemi
Lemna minor
duckweed
Loin
Lonicera involucrata
twinberry
Lyam
Lvsichitum americanum
skunk cabbage
Madi
Maianthemum dilatatum
false lily-of-the-valley
Mefe
Menziesia ferruqinea
fool1s huckleberry
Mo sp.
Montia sibirica
candy flower
Oesa
Oenanthe sarmentosa
water parsley
Chgl
Chrvsosplenium qlechomaefolium
golden carpet
Oxor
Oxalis oreqanum
wood sorrel
Tome
Tolmiea menziesii
youth-on-age
Pisi
Picea sitchensis
Sitka spruce
Povu
Polvpodium vulqare
licorice fern
Pomu
Polvstichum munitum
sword fern
Pyfu
Pyrus fusca
wild crabapple
Rumex sp.
Rumex sp.
dock
Rupa
Rubus parviflorus
thimble berry
Rusp
Rubus spectabilis
salmon berry
Ruur
Rubus ursinus
pacific blackberry
Saca
Sambucus callicarpa
red elderberry
Seja
Senecio iacobaea
groundsel
Stco
Stachvs coolevae
hedge nettle
Tiun
Tiarella unifoliata
coolwort
Vapa
Vaccinium parvifolium
huckleberry
Vigi
Vicia qiqantea
giant vetch
84
-------
APPENDIX C
Species located within the wastewater wetland boundaries but not occuring in
the nested frequency/permanent plot sampling survey:
Semi Scirpus microcarpus small-fruit bull rush
Juef Juncus effusus soft rush
Tyla Typha latifolia cattail
Ve sp. Veronica sp. speedwell
Poam Polygonum amphibium water smartweed
Pohy Polygonum hydropiper smartweed
85
------- |