Applied Wetlands Science - Chapter 5 ppsx

Kent, Donald M. et al “Avoiding and Minimizing Impacts to Wetlands”
Applied Wetlands Science and Technology
Editor Donald M. Kent
Boca Raton: CRC Press LLC,2001

©2001 CRC Press LLC

CHAPTER

5
Avoiding and Minimizing Impacts to
Wetlands

Donald M. Kent and Kevin McManus

CONTENTS

Planning
Design and Construction
Design
Construction
Erosion and Sedimentation
Nitrogen Loading
Planning Guidelines
Estimating Nitrogen Loads
Stormwater Runoff
Planning and Nonstructural Practices
Structural BMPs
Pretreatment
Detention Basins/Retention Ponds
Vegetated Treatment

Infiltration
Filtration
References
Recent estimates of the extent of global wetlands range from 5 to 8.6 million ha
(Mitsch, 1995). Increasing evidence suggests that the historic extent of global wet-
lands was substantially greater. For example, in Japan, 45 percent of tidal flats have
been destroyed since 1945 (Hollis and Bedding, 1994). Northern Greece has lost

©2001 CRC Press LLC

94 percent of its marshland since 1930. In the conterminous United States, an
estimated 47 million ha of wetlands have been lost over the last 200 years — an
average rate of 235,000 ha per year (U.S. Office of Technology Assessment, 1984;
Dahl, 1990; Hollis and Bedding, 1994). This rate of loss appears to have decreased
dramatically in recent years, to about 32,000 ha per year, coincident with recognition
of the importance of wetlands and a “no net loss” government policy (Heimlich and
Melanson, 1995). Wetland losses are attributed to filling and draining, primarily in
support of development and agricultural activities.
An unknown number of wetlands, not filled or drained, have been otherwise
impacted by changes in watersheds or adjacent land uses. Alterations to wetland
plant communities lead to increased erosion and sedimentation. Construction of
buildings, parking lots, and other impervious surfaces increases the quantity and
decreases the quality of surface runoff to wetlands. Septic systems and fertilizers
increase the concentration of nitrogen in groundwater flow to wetlands. Activities
adjacent to wetlands can disturb wildlife.
Wetland impacts, both direct and indirect, can be avoided or minimized by
appropriate planning, design, and construction. In this chapter, planning is discussed
as a means for avoiding or minimizing direct impacts to wetlands. Design and
construction techniques are discussed as a means to avoid or minimize indirect
impacts to wetlands. Discussed in some detail are three design and construction

issues. They are erosion and sedimentation, nitrogen loading, and stormwater.

PLANNING

Planning to avoid or minimize direct impacts to wetlands is fundamentally a
three-step process. The first step is to identify the wetland resource. Discussed in
detail in Chapter 2, this step requires applying hydrology, soils, and vegetation
criteria to undeveloped areas. For large areas, off-site resource identification is an
effective and appropriate approach for preliminary planning. Greater resource res-
olution, typically requiring on-site identification, is more appropriate for smaller
areas and for detailed planning. Characterization and classification (e.g., palustrine
forested wetland, emergent marsh; see Chapter 1) of wetland resources are also
helpful at this stage.
The second step in effective planning is to assign functions and values to iden-
tified wetland resources. Common techniques for determining functions and values
include professional opinion, the use of indicators, direct measurement, and eco-
nomic analysis (see Chapter 3). As with resource identification, off-site and less
detailed approaches are most appropriate for large areas during preliminary planning,
whereas on-site assessments are most appropriate for small areas and detailed plan-
ning. Assigning functions and values will facilitate prioritization in the event that
not all resource areas can be preserved and reveal functions and values that need to
be protected or replaced during construction and operation.
Finally, wetlands identified and evaluated for functions and values are incorpo-
rated into a site selection process. Site selection typically includes identification of

©2001 CRC Press LLC

several alternative sites and development of site selection criteria. Alternative sites
satisfy minimal, implicit criteria such as availability and location.
At a minimum, the site selection process should consider the criteria listed in

Table 1 (McManus, 1994). Direct and indirect impacts to wetland and other envi-
ronmental resources should be identified. Other environmental resources include fish
and wildlife, navigation channels, and recreation areas. Projects not dependent upon
access to water should be sited elsewhere. The minimum size required to satisfy the
project purpose should be determined and project configuration and layout evaluated
to further reduce project size. Constructability refers to project topographic, slope,
soil, and backfill requirements. Extensive grading, blasting, or filling are typically
associated with environmental impacts and should be avoided. Proximity to support-
ing infrastructure, such as utilities and roadways, affects project size, configuration
and layout, and cost. Cost prohibitive sites should be eliminated; thereafter, the costs
of development should be weighed against the costs of environmental impacts. The
opportunity for successfully satisfying the requirements of various international,
national, regional, and local entities such as regulatory agencies and lending insti-
tutions should also be evaluated.
Larger and more complex projects will require a more detailed site selection
process. In the United States, the National Environmental Policy Act (U.S. Con-
gress/NEPA, 1978) provides guidance as to appropriate criteria for evaluating project
impacts to wetlands and other environmental resources. In addition to environmental
impacts, this approach considers impacts to human uses and the technical, economic,
and institutional feasibility and merits of the site. Table 2 represents a hypothetical
site-screening matrix consistent with the NEPA (McManus, 1994). In the example,
Site 1 is technically and economically feasible, but will likely impact the environment
and human use of the site, and is not publicly acceptable. Site 2 has no significant
environmental, human use, or institutional constraints but has technical and eco-
nomical issues. Site 3 is the preferred site, having no significant environmental or
human use impacts, being technically and economically feasible and acceptable to
the public.

Table 1 Representative Site
Selection Criteria

(Adapted from

McManus, 1994)

Wetland impacts
Other environmental impacts
Water dependency
Site size
Constructability
Supporting infrastructure
Costs
Regulatory/institutional issues

©2001 CRC Press LLC

Table 2 A Hypothetical Site Selection Matrix (Adapted from

McManus, 1994)
Screening Criteria Site 1 Site 2 Site 3

Environmental

Aquatic Ecosystem
Substrate 0 0 0
Water quality 0 0 0
Water circulation 0 0 0
Normal water fluctuations – 0 0
Threatened and endangered species – 0 0
Other aquatic organisms and wildlife – 0 0
Special Aquatic Sites

Sanctuaries/refuges – 0 0
Wetlands – 0 0
Mudflats 0 0 0
Vegetated shallows 0 0 0
Riffle and pool complexes – 0 0

Human Uses

Water supplies 0 0 0
Recreational and commercial fisheries – 0 0
Water–related recreation – 0 0
Aesthetics – 0 0
Parks, preserves, wilderness areas – 0 0
Archaeological or historical sites 0 0 0
Compatibility with adjacent land uses – 0 0
Potential noise impacts 0 0 0
Potential odor impacts 0 0 0
Public health 0 0 0
Traffic increase 0 0 0

Technical

Suitable foundation/soils conditions + – +
Adequate land area + – +
Access to existing roads and utilities – +

Economic

Land acquisition + – +
Operation and maintenance + – +

Capital cost—construction 0 – 0

Institutional

Public acceptance – 0 +
Compliance with existing regulations 0 0 0

Note:

+ indicates an expected positive impact; – is an expected negative
impact; 0 is an insignificant or no impact.

©2001 CRC Press LLC

DESIGN AND CONSTRUCTION
Design

Once the site selection process has been completed, the focus can shift to design
details, site layouts, construction methods, and other specific engineering require-
ments to minimize unavoidable wetland impacts. A reasoned assessment of the
minimum economically and functionally viable size for a proposed structure(s)
should be made, particularly if the project is not water dependent. Even for water
dependent projects, such as marinas or dredging projects, project scope should be
evaluated with an eye toward minimizing wetland impacts. The project should have
an accurate wetland delineation line depicted on site plans to facilitate evaluation
of layout options.
For projects that may involve clearing of trees and other existing vegetation, care
should be taken to minimize the limits of clearing to the minimum acreage needed
for the project. Maintenance of existing vegetative buffers, particularly within wet-
land areas, is not only a valuable means of providing a visual and auditory buffer

for the facility, but it also may reduce overall facility wetland impacts. This is
particularly true along active coastal shorelines, such as eroding bluffs, beaches, and
dune environments.
The orientation and layout of a project are generally a function of its intended
purpose and use. Many projects, such as railways, roads, and retaining walls, being
linear features, have limited flexibility with regard to basic configuration. However,
their actual alignment, relative to wetland areas, can often be optimized to reduce
impacts to insignificant levels. Similarly, layouts of buildings and ancillary struc-
tures such as garages, walkways, and decks can be adjusted to minimize direct
wetland impacts.
Specific design details for a project can also be important factors in reducing
wetland impacts. For example, use of the maximum safe slopes for site preparation
will minimize incursions into wetland areas. Maximum safe slopes can be achieved
using vertical retaining walls, cellular confinement, sheet piling, or gabion rock
walls. Backfill and other construction materials should ensure good drainage and
scour protection (Nelson, 1995). Another method for minimizing impacts is to use
boardwalks supported by posts or post-like anchors.
Waterway crossings offer another opportunity to minimize wetland impacts.
Typically, culverts are used when crossing small waterways. Culverts should be
designed to pass expected flows (e.g., 100-year flood event), and to avoid changes
to flow velocity and increased erosion and scour. Bridges can minimize impacts to
larger waterways, especially if construction is accomplished in midair using a
crawler crane.

Construction

For many projects, such as subsurface water, sewage, and other utility pipelines,
the primary impacts to wetlands occur during construction. The use of temporary

©2001 CRC Press LLC

access materials, specialized construction equipment, and the placement of staging
areas can all affect the level of wetland impacts.
Temporary pile-supported construction trestles can be used to significantly
reduce direct wetland impacts through ecologically sensitive wetland areas such as
estuarine and fresh-water marshes, beach or dune environments, and peat bogs
(Figure 1). These trestles can be located either directly above, or directly adjacent
to, the work area. Equipment can be brought to the work area using rail-mounted
transport platforms, and the trestle can be constructed in stages to accommodate the
construction schedule. Trestles provide a stable temporary work platform that
directly impacts little wetland acreage.

Figure 1

Temporary pile-supported construction trestles can be used to significantly reduce
wetland impacts. The trestles may be located either directly above or immediately
adjacent to the work area.
PLAN VIEW
pipeline
wood planking
sheet piling
construction trench
sheet piling
wood planking
pipeline
wetland
PROFILE
upland

©2001 CRC Press LLC

Another effective construction technique uses steel sheet piling to isolate the
active work area, and temporary wood decking placed directly on top of the sheet
piling. This allows construction equipment to access the work areas without com-
pacting wetland soils. Compacted wetland soils lose their original productivity and
hydrologic functions. For smaller projects that may not warrant the use of sheet
piling, geotextile fabric, clean granular material, and wood decking can be placed
within the project alignment.
Sheet piling can also be used in intertidal or shallow freshwater areas. Combined
with siltation curtains, piling can prevent the release of sediment-laden water to
surrounding wetlands and waterways. The use of barge-mounted equipment can also
be used in intertidal and shallow freshwater areas to access sensitive sites. Work
barges can be floated into place on rising tides, and grounded out to provide suitable
access with minimal or no long-term impacts.
For construction of trenches in wetland areas, utility workers have developed
specialized, tracked, trenching vehicles that can operate on soft, unstable soils. The
vehicles work directly within the project alignment. Wide, low-pressure tires on
vehicles that distribute loads across wetland soils and vegetation also reduce vehicle
impacts. For dredging within wetlands, waterways, and waterbodies, clamshell
dredge equipment fitted with covers and watertight buckets minimizes sediment
washout and turbidity.
Large construction projects typically require staging areas. Staging areas should
be located outside wetlands and their designated buffer zones and should be paved
to minimize erosion and groundwater impacts. Also, staging areas should include
stormwater management systems designed to trap suspended sediments and to con-
tain accidental releases of fuel oil, lubricants, and other potentially hazardous
releases from equipment.
Scheduling can minimize temporary, construction-related impacts. As a general
rule, wetland work in temperate climates should be scheduled during winter and
early spring when plants are dormant and the soils are frozen or well consolidated.

Soil compaction is minimized, and site cleanup and rehabilitation during the coming
peak growing season are facilitated. Other seasonal restrictions are often applied for
work within coastal environments based upon the expected presence of commercially
and recreationally important fish and wildlife species. Species susceptible to ill-
timed construction include spawning and migrating anadromous fish and shrimp,
overwintering groundfish, and migratory waterfowl.
Another method for minimizing the impacts of construction within wetlands is
proper work sequencing. For example, minimizing the extent of clearing in front of
the active trenching operation will reduce the potential for soil erosion into adjacent
wetlands and reduce impacts to wildlife using the existing vegetative cover. Wherever
possible, work that is required within wetland areas should be completed as quickly
as possible, without excessive delays between the initial disturbance and rehabilita-
tion. Trenching should be conducted as a single, continuous operation, involving
clearing, installation, backfilling, and soil restoration. An open trench can act as a
channel to dewater adjacent wetland areas and increase erosion and runoff impacts.

©2001 CRC Press LLC

EROSION AND SEDIMENTATION

Sedimentation of wetlands can be avoided or minimized by preventing soil
erosion and controlling already eroded sediments. There are numerous methods for
erosion and sedimentation control, all of which seek to isolate and contain, to the
maximum extent possible, sediment-laden runoff generated during project construc-
tion activities. The performance of these various methods in the field varies consid-
erably depending upon the type of soils, water flows, exposure, and other site specific
factors. Figure 2 summarizes some of the more popular sedimentation control meth-
ods. Critical elements of effective erosion and sediment control plans are listed in
Table 3 (Brown and Caraco, 1997).
Erosion and sedimentation control methods can be used singly or in combination.

By limiting the amount of incremental and total land clearing, and maintaining
existing ground cover to the maximum extent possible, potential runoff, gully cre-
ation, rutting, and airborne dust formation can be reduced to acceptable levels.
Cleared land produces as much as 2000 times more sediment than uncleared land
(Paterson et al., 1993). Where feasible, a project site layout should take advantage
of existing vegetation between the clearing limits and adjacent wetlands. Buffers of
at least 25 m in width are the most effective in filtering sediment from construction
site runoff (Woodward, 1989). Vegetative buffers should also be preserved for
projects with shoreline frontage to protect structures from wave and flooding impacts.
Installation of hay bales within shallow cut-off trenches upgradient of wetland
areas can be an effective and inexpensive perimeter control method. Bales should
be staked to the ground, without gaps between bales. Bales should be routinely
monitored, and bales damaged, moved, or destroyed during construction should be
repaired. Construction specifications should provide for regular checks of the con-
dition and effectiveness of the hay bale protection systems. Geotextile siltation fences
can be wrapped around hay bales and staked into the ground to provide an extra
measure of protection against the release of fine-grained materials. Siltation fence
efficiency ranges from 35 to 86 percent depending upon site conditions (Horner
et al., 1990; W&H Pacific and CH2M-Hill, 1993).
Siltation curtains can also be used effectively in both wetlands and open water
environments. Curtains can be used to surround subaqueous dredging operations,
particularly those occurring within sheet piling, to isolate trench water from the
surrounding environment. Curtains with flotation can also be installed around shore-
line construction projects and anchored in place to isolate the work area. However,
the effectiveness of these structures decreases significantly in areas of strong river
currents, tidal flows, and large tidal ranges, particularly if the curtain is installed
perpendicular to the current flow. In such cases, the siltation curtain experiences
rollover or submergence and is susceptible to damage from debris. Therefore, silt-
ation curtains are most effective in ponds, lakes, and other sheltered water bodies
with little or no variation in water height.

In any construction project, regardless of the proximity to wetlands or other
adjacent sensitive habitats, construction specifications should require prompt stabi-
lization of newly exposed soils, including stockpiled soil. Seeding and sodding are
relatively inexpensive, and up to 99 percent effective in reducing erosion (Brown

Figure 2

Erosion and sedimentation control methods (McManus, 1994). Black indicates the method is suitable for use in the
environment; gray indicates the method is suitable with limitations.
©2001 CRC Press LLC

©2001 CRC Press LLC

and Caraco, 1997). Seeding is the least expensive option and is appropriate when
temporary stabilization is required. Seeds can be broadcast by hand or hydroseeded.
The latter is a mixture of seeds, water, fertilizer, lime, and mulch sprayed onto the
soil. Sodding is more appropriate for permanently vegetated areas and provides
immediate cover and greater resistance to higher flow velocities.
The construction schedule should allow time for vegetation to become reestab-
lished prior to the end of the growing season. In cases where this is not possible,
more expensive but generally less effective measures, such as mulching or covering
exposed areas with erosion control blankets, jute mats, or geotextile mats, should
be employed. Mulches, blankets, and mats protect seeds from erosion, dehydration,
and animals until the next growing season (Brown and Caraco, 1997). Mulches,
consisting of straw, hay, fiber, or wood chips, are effective on flat or gently sloping
areas. Erosion control blankets consist of a mulch material held together by a plastic
netting, and jute mats are sheets of woven jute fiber. Effective on relatively level
ground, both the blankets and the mats are stapled to the ground after seeding and
degrade over time. Geotextile mats are more appropriate for steeper slopes and
channels. The mats are typically laid on the soil surface and covered with topsoil

and seed.
As previously discussed, isolation of the active work area in both wetlands and
open water areas is an effective method to limit the horizontal extent of disturbance,
particularly in areas where significant dredging is required. In such cases, dredging
open trenches beyond 1 m in depth requires side slopes which can range from 3:1
to 5:1 or greater, meaning that a 3-m-deep trench would disturb a minimum 20- to
33-m width of sediments. Clearly, this size dredging operation would require the
handling and disposal of large amounts of excess dredged material. Conducting this
work within sheet pilings allows a vertical sidewall, thereby reducing the volume
of material to be handled and isolating the silt-laden trench water from surrounding
marsh and other wetland areas.
Open dredging within or adjacent to wetland areas can produce significant
amounts of turbidity. If typical dredging equipment is used, for example, a barge-
mounted crane with a clamshell dredge bucket, methods are available which can
reduce turbidity. These include establishing requirements that all lifts of a clamshell
dredge bucket through the water column are vertical, that dredge buckets be used

Table 3 Critical Elements of an Erosion and
Sediment Control Plan (Adapted from

Brown and Caraco, 1997)

Minimize clearing and grading
Protect waterways and stabilize drainage ways
Phase construction to limit soil exposure
Stabilize exposed soils immediately
Protect steep slopes and cuts
Install site perimeter control to filter sediments
Use settlement traps and basins for larger volumes
Use experienced contractors to implement the plan

Tailor the plan to specific site conditions
Assess plan effectiveness after storms

©2001 CRC Press LLC

with covers and gasket seals to prevent washout of sediments and suitable filtering
of water released from stockpiled dredged material.
Hydraulic dredging can also be used in certain unconsolidated sediments to
reduce turbidity. With this method, sediments are removed and pumped as a slurry
to a settling barge or disposal site. While initial turbidity at the point of dredging is
minimal, large amounts of water must be filtered and removed from sediments at
the disposal site, and pumping limitations require that disposal occur in close prox-
imity to the point of dredging.
Construction will often require temporary stockpiling of soils, and care should
be taken to continually spray these piles with water, or cover them, in order to
prevent wind erosion and transport of fines. Similarly, newly graded access roads
should be frequently sprayed with water or dust suppressants to reduce dust forma-
tion. The construction schedule should attempt to minimize the period of time where
exposed stockpiles or unpaved road surfaces are required.
Site grading and excavation activities in areas already served by drainage systems
are a potential concern for sedimentation. Many parking lots, roadways, and other
facilities use stormwater drainage systems that discharge directly into adjacent
wetland areas. In order to minimize the impacts from run-off of sediment-laden
water, existing catch basins and storm drains should be completely ringed with staked
haybales and a layer of filter fabric. Other inlet protection methods include concrete
block wrapped with wire and stones and placing geotextile fabric and stones directly
over the inlet (Brown and Caraco, 1997). These sediment traps will allow stormwater
flow to pass through, but will filter out significant amounts of suspended sediments.
These structures also provide protection in the event of an accidental fuel oil spill,
hydraulic hose rupture, or other hazardous material release, providing some measure

of initial containment upgradient of adjacent wetland areas. As with all hay bale
structures, the sediment traps need to be maintained and periodically replaced to
ensure their effectiveness.
Excavation for foundations, utility trenches, and other facilities will often extend
below the existing water table, resulting in collection of groundwater within the
excavation. In order to dewater these areas and prevent discharge of sediment-laden
water into surrounding areas, various types of settling basins and detention structures
can be constructed. Sediment removal efficiencies generally range from 60 to 90
percent, with higher efficiencies associated with wet storage (Brown and Caraco,
1997). These structures allow particulate matter to settle and gradually discharge
filtered runoff. Figure 3 is a schematic representation of a typical settling basin which
can be constructed upgradient of a wetland area using filter fabric and clean rip-rap
material to effectively filter silts and sediments at a construction site. Concrete or
fiberglass settling basins are also available for use as sedimentation control structures
during dewatering operations and are often used on barges during dredging opera-
tions to filter water discharged from stockpiled dredged materials. Geotextile wetland
filter bags have also been developed to serve as sedimentation and erosion control
devices on construction sites (Figure 4).
The true test of any sedimentation and erosion control plan will occur during
the first significant rainfall event during construction. Thus, it is recommended that
on-site resident inspectors monitor the success of the installed erosion control devices

©2001 CRC Press LLC

during and immediately after a rainstorm or snowmelt. The hay bales, siltation
fences, and other structures should be observed on, at least, a weekly basis to detect
damage from wildlife, machinery, or other activities on site.
Equally important, resident inspectors should conduct frequent visual observa-
tions of the adjacent wetlands or open water bodies to detect turbidity plumes
resulting from on-site runoff. For certain subaqueous activities, significant short-

term increases in turbidity are unavoidable. Nevertheless, attention should focus
on the effectiveness of the siltation curtains, dredging methods, and dewatering

Figure 3

Settling basins are used in conjunction with dewatering operations to prevent
discharge of sediment-laden water into wetlands. The basins are constructed
upgradient of wetlands using filter fabric and clean rip-rap material.
Ground Slope
Sediment Laden Water
Pump Discharge
Flat Stone
Approved Filter fabric Mat
10'-15' (Typ.) or as Direcled
15' - 20 (Typ.)
or as Directed
To Natural
Water Course
Sediment Free Water
Ground Slope
Baled Hay or Straw
Pump Discharge Line
Flat Stone
Approved Filter Fabric Mat
Sediment
Clean Stones
(If Required)
Suitable Velocity Dissipator
Baled Hay or Straw
as Directed

TYPICAL SECTION
SEDIMENT TRAP
Suitable Device
to Dissipate Velocity

Figure 4

A geotextile bag can be used on construction sites to remove sediments from site runoff.
Wetland Filter Bag
Water flowing
out from the bag
Pipe
Flow
©2001 CRC Press LLC

©2001 CRC Press LLC

practices to ensure that surrounding background levels of turbidity are not signif-
icantly increased. For deepwater areas, the use of a Secchi disk or similar device
will provide a qualitative measure of the water clarity and amount of suspended
sediments during construction.

NITROGEN LOADING

Nitrogen occurs in wetlands in various inorganic and organic forms (Mitsch and
Knight, 1997). Ammonia, nitrate, and nitrite are the most important forms for
wetland processes. Ammonia is an important nutrient for most wetland plants and
autotrophic bacteria and is a growth limiting compound in coastal waters. Coastal
waters are the most highly fertilized ecosystems on earth (Nixon, 1986; Kelly and
Levin, 1986). In natural waters, ammonia is readily oxidized resulting in oxygen

consumption. Ammonia in its unionized form (NH

3

) is toxic to many forms of aquatic
life at low concentrations (0.2 ppm).
Nitrate is reduced to nitrite in oxygen-poor environments and is conservative in
groundwater. In infants, nitrite combines with fetal hemoglobin preventing oxygen
transport. This potentially fatal condition is known as methemoglobinemia. High
nitrate concentrations have also been linked to carcinogenic effects (U.S. Environ-
mental Protection Agency, 1990).
The primary source of nitrogen, particularly in residential areas, is domestic
wastewater (Cape Cod Commission Water Resources Office, 1992; Valiela et al.,
1997). For residential septic systems, the concentration of nitrogen depends upon
soil characteristics, loading rates, distance to the impervious stratum, distance to the
water table, time of year, and depth below the leach field (Suffolk County Department
of Health Services, 1983; Canter and Knox, 1985). Typical nitrate nitrogen concen-
trations range from 33 to 41 ppm (Cape Cod Planning and Economic Development
Commission and U.S. Environmental Protection Agency, 1978; Nassau-Suffolk
Regional Planning Board, 1978; Suffolk County Department of Health Services,
1983; IEP, 1988; Robertson et al., 1991). Nitrogen concentrations in nonresidential
areas are less well known and vary widely in character and quantity. In general,
nitrate nitrogen concentrations are higher when no gray water (i.e., sinks and show-
ers) is present.
Secondary sources of nitrogen include lawn fertilizer, atmospheric nitrogen,
and runoff from impervious surfaces. Nitrogen levels from lawn fertilizer vary
with soil type, application rate, precipitation, temperature, turf type, and nitrogen
form. Typical fertilizer application rates range from 0.8 to 1.7 kg/100 m

2

(1.7 to
3.8 lb nitrogen/1000 ft

2

) up to 4.7 kg/100 m

2

(9.6 lb nitrogen/1000 ft

2

) for golf
course greens (Nassau-Suffolk Regional Planning Board, 1978; Cape Cod Planning
and Economic Development Commission, 1979; Eichner and Cambareri, 1990).
Nitrogen leaching rates vary widely, from 0 to 60 percent (Nassau-Suffolk
Regional Planning Board, 1978; Brown et al., 1982; IEP, 1988; Petrovic, 1990).
Controlled application of fertilizer to healthy turf can eliminate or minimize
leaching (Petrovic, 1990).

©2001 CRC Press LLC

Atmospheric nitrogen loading, largely from precipitation, is relatively minor
compared to potential loading from wastewater and fertilizers. Precipitation concen-
trations in the United States range from 0.14 to 1.15 ppm nitrate nitrogen (Loehr,
1974). Dry deposition of nitrogen may double this concentration (Valiela et al.,
1997). Nitrogen loading off of impervious surfaces is, however, significant, ranging
from 0.41 to 1.75 ppm nitrate nitrogen and 1.13 to 10 ppm total nitrogen (IEP, 1988).

Recharge rate off of impervious surfaces is poorly understood. The TR-55 storm-
water modeling program assumes a recharge rate of 98 percent (Soil Conservation
Service, 1986) and the Water Resources Office of the Cape Cod Commission (1992)
assumes a 90 percent recharge rate.

Planning Guidelines

Based upon the threat of methemoglobinemia and cancer, the U.S. Environmental
Protection Agency has established a limit of 10 ppm nitrate nitrogen in drinking
water (U.S. Environmental Protection Agency, 1990). Studies on Long Island, NY,
revealed that average nitrate nitrogen concentrations of 6 ppm led to violation of
the 10 ppm criteria 10 percent of the time, and that average concentrations of 3 ppm
led to violation of the 10 ppm criteria 1 percent of the time (Nassau-Suffolk Regional
Planning Board, 1978; Long Island Regional Planning Board, 1986). Based upon
this information, Long Island recommended that areas be sewered if the average
nitrate nitrogen concentration exceeds 6 ppm. The Cape Cod Planning and Economic
Development Commission (1978) and Cape Cod Commission Water Resource Office
(1992) adopted a 5 ppm nitrate nitrogen standard.
A second consideration in the establishment of nitrogen standards is protection
of coastal embayments. Each embayment has a unique critical nitrogen loading rate
dependent upon embayment morphology and flushing rate. For example, U.S. Envi-
ronmental Protection Agency and the Massachusetts Executive Office of Environ-
mental Affairs (1991) developed recommended nitrogen loading limits for Buzzards
Bay (Table 4). Recommended nitrogen loads are lower in shallower embayments
than deep embayments and in higher quality waters than lower quality waters. Based
upon studies of Waquoit Bay, MA, Valiela et al. (1997) offer general recommenda-
tions that wastewater disposal within 200 m of shore be limited, that homes be
required to use multiple leaching fields or septic systems, and that fertilizer use be
controlled on near-shore lawns.

Estimating Nitrogen Loads

Several methods have been used to estimate nitrogen loading to groundwater and
coastal embayments. For example, the Cape Cod Commission Water Resources Office
(1992) recommends a site-specific mass balance analysis for relatively small sources
and a cumulative loading analysis for proposed sources in groundwater recharge zones
and relatively large sources. The former estimates nitrogen and water uses within the
boundaries of a development, whereas the latter is a recharge zone-wide analysis for

©2001 CRC Press LLC

existing and proposed conditions. The models estimate the nitrate nitrogen load by
totaling the nitrogen inputs from wastewater, impervious surfaces (roof and paved),
and fertilizer, and dividing nitrogen inputs by total water inputs. Figure 5 illustrates
a mass balance analysis process for a hypothetical 20 house residential development.
In this example, as typically occurs, the majority of nitrate nitrogen originates in
wastewater.
Valiela et al. (1997) developed a model to estimate atmospheric, fertilizer, and
wastewater nitrogen loading to watersheds and receiving waters. Based upon data
from the Waquoit Bay Land Margin Ecosystems Research Project and syntheses
of published information, the model estimates nitrogen inputs to surfaces of the
major types of land use within the landscape. Nitrogen losses in the various water-
shed compartments are then estimated. For example, atmospheric and fertilizer
nitrogen are lost in vegetation, soils, the vadose zone, and aquifer. Wastewater
nitrogen losses occur in septic systems and effluent plumes and during diffuse
transport in aquifers. Nitrogen loss calculations are conducted separately for each
major type of land cover. The model was developed for Waquoit Bay, MA, but is
believed applicable to other rural to suburban watersheds underlain by unconsoli-
dated sandy sediments.
According to the model, the atmosphere is the largest contributor of nitrogen to

the watershed, but wastewater is the largest source of nitrogen to receiving estuaries
(Table 5). The model implies that estuary management should focus on wastewater
disposal, particularly within 200 m of shore. The authors also suggest that installation
of multiple conventional leaching fields or septic systems in high flow parcels could
be beneficial. Other recommendations include control of fertilizer use on near-shore
lawns and conservation of parcels of accreting natural vegetation. The latter effec-
tively intercept atmospheric nitrogen.

Table 4 Recommended Nitrogen Loading Limits for Coastal Embayments
(Adapted from U.S. Environmental Protection Agency and Massachusetts

Executive Office of Environmental Affairs, 1991)
Embayment SB Waters

1

SA Waters

2

Outstanding
Resource Areas

3

Shallow
Flushing < 4.5 350 mg/m

3

/Vr

4

200 mg/m

3

/Vr 100 mg/m

3

/Vr
Flushing > 4.5 30 g/m

2

/yr 15 g/m

2

/yr 5 g/m

2

/yr
Deep
Select rate resulting 500 mg/m

3

/Vr 260 mg/m

3

/Vr 130 mg/m

3

/Vr
in lesser annual loading 45 g/m

2

/yr 20 g/m

2

/yr 10 g/m

2

/yr

1

Excellent for fish, other aquatic life and wildlife, primary and secondary contact recreation,
and shellfish harvesting with depuration.

2

Excellent for fish, other aquatic life and wildlife, primary and secondary contact recreation,
and shellfish harvesting without depuration.

3

Outstanding socioeconomic, recreational, ecological, or aesthetic value.

4

Vollenweider flushing term;

Vr

equals

r

/1 +

r

equals flushing time (years).r

©2001 CRC Press LLC

STORMWATER RUNOFF

Development is accompanied by an increase in impervious surfaces which
increases stormwater runoff and decreases infiltration and evapotranspiration. This

decreases the time for the runoff to reach wetlands and streams, increasing the

Model Parameters

Impervious Surfaces Wastewater NO

3

35 mg/L
Roof Area 2,000 m

2

Roof runoff NO

3

0.75 mg/L
Paved Area 4,000 Paved runoff NO

3

1.5 mg/L
Natural Area 20,000 m

2

Fertilizer 1000 g/100 m

2

Lawn Area 10,000 m

2

Fertilizer leach rate 0.25
Wastewater 400 L/bedroom
Impervious surface recharge rate 1 meter/year
Natural area recharge rate 0.45 meter/year
Roof
2,000 m

2

×

1 m/yr

×

1,011 L/m

3



×

1 yr/365 day 5,539.7 L/day
L runoff/day

×

0.75 mg NO3/L = 4,154.8 mg/day
Paved
4,000 m

2



×

1 m/yr

×

1,011 L/m

3



×

1 yr/365 day 11079.5 L/day
L runoff/day

×

1.5 mg NO3/L = 16,619.3 mg/day
Natural
20,000 m

2



×

0.45 m/yr

×

1,011 L/m

3



×

1 yr/365 24,928.8 L/day
Lawn
10,000 m

2

×

1000 mg/100 m

2

/yr

×

1 yr/365 days 68.5 mg/day
Wastewater
3.5 bedrooms

×

400 L/bedroom

×

20 bedrooms 28,000 L wastewater per day
L wastewater per day

×

35 mg NO

3

/L= 980,000 mg/day

Cumulative nitrate nitrogen load

Figure 5

Nitrate nitrogen loading calculations for a hypothetical 20 house
residential development with an average of 3.5 bedrooms per
house.

Table 5 Percent Nitrogen Input to the Watershed, Loss within the
Watershed, and Input to Estuaries According to the
Waquoit Bay Land Margin Ecosystems Research Model

(Adapted from Valiela et al., 1997)
Source
Input to
Watershed
Losses within
Watershed
Input to
Estuaries

Atmospheric 56 89 30
Fertilizer 14 79 15
Wastewater 27 65 48
4 154.8 16 619.3 68.5 980 000 mg,++,+,
5 539.7 11079.5 24 928 28 000 liters,+,++,
14.39 mg/L=

©2001 CRC Press LLC

frequency and severity of erosion and downstream flooding. During periods of
prolonged dry weather, water tables and stream flows are reduced leading to a loss
of wetland and aquatic habitats.
Hydraulic and biological changes to streams occur when 10 to 20 percent of a
watershed has impervious surfaces [Massachusetts Department of Environmental
Protection and Massachusetts Office of Coastal Zone Management (DEP/CZM),
1997]. Typical impervious area percentages range from 20 to 40 percent for low
density residential developments to 95 to 100 percent for business districts
(Brach, 1989).
Stormwater runoff is contaminated with a variety of pollutants that have various
effects on wetland and aquatic habitats (Bingham, 1994; DEP/CZM, 1997). Nutri-
ents from animal wastes, human wastes, and fertilizers induce algal growth and
lower dissolved oxygen. Sedimentation also lowers dissolved oxygen as well as
increasing turbidity and smothering aquatic life. Pathogens contaminate drinking
water, swimming areas, and shellfish. Metals, hydrocarbons, organic chemicals,
and salt increase the toxicity of the water column and sediments and may bioac-
cumulate in aquatic organisms.
Stormwater impacts to wetland and aquatic habitats can be avoided or mini-
mized through careful planning, use of nonstructural practices, and use of structural
best management practices (BMPs). These measures reduce the volume of runoff,
store runoff water, promote infiltration of stormwater to groundwater, and remove
pollutants.

Planning and Nonstructural Practices

Effective stormwater management planning will minimize the size and cost of
structural requirements. Planning and nonstructural practices can mitigate most
stormwater impacts to wetlands and aquatic habitats for small developments. For
larger developments, planning and nonstructural practices can significantly reduce
the extent and, therefore, the cost of structural BMPs.

Perhaps the most important planning technique available is minimizing imper-
vious surfaces. Minimization of impervious surfaces is critical in recharge areas,
especially those associated with drinking water supplies. Methods for minimizing
impervious surfaces include the maintenance of natural buffers and drainageways.
This allows infiltration of runoff, reduces runoff velocity, and removes suspended
solids. Other methods include minimizing steep slopes, reducing building footprints
and parking areas, limiting the width of roadways and the use of sidewalks, using
shallow grassed roadside swales and parking lot islands, using turf pavers, gravel,
and other porous surfaces, and maintaining as much predevelopment vegetation as
possible (DEP/CZM, 1987).
Other planning techniques are also available for managing stormwater. Devel-
opments can be “fit” to the terrain by designing road patterns that match the landform.
Grassed waterways, vegetated drainage channels, and water quality swales can be
constructed along roadways to channel runoff (DEP/CZM, 1987). Similarly, natural,

©2001 CRC Press LLC

vegetated drainageways can be preserved, helping to maintain predevelopment flood
volumes, peak discharges, and base flows. Pollutants will be filtered by vegetation
and will bind to underlying soils and organic matter. The planning process should
also attempt to mimic predevelopment hydrologic conditions including peak dis-
charge, runoff volume, infiltration capacity, base flow levels, groundwater recharge,
and water quality.
Several nonstructural techniques can also be effective in managing stormwater
quantity and quality. Developments, especially commercial developments, will benefit
from preparation of a pollution prevention plan that identifies potential sources of
pollution and ensures implementation of practices that reduce pollutants in stormwater
discharges (DEP/CZM, 1987). Example techniques include proper pesticide and
fertilizer application, pet waste management, the proper storage, use, and disposal of
hazardous chemicals, and proper operation and maintenance of septic systems.

Areas accustomed to winter snowfalls use sand and salt to mitigate icy roadways.
Street and parking lot sweeping can reduce total suspended solids like sand and salt
by 5 to 80 percent. Vacuum sweepers tend to be more effective than mechanical
sweepers. Much of the solids also ends up in catch basins, which benefit from regular
cleaning. Salt (NaCl) toxicity can be minimized by using alternative de-icing com-
pounds such as calcium chloride (CaCl

2

) and calcium magnesium acetate (CMA),
and by designating “low salt” areas on roadways near wetlands and streams. De-
icing compounds should be stored on sheltered, impervious pads, and stored snow
should be placed where it can slowly infiltrate into the ground.

Structural BMPs

Structural BMPs are required when planning and nonstructural practices alone
are insufficient to mitigate stormwater impacts. There are five major categories of
stormwater structural BMPs: pretreatment, detention basins/retention ponds, vege-
tated treatment, infiltration, and filtration (Table 6). As with planning and nonstruc-
tural practices, the goals of a stormwater management design using structural BMPs
should be to approximate predevelopment runoff rates and volume and to maximize
pollutant removal. In many instances, the most effective design will incorporate
several BMPs in series.
Selecting a BMP requires consideration of the quantity of stormwater runoff to
be produced, the water quality to be achieved, the proximity of critical areas (e.g.,
wetlands, aquatic habitats), maintenance requirements, aesthetics, cost, and site
constraints (DEP/CZM, 1987; Schueler, 1987; Horner et al., 1994). In many
instances, site physical characteristics limit or determine BMP selection. For exam-
ple, sandy soils will inhibit the use of ponds but will facilitate infiltration BMPs.

Pond BMPs require a relatively large contributing drainage area, whereas infiltration
BMPs are restricted to a relatively small drainage area. A water table at or near the
surface is essential for wetlands and wet ponds but will preclude infiltration BMPs.
Swales and trenches are most effective when slopes are greater than 5 percent but
less than 20 percent. Wet ponds and wetlands should not outflow to cold water
streams, and infiltration BMPs should not be located close to foundations.

©2001 CRC Press LLC

Pretreatment

Sediment traps, water quality inlets, and catch basins remove debris, oil, and
grease, and sediment and associated pollutants. Settling is the primary treatment
mechanism. Sediment traps are on-line units, whereas water quality inlets and catch
basins are off-line units. Pretreatment BMPs should only be used as pretreatment
devices for other stormwater management technologies because they have limited
storage capacity, short detention times, and do not remove soluble pollutants. Essen-
tial elements of detention basin/retention pond and vegetated systems, pretreatment
BMPs are applicable to other BMPs as well. The longevity of pretreatment BMPs
is high with frequent maintenance.
A sediment trap is an excavated pit or cast structure 1 to 2 m (3 to 6 ft) deep.
Typically, sediment traps can accommodate the 2- and 10-year storms. Water quality
inlets and catch basins are chambered, underground retention systems. Water quality
inlets have multiple chambers with permanent pools of water in the first couple of
chambers. Floatable debris and sediments are trapped in the first chamber, oil and
grease are trapped in a second chamber, and water is routed out of a third chamber
into the storm drain or another BMP. Catch basins operate similarly, but only a single
chamber is present. Water quality inlets and catch basins are particularly applicable
to parking lots and other areas with substantial vehicular traffic.

Detention Basins/Retention Ponds

Detention basins capture and hold stormwater for 24 h or more, which permits
solids to settle and downstream flooding to be reduced. Typically, a detention basin
will have a lower stage capable of detaining smaller storms, and an upper stage
capable of detaining larger, less frequent storms. One of the less expensive BMPs

Table 6 The Major Types of Structural
Best Management Practices for

Stormwater

Pretreatment
Sediment traps
Water quality inlets
Catch basins
Detention/retention ponds
Extended detention ponds
Wet retention ponds
Vegetated treatment
Drainage channels
Water quality swales
Constructed wetlands
Infiltration
Dry wells
Trenches
Basins
Filtration
Basins

©2001 CRC Press LLC

capable of controlling both stormwater quantity and quality, detention basins remove
significant levels of sediment and sorbed pollutants. Detention basins are largely
ineffective at removing soluble pollutants.
Retention ponds use a deep, permanent pool of water to remove both solid and
soluble pollutants (Figure 6). Soluble pollutants, such as nutrients, are removed by
the biological activity of algae and fringing wetland. Retention ponds also have
additional storage capacity to control peak discharge rates. A pool depth of 0.9 to
1.8 m (3 to 6 ft) is recommended to optimize particle settling.
Both detention basins and retention ponds require a contributing watershed of
at least 4 ha (10 acres). Inflow points should have energy dissipaters and a forebay
or settling zone to trap course sediments. The original design should account for the
gradual accumulation of sediment. A routine inspection should be conducted at least
once a year, and sediment should be removed as necessary.

Vegetated Treatment

Treatment by vegetated BMPs ranges from simple drainage channels to con-
structed wetlands. Each vegetated BMP, to a degree, controls peak discharges by
reducing runoff velocity and promoting infiltration and reduces pollutants by trap-
ping, filtering, and infiltration. Both drainage channels and water quality swales are
most effective when the percentage of impervious cover in the contributing area is
relatively small and the slope is minimal (0 to 5 percent). Constructed wetlands
require relatively large contributing drainage areas to maintain dry weather base
flows. All vegetated BMPs benefit from inclusion of a sediment forebay to settle
large particles.
Drainage channels have grass or some other channel lining so runoff can be
conveyed during large storm events without causing erosion. Channels are applicable
to residential and other low to moderate density areas and can be used in parking

lots. A minimum channel length of 30 m (10 ft) is recommended to optimize pol-
lutant removal.

Figure 6

Retention ponds use a deep, permanent pool of water to remove solid and soluble
pollutants.

©2001 CRC Press LLC

Water quality swales are drainage channels enhanced to remove stormwater pol-
lutants and are typically sized to accommodate the 10-year storm. The three major
types of water quality swales are dry swales, wet swales, and grassed swales
(DEP/CZM, 1987; Claytor and Schueler, 1996). Dry swales allow for filtering or
infiltration through the bottom of the swale. Underlying soils should be permeable,
and the seasonal high water table should not be within 0.7 to 1.3 m of the swale bottom.
Wet swales are useful when the water table is at or near the soil surface, or soils are
poorly drained (Figure 7). Sediment accumulation, filtration, and vegetation uptake
remove pollutants. Grassed swales resemble dry swales in that underlying soils are
relatively permeable, but they are vegetated with moisture-tolerant grass species that
produce a fine, uniform, dense cover. Filtration, vegetation uptake, sediment accumu-
lation, and to some extent infiltration are the pollutant removal mechanisms.
Constructed wetlands are the most complex, and most expensive, vegetated
treatment BMP. Designed to mimic elements of natural wetlands, constructed wet-
lands reduce peak discharge and reduce the occurrence of downstream flooding,
settle particulate pollutants, and facilitate the uptake of pollutants by vegetation.
Constructed wetlands require relatively large contributing drainage areas to maintain
dry weather base flows, and construction costs are relatively high. Chapter 10 dis-
cusses constructed wetlands for the treatment of stormwater and other waste waters.

Infiltration

Infiltration BMPs are aggregate-filled devices which capture stormwater runoff
and gradually exfiltrate the runoff through the bottom of the device into the subsoil

Figure 7

Water quality swales remove stormwater pollutants by filtration, vegetation uptake,
sediment accumulation, and infiltration.

©2001 CRC Press LLC

and groundwater. Examples include dry wells and infiltration trenches and basins.
Infiltration BMPs require permeable underlying soils (minimum 1.3 cm/h) and a
groundwater level at least 0.7 m (2 ft) below the bottom of the infiltration device.
To avoid contaminating groundwater, infiltration BMPs should not be used when
runoff is highly contaminated. Infiltration BMPs can be used to manage peak dis-
charges and reduce runoff volume. Regular maintenance is required because infil-
tration BMPs are susceptible to clogging by sediments.
Dry wells are small pits used for infiltrating relatively good quality water such
as roof runoff. With a storage time of 48 to 72 h, dry wells typically have a contrib-
uting drainage area of less than 0.4 ha (1 acre).
Infiltration trenches are applicable to sites with gentle slopes (5 percent or less),
groundwater levels at least 1.2 m (4 ft) below the surface, and contributing drainage
areas of 2 ha (5 acres) or less. In addition to reducing runoff volume and peak
discharge, infiltration trenches remove soluble and particulate pollutants from runoff.
Infiltration trenches require pretreatment by inlets, sumps, swales, or forebays to
remove sediment and oil, and grease which may clog the trench.
Infiltration basins resemble detention/retention ponds but are constructed over
permeable soils. The contributing drainage area should be 6 ha (15 acres) or less,

depth to seasonal high water table at least 0.7 m (2 ft), and the soil infiltration rate
1.3 to 6 cm (0.5 to 2.4 in.) per hour. A sediment forebay or other pretreatment device
is necessary to capture coarse particulate pollutants. Infiltration basins are designed
to have a retention time of 48 to 72 h.

Filtration

Filtration basins consist of sand, peat, or compost underlain by gravel and
perforated underdrains. Filter fabrics may be installed at the top of the bed, and
between the filter media and gravel bed, to minimize clogging. Pollutant removal is
achieved by settling on top of the basin and by straining pollutants through the
filtering media. A sedimentation chamber designed to remove coarse pollutants
precedes the filter basin. Filter basins are applicable to small drainage areas of 0.2
to 2 ha (0.4 to 4 acres) for most development situations. A design filtration rate of
5 cm (2 in.) per hour is typical, and the filter should drain within 24 h.

REFERENCES

Bingham, D., Wetlands for stormwater treatment, in

Applied Wetlands Science and Technology

,
Kent, D. M., Ed., Lewis Publishers, Boca Raton, FL, 1995.
Brach, J.,

Protecting Water Quality in Urban Areas: Best Management Practices for Minne-
sota

, Minnesota Pollution Control Agency, Division of Water Quality, St. Paul,

MN, 1989.
Brown, K. W., Thomas, J. C., and Duble, R. L., Nitrogen source effect on nitrate and
ammonium leaching and runoff losses from greens,

J. Agron

., 74, 947, 1982.
Brown, W. E. and Caraco, D. S., Muddy water in - muddy water out?: a critique of erosion
and sediment control plans,

Watershed Prot. Tech

., 2(3), 393, 1997.

©2001 CRC Press LLC
Canter, L. W. and Knox, R. C.,

Septic Tank System Effects on Groundwater Quality

, Lewis
Publishers, Chelsea, MI, 1985.
Cape Cod Commission Water Resources Office, Nitrogen Loading, Technical Bulletin 91-001,
Barnstable, MA, 1992.
Cape Cod Planning and Economic Development Commission, Water Supply Protection
Project, final report, Barnstable, Bourne, Brewster, Dennis, and Yarmouth, Barnstable,
MA, 1979.
Cape Cod Planning and Economic Development Commission and U.S. Environmental
Protection Agency, Draft environmental impact statement and proposed 208 water
quality management plan for Cape Cod, Barnstable, MA, 1978.
Claytor, R. A. and Schueler, T. R., Design of Stormwater Filtering Systems, Center for

Watershed Protection, prepared for Chesapeake Research Consortium, Solomons,
MD, 1996.
Dahl, T. E., Wetland Losses in the United States 1780s to 1980s, U.S. Department of the
Interior, Fish and Wildlife Service, Washington, D.C., 1990.
Eichner, E. M. and Cambareri, T. C., The Cape Cod Golf Course Monitoring Project, Cape
Cod Commission, Water Resources Office, Barnstable, MA, 1990.
Heimlich, R. and Melanson, J., Wetlands lost, wetlands gained,

Natl. Wetlands Newsl.

,
May–June, 17(3), 1, 23, 1995.
Hollis, T. and Bedding, J., Can we stop wetlands from drying up?,

New Sci

., July, 31, 1994.
Horner, R. R., Skupien, J. J., Livingston, E. H., and Shaver, H. E.,

Fundamentals of Urban
Runoff Management: Technical and Institutional Issues

, Terrene Institute, Washington,
D.C., 1994.
Horner, R. R., Guedry, J., and Kortenhog, M. H., Improving the Cost Effectiveness of Highway
Construction Site Erosion and Pollution Control, Washington State Transportation
Center, Federal Highway Administration, Seattle, WA, 1990.
IEP,

Water Resources Protection Study: Town of Yarmouth, Massachusetts

, Sandwich,
MA, 1988.
Kelly, J. R. and Levin, S. A., A comparison of aquatic and terrestrial nutrient cycling and
production processes in natural ecosystems, with reference to ecological concepts of
relevance to some waste disposal issues, in

The Role of the Oceans as a Waste Disposal
Option

, Kullenberg, G., Ed., NATO Advanced Research Workshop Series, Reidel,
Dordrecht, The Netherlands, 1986, 154.
Loehr, R. C., Characteristics and comparative magnitude of nonpoint sources,

J. Water Pollut.
Control Fed.

, 46, 1849, 1974.
Long Island Regional Planning Board, Special Groundwater Protection Area Project for the
Oyster Bay Pilot Area and Brookhaven Pilot Area, 1986.
Massachusetts Department of Environmental Protection and Massachusetts Office of Coastal
Zone Management,

Stormwater management

, Vol. 2,

Stormwater Technical Hand-
book

, MADEP, 1997.
McManus, K., Wetlands avoidance and impact minimization, in

Applied Wetlands Science
and Technology

, Kent, D. M., Ed., Lewis Publishers, Boca Raton, FL, 1994, 105.
Mitsch, W., The world’s wetlands and SWS—a call for an international view,

Wetlands Bull

.,
September, 1, 4, 1995.
Mitsch, W. H. and Knight, R. L.,

Treatment Wetlands

, Lewis Publishers, Boca Raton, FL, 1997.
Nassau-Suffolk Regional Planning Board, The Long Island Comprehensive Waste Treatment
Management Plan, Hauppauge, NY, 1978.
Nelson, K. A., Design considerations for segmental retaining walls in water environments,

Land Water

, July/August, 22, 1995.