Assessing cumulative loss of wetland functions in the Nanticoke River Watershed uing enhanced national wetlands inventory data

              405
WETLANDS, Vol. 25, No. 2, June 2005, pp. 405–419
  2005, The Society of Wetland Scientists
ASSESSING CUMULATIVE LOSS OF WETLAND FUNCTIONS IN THE
NANTICOKE RIVER WATERSHED USING ENHANCED NATIONAL WETLANDS
INVENTORY DATA
Ralph W. Tiner
U.S. Fish and Wildlife Service
Northeast Region
300 Westgate Center Drive
Hadley, Massachusetts, USA 01035
E-mail: ralph tiner@fws.gov
Abstract: The coterminous U.S. has lost more than 50% of its wetlands since colonial times. Today, wet-lands
are highly valued for many functions including temporary storage of surface water, streamflow main-tenance,
nutrient transformation, sediment retention, shoreline stabilization, and provision of fish and wildlife
habitat. Government agencies and other organizations are actively developing plans to help protect, conserve,
and restore wetlands in watersheds. The U.S. Fish and Wildlife Service’s National Wetlands Inventory
Program (NWI) has produced wetland maps, digital geospatial data, and wetland trends data to aid these
and other conservation efforts. Most recently, the NWI has developed procedures to expand the amount of
information contained within its digital databases to characterize wetlands better. It has also developed
techniques to use these data to predict wetland functions at the watershed level. Working with the states of
Delaware and Maryland, the NWI applied these techniques to the Nanticoke River watershed to aid those
states in developing a watershed-wide wetland conservation strategy. Wetland databases for pre-settlement
and contemporary conditions were prepared. An assessment of wetland functions was conducted for both
time periods and comparisons made. Before European settlement, the Nanticoke watershed had an estimated
93,000 ha of wetlands covering 45% of the watershed. By 1998, the wetland area had been reduced to 62%
of its original extent. Sea-level rise and wetland conversion to farmland were the principal causes of wetland
loss. From the functional standpoint, the watershed lost over 60% of its original capacity for streamflow
maintenance and over 35% for four other functions (surface-water detention, nutrient transformation, sedi-ment
and particulate retention, and provision of other wildlife habitat). This study demonstrated the value
of enhanced NWI data and its use for providing watershed-level information on wetland functions and for
assessing the cumulative impacts to wetlands. It provides natural resource managers and planners with a tool
that can be applied consistently to watersheds and large geographic areas to show the extent of wetland
change and its projected effect on wetland functions.
Key Words: cumulative wetland impacts, historic wetlands, Nanticoke River watershed, National Wetlands
Inventory, wetland classification, wetland functional assessment, wetland trends
INTRODUCTION
Many investigators have reported significant losses
of wetlands in the United States (e.g., Frayer et al.
1983, Tiner and Finn 1986, Dahl and Johnson 1991,
Hefner et al. 1994, Tiner et al. 1994, Dahl 2000).
These reports address wetland trends in terms of area
lost or area gained but do not address the significance
of the loss in functional terms. In the past decade, there
has been considerable interest in wetland functional
assessment at both the site-specific and landscape or
watershed levels. The latter assessments require the
use of geospatial data and geographic information
technology (GIS). Several states in the Northeast with
interest in landscape-level analysis have cooperated
with the U.S. Fish and Wildlife Service (FWS) in pro-ducing
watershed-level assessments of wetland func-tions.
Among the areas evaluated were watersheds as-sociated
with Maine’s Casco Bay, New York City’s
water supply system, the Nanticoke River of Maryland
and Delaware, and Maryland’s Coastal Bays plus
Pennsylvania’s Coastal Zone (Tiner et al. 1999, 2000,
2001, 2002, 2004, Tiner and DeAlessio 2002, Tiner
and Stewart 2004). To accomplish this work, the
FWS’s Northeast Region developed a technique to
prepare preliminary assessments of wetland functions
for watersheds and large geographic areas (Tiner
2002). The technique requires enhancing digital Na-tional
Wetlands Inventory (NWI) data by adding de-scriptors
for landscape position, landform, water flow
path, and waterbody type (LLWW) to the NWI digital
database and then applying correlations between wet-
406 WETLANDS, Volume 25, No. 2, 2005
Figure 1. Locus map showing the Nanticoke River water-shed
on the Delmarva Peninsula.
land characteristics and functions to identify wetlands
of potential significance for various functions. When
applied to different-era datasets for wetlands in the
same watershed, this assessment approach provides a
perspective on the magnitude of the losses from a
functional standpoint.
The states of Maryland and Delaware are working
cooperatively to develop a watershed-based strategy
for wetland conservation and restoration for the Nan-ticoke
River watershed. They contacted the FWS for
assistance in conducting watershed-level assessments
of wetlands, first for the present era and then for the
pre-settlement period. The purpose of the investigation
was to produce an inventory and analysis of historic
wetlands and their functions for the Nanticoke River
watershed and to compare these findings to present-day
conditions. The specific objectives were 1) to pro-duce
a map showing the general extent of wetlands
prior to European colonization, 2) to prepare a prelim-inary
functional assessment of pre-settlement wet-lands,
3) to create a consistent database of contem-porary
wetlands for the entire watershed from existing
enhanced NWI data, 4) to prepare a preliminary wet-land
functional assessment for the present-day water-shed,
and 5) to compare the changes in wetland extent
and functions based on the pre-settlement and contem-porary
wetland assessments. This paper generally de-scribes
the assessment method and demonstrates its
use for predicting the cumulative effect of historic wet-land
losses on wetland functions for the Nanticoke
River watershed.
Study Area
The study area is the Nanticoke River watershed, a
tributary of the Chesapeake Bay, beginning in western
Delaware on the Delmarva Peninsula and flowing in a
southwesterly direction into Chesapeake Bay (Figure
1). This watershed is roughly 2,070-km2 in size and
includes about 25% of the state of Delaware. Major
tributaries include five in Delaware (Broad Creek,
Deep Creek, Gravelly Branch, Gum Branch, and Mar-shyhope
Creek) and four in Maryland (Marshyhope
Creek, Rewastico Creek, Quantico Creek, and Wetip-quin
Creek).
METHODS
Pre-settlement Wetland Inventory
Reconstructing the distribution of historic wetlands
requires using varied sources of information and mak-ing
certain assumptions. Regardless of the procedures
employed, the outcome is an approximation and not
an exact replication of pre-settlement conditions. For
this study, the distribution and extent of pre-settlement
wetlands were derived from two sources: 1) soil sur-vey
data from the U.S.D.A. Natural Resource Conser-vation
Service (NRCS) and the Delaware Department
of Natural Resources and Environmental Control
(DNREC) based on 1:15,840 to 1:20,000 soil maps
and 2) U.S. Geological Survey orthophotomaps (1:
24,000). The former source was the primary source,
and most historic wetlands were identified from this
material, since urban development was minor com-pared
to agricultural impacts. The orthophotomaps
were used to locate ‘‘lost’’ estuarine wetlands that are
now shallow water.
Hydric soil map units from soil survey data were
identified as historic wetlands. A digital database of
hydric soil map units was created for the Nanticoke
watershed from existing digital soil survey data and
from soil map unit data in published soil surveys. Two
counties had digital soils data available: Dorchester
(SSURGO data from NRCS based on Brewer et al.
1998) and Sussex (from DNREC). For other counties
(Caroline, Wicomico, and Kent), hydric soil digital
data were created by scanning individual soil survey
maps from county soil survey reports (Matthews 1964,
Hall 1970, Matthews and Ireland 1971, respectively).
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS���NANTICOKE RIVER WATERSHED 407
Scanning was done at 300 dots per inch (dpi) and
saved as TIFF images. The black color band (all line-work)
was selected in each image and copied to form
a composite image (mosaic) for the county. Mosaics
were georeferenced in ArcGIS 8.0 using the georefer-encing
extension, with a 1:24,000 digital raster graph-ics
(DRG) serving as the base. These mosaics were
then converted to georeferenced GRIDS and then to
linear coverages, which were converted to polygonal
coverages and finally to shapes. The shapes were ed-ited
and hydric soil map units labeled using the geo-referencing
image to code ID in the background in
ArcGIS 8.3.
The soil-based historic wetland data were compared
with existing NWI data to identify possible large wet-land
complexes (typically forested wetlands) that were
not recorded as historic wetlands based on soils map-ping
(e.g., likely hydric inclusions in larger nonhydric
soil units). Due to alignment issues caused by merging
data sources, a 5-ha threshold was established for iden-tifying
significant omissions. These larger NWI wet-lands
were added to the historic data base. The pre-sumption
was that if the area is a large forested wet-land
today, it was likely a forested wetland at the time
of European settlement.
Estuarine wetlands have migrated landward and up-river
due to sea-level rise over the past 500 years,
while others have become permanently inundated.
Consequently, the pre-settlement estuarine-riverine
break had to be relocated further downriver than its
current location, and ‘‘lost’’ estuarine wetlands had to
be added to the database. For the former, the presence
of soils recognized as submerged uplands and the ap-pearance
of salt-stressed forests were used to establish
this break at the mouth of the Baron Creek. Under-standably,
this is a conservative demarcation, as it is
likely that freshwater forested wetlands also occurred
downstream along the edges of estuarine wetlands.
The Honga and Sunken series (submerged ‘‘uplands,’’
now brackish tidal wetlands) both represent former
lowland forests (likely palustrine forested wetlands or
wet flatwoods similar to those occurring today on
Othello and Elkton soils) that became estuarine wet-lands
with rising sea level over the past few hundred
years. The former is an organic soil (Terric Sulfihe-mists)
with more than 40 cm of organic matter, where-as
the latter is a mucky silt loam soil (Typic Ochra-quults)
with a surface layer of only 5–20 cm of organic
matter (Brewer et al. 1998). The Sunken series is typ-ified
by salt-stressed (dying or dead) stands of loblolly
pine (Pinus taeda L.) and some areas have become
salt/brackish marshes. While both series represent for-mer
forest, for purposes of this study, only the Sunken
series was identified as pre-settlement freshwater for-ested
wetlands. Given the thickness of its organic ho-rizon,
the Honga series most likely became estuarine
wetland more than 300 years ago (e.g., wood found in
the organic and mineral horizons was carbon-dated at
less than 700 years before present; Brewer et al. 1998).
Pone soils were designated as temporarily flooded-tid-al
forested wetlands where contiguous with tidal marsh
soils; in other places, they were designated as nontidal
temporarily flooded forested wetlands. Muck soils and
contiguous soils that are now estuarine wetlands were
also identified as historic tidal forested wetlands. Else-where,
muck soil map units were regarded as non-tidal
forested wetlands. The Nanticoke series and the tidal
marsh map units from the soil surveys were considered
pre-settlement freshwater tidal marshes. The pre-co-lonial
limits of estuarine and freshwater tidal reaches
represent approximate boundaries, mainly used to in-dicate
a significant ecological and hydrologic change
in this watershed over time. It is further recognized
that the upstream limit of tidal influence was probably
downstream from its current location, but approximat-ing
this limit was not possible.
To identify ‘‘lost’’ estuarine wetlands due to sea-level
rise over the past few hundred years, U.S. Geo-logical
Survey 1:24,000 orthophotomaps (Deal Island
1972, Mardela Springs 1982, Nanticoke 1983, and
Wetipquin 1983) were consulted. The 2-m depth
shown on these maps represents a convenient approx-imation
of the lower limit of the intertidal zone 600
years ago; recorded depths within this boundary are
mostly listed as 1 m below mean low water. Given a
spring tide range of 0.8 to 0.9 m for the Nanticoke
River (http://co-ops.nos.noaa.gov/tide pred.html) and a
near constant rate of sea-level rise of 1.4 mm/yr in
Chesapeake Bay over the past 6,000 years (Curtis Lar-sen,
U.S. Geological Survey, pers. comm.; Larsen
1998), these shallow water areas were predicted to be
estuarine wetlands (probably some combination of tid-al
marshes and flats) around 1400 AD.
Impounded sections of rivers (i.e., artificial in-stream
ponds and lakes) shown on the soil surveys
were classified as forested wetlands similar to contig-uous
wetlands above and below the impoundment.
Some minor area of open water was probably included
in the wetland area estimate following this interpreta-tion.
After pre-settlement wetlands were identified, they
were classified according to NWI types (Cowardin et
al. 1979). All inland wetlands were classified as pal-ustrine
forested wetlands, recognizing that periodic
wildfires would have created a succession of types
from emergent wetlands through shrub swamps to for-ested
wetlands, much like we observe today after tim-ber
harvest. According to the 1920s soil surveys, most
of the soils were forested in their original state (e.g.,
Wicomico County was ‘‘practically’’ all forest until
408 WETLANDS, Volume 25, No. 2, 2005
‘‘reclaimed for agricultural purposes;’’ Snyder and
Gillett 1925). Water regimes were based on hydrology
data for soil map units published in the soil survey
reports.
The condition of the historic landscape is therefore
much simplified. No attempt was made to separate for-ested
wetlands into different types at the subclass level
according to Cowardin et al. (1979) or to account for
the effect of increased sedimentation on estuarine wet-lands
following conversion of forests to agricultural
land, since these patterns were impossible to predict.
1998 Wetland Inventory
The distribution, extent, and classification of pre-sent-
day wetlands were based on NWI mapping. NWI
data for the Nanticoke watershed were recently updat-ed
using spring 1998–1:40,000 black and white pho-tography
(see Tiner et al. 2001, 2000 for details). Wet-lands
were classified according to the FWS’s official
wetland classification system (Cowardin et al. 1979).
Enhanced Wetland Classification
The NWI database was expanded to include descrip-tors
for landscape position, landform, water flow path,
and waterbody types (LLWW descriptors). They were
applied to all wetlands and deepwater habitats in the
NWI digital database by merging NWI data with on-line
U.S. Geological Survey topographic maps (digital
raster graphics), consulting aerial photography where
necessary, and interpreting dichotomous keys to the
descriptors (Tiner 2003a; Table 1). Enhanced classi-fication
was applied to both the pre-settlement and
1998 wetlands.
Preliminary Assessment of Wetland Functions
This study employed a landscape-level wetland as-sessment
approach called ‘‘Watershed-based Prelimi-nary
Assessment of Wetland Functions’’ (W-PAWF).
W-PAWF applies general knowledge about wetlands
and their functions to produce a watershed profile
highlighting wetlands of potential significance for nu-merous
functions. The method was developed to pre-dict
wetland functions for large geographic areas, par-ticularly
watersheds, from NWI data. To do this, two
steps must be undertaken: 1) the digital NWI database
must be expanded by adding LLWW descriptors, and
2) correlations between wetland characteristics in the
database and wetland functions must be developed.
Many wetland functions are related to physical prop-erties,
while others are dependent on a combination of
biological and physical characteristics. For example,
floodplain and depressional wetlands temporarily store
surface water, whereas slope wetlands do not; wet-lands
that are sources of streams are vital for stream-flow
maintenance; marshes provide habitat for water-fowl
and waterbirds.
In W-PAWF, ten wetland functions are evaluated:
1) surface-water detention, 2) streamflow maintenance,
3) nutrient transformation, 4) sediment and other par-ticulate
retention, 5) coastal storm-surge detention (for
tidal regions only), 6) shoreline stabilization, 7) pro-vision
of fish and shellfish habitat, 8) provision of wa-terfowl
and waterbird habitat, 9) provision of other
wildlife habitat, and 10) conservation of biodiversity
(e.g., rare or uncommon wetland types in the water-shed
based on NWI mapping or photointerpretable
wetland types of regional significance for biodiversi-ty).
The rationale for correlating wetland characteris-tics
with these functions for the Northeast is described
in Tiner (2003b). Correlations are based on a review
of the literature and application of best professional
judgment from many wetland biologists and resource
specialists in the Northeast.
After the digital databases for pre-settlement and
contemporary wetlands were constructed (including
LLWW descriptors), analyses were performed to pro-duce
a preliminary assessment of wetland functions for
the watershed for each era. Correlations between wet-land
functions and characteristics were applied to the
enhanced NWI database to identify wetlands that may
be performing each function at significant levels. The
conservation of biodiversity function was not evalu-ated
for the pre-settlement era since source data were
limited.
After running the analyses, a series of maps were
generated by ArcView 3.x to highlight wetlands that
may perform these functions at high or other signifi-cant
levels. Area summaries for each function were
generated from Microsoft’s Access program. The tar-geted
wetlands were predicted to perform a given
function at a significant level presumably important to
the watershed’s ability to provide that function. ‘‘Sig-nificance’’
is a relative term and is used in this analysis
to identify wetlands that are likely to perform a given
function at a level above that of wetlands not desig-nated.
Function Comparison: Pre-settlement vs. 1998
To assess the impact of cumulative loss of wetlands
on specific functions, one can simply examine the
change in area of functionally significant wetlands.
This was done, but the area difference alone may not
adequately convey the cumulative impact on wetland
functions. To address the latter, a simple weighting
scale for wetlands of potential significance for each
function was devised. A ‘‘high’’ potential was given
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 409
Table 1. Simplified keys for classifying wetlands by landscape position, landform, and water flow path. (Adapted from Tiner 2003a)
Landscape Position
1. Wetland borders a river, stream, in-stream pond, lake, reservoir, estuary, or ocean . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2
1. Wetland does not border one of these waterbodies; it is completely surrounded by upland or borders a pond surrounded by
upland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Terrene
2. Wetland lies along an ocean shore and is subject to tidal flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marine
2. Wetland does not lie along an ocean shore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
3. Wetland lies along an estuary (salt to brackish tidal waters) and is subject to tidal flooding . . . . . . . . . . . . . . . . . . . . . . . .Estuarine
3. Wetland does not lie along an estuary or if so, it is not subject to tidal flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
4. Wetland lies along a lake or reservoir or within its basin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lentic
4. Wetland lies along a river, stream, or in-stream pond, or borders an estuarine wetland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
5. Wetland is the source of a river or stream and this watercourse does not extend through the wetland . . . . . . . . . . . . . . . . . . Terrene
5. River or stream flows through the wetland, or wetland borders an estuarine wetland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
6. Wetland is periodically flooded by river or stream. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lotic1
6. Wetland is not periodically flooded by the river or stream or by tides (episodic flooding may occur). . . . . . . . . . . . . . . . . . . Terrene
Landform
1. Wetland occurs on a slope  2%. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Slope
1. Wetland does not occur on a slope  2% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2
2. Wetland forms an island completely surrounded by water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Island
2. Wetland does not form an island . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
3. Wetland occurs in the shallow water zone of a permanent non-tidal waterbody, the intertidal zone of an estuary with unrestricted tidal
flow, or the regularly flooded (daily tidal inundation) zone of freshwater tidal wetlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fringe
3. Wetland does not occur in these waters or intertidal zones with unrestricted tidal flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
4. Wetland occurs in a portion of an estuary with restricted tidal flow due to tide gates, undersized culverts, dikes, or similar
obstructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basin
4. Wetland does not occur in such location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
5. Wetland forms a non-vegetated bank or is within the banks of a river or stream. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fringe
5. Wetland is not a non-vegetated riverbank or streambank or within the banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
6. Wetland occurs on an active alluvial plain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Floodplain*
6. Wetland does not occur on an active floodplain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
7. Wetland occurs on a broad interstream divide (including headwater positions) associated with coastal or glaciolacustrine plains or
similar plains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Interfluve*
7. Wetland does not occur on such a landform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8
8. Wetland occurs in a distinct depression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basin
8. Wetland occurs on a nearly level landform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Flat
Water Flow Path2
1. Wetland is typically surrounded by upland (non-hydric soil); receives precipitation and runoff from adjacent areas with no apparent
outflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isolated**
1. Wetland is not geographically isolated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2
2. Wetland is a sink receiving water from a river, stream, or other surface-water source, and lacking surface-water outflow . . . . . Inflow
2. Wetland is not a sink; surface water flows through or out of the wetland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
3. Wetland is subjected to tidal flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bidirectional-Tidal
3. Wetland is not tidally influenced. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
4. Water flows out of the wetland, but does not flow into this wetland from another source . . . . . . . . . . . . . . . . . . . . . . . . . . .Outflow
4. Water flows in and out of the wetland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
5. Water flows through the wetland, often coming from upstream or uphill sources (typically wetlands along rivers and
streams) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Throughflow
5. Wetland is along a lake or reservoir and its water levels are subjected to the rise and fall of this waterbody . . Bidirectional-Nontidal
1 Lotic wetlands are separated into river and stream sections (based on watercourse width at map scale of 1:24,000 – polygon   Lotic River vs. linear  
Lotic Stream) and then divided into one of five gradients: 1) high (e.g., shallow mountain streams on steep slopes), 2) middle (e.g., streams with moderate
slopes), 3) low (e.g., mainstem rivers with considerable floodplain development), 4) intermittent (subject to periodic flows), and 5) tidal (hydrology under
influence of the tides).
2 Surface-water connections are emphasized because they are more readily identified than groundwater linkages.
* Basin and Flat sub-landforms can be identified within these landforms when desirable.
** Wetland is geographically isolated; hydrological relationship to other wetlands and watercourses may be more complex than can be determined by
simple visual assessment of surface-water conditions.
410 WETLANDS, Volume 25, No. 2, 2005
a weight of 2, while a ‘‘moderate’’ potential and other
potentially significant wetlands (i.e., shading for fish
habitat and wood duck habitat) were assigned a weight
of 1. By multiplying the wetland area listed as high,
moderate, or other potential by the weighting factor, a
total number of functional units was calculated for
each function at pre-settlement and 1998. This allowed
comparison between pre-settlement functional capacity
(total functional units for time one) and the 1998 ca-pacity
(total functional units for time two) and could
demonstrate a percent loss of pre-settlement function.
This provides an interesting perspective on the current
conditions from a functional capacity standpoint and
may give a better sense of the relative magnitude of
the functional loss than change in wetland area alone.
RESULTS
The wetland database created for this project al-lowed
production of wetland maps and statistics on
wetland extent and predicted functions for two time
periods (pre-settlement and 1998). Two sets of water-shed-
scale maps (1:110,000) were produced to profile
the Nanticoke’s wetlands—one set showing estimated
pre-settlement conditions and predicted wetlands of
significance for nine functions (excluding conservation
of biodiversity) and the other set showing 1998 con-ditions
and predicted wetlands of significance for ten
functions. These maps are multi-colored and too de-tailed
to present in this paper; they display wetlands
by NWI types, landscape position, landform, water
flow path, and potential significance for each of ten
functions. An example of a reduced version is pre-sented
as Figure 2; examples of similar maps for the
Maryland portion of the watershed can be viewed on
the web at: http://wetlands.fws.gov/Pubs Reports/
Md Watershed/Md watershed.htm.
Wetland Extent Comparison
Trends by Generalized NWI Types. There have been
significant changes in wetland and aquatic resources
since pre-settlement times (Figure 3). Prior to Euro-pean
settlement, an estimated 93,125 ha of wetlands
may have existed in the Nanticoke watershed (Table
2). Ninety percent of the predicted wetland area was
represented by palustrine (freshwater) wetlands, most-ly
nontidal (79,537.5 ha). Most (88.5%) of the wet-lands
were forested, with the rest being classified as
emergent (10.3% as estuarine and 1.2% as palustrine).
The actual extent of palustrine emergent wetlands was
undoubtedly greater than estimated due to fire impacts,
but there were no data to predict this effect. The es-timates
also do not include any predicted area of pal-ustrine
scrub-shrub wetlands for similar reasons.
By 1998, the Nanticoke’s wetland area had fallen to
57,492 ha (about 62% of the pre-settlement total).
Eighty-eight percent was palustrine wetland, with for-ested
and scrub-shrub wetlands accounting for over
46,000 ha. This figure includes many wetlands in post-harvest
succession. Estuarine wetlands accounted for
nearly 12% of the watershed’s wetland area. Irregu-larly
flooded emergent wetlands predominated, occu-pying
over 6,000 ha (about 93% of the Nanticoke’s
estuarine wetlands; Table 3). Although Table 2 shows
a tremendous increase in palustrine non-tidal emergent
wetland area, the huge difference is an artifact, related
more to the detailed wetland mapping in 1998 vs. gen-eralized
pre-settlement data. Much emergent wetland
area in 1998 resulted from timber harvest operations
converting forested wetland to emergent wetland and
with some increase in emergent wetland also due to
pond construction. Table 3 gives a more detailed ac-counting
of present-day wetlands by NWI types.
Trends by LLWW Types. At pre-settlement, an esti-mated
2,809 wetlands occupied over 93,000 ha of the
watershed (Table 4). Seventy-eight percent of the wet-land
area was represented by terrene wetlands, while
lotic wetlands comprised 12% of the area; estuarine
wetlands made up 10%. About 77% of the wetlands
were interfluve types. Fringe wetlands constituted 11%
and floodplain wetlands about 10% of the wetland
area. From the water flow perspective, 73% of the wet-land
area experienced outflow, 15% bidirectional-tidal
flow, 7% throughflow, and 5% was isolated (complete-ly
surrounded by nonwetland).
By 1998, the Nanticoke’s wetland area had been
reduced by 39%, while the number of wetlands (ex-cluding
ponds) increased 1.75 times to 4,920 due
largely to fragmentation by roads and agricultural
fields. The ratio of wetland types comprising the Nan-ticoke’s
wetlands changed slightly with the significant
decrease in wetland area. Terrene wetlands now rep-resent
about 72% of the wetland area (excluding
ponds), while estuarine wetlands comprise 16% and
lotic wetlands 12%. Lentic wetlands created from
dammed rivers or streams or by excavating and diking
terrene wetlands occupy only 0.2% of the area. From
the landform standpoint, interfluve wetlands account
for 71% of the wetland area, followed by fringe wet-lands
(17%) and floodplain wetlands (11%). Other
wetland landforms now represent less than 2% of the
area (flats  1.1%; basins  0.5%, and islands  0.2%).
Outflow wetlands remain the predominant water-flow-path
type, totaling 38,539 ha (68% of the wetland
area). Bidirectional-tidal wetlands are second-ranked
with 10,434 ha (18% of the wetland area), followed
by throughflow wetlands with 5,917 ha (10%). Isolated
wetlands amount to 2,029 ha (4%) and bidirectional-
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 411
Figure 2. Example of published thematic map (reduced in size and black & white copy of color map) highlighting pre-settlement
wetlands of potential significance for streamflow maintenance. Black areas represent wetlands with high potential
for contributing significantly to stream flow. (Tiner and Bergquist 2003)
412 WETLANDS, Volume 25, No. 2, 2005
Figure 3. Nanticoke River watershed’s wetlands and deepwater habitats at pre-settlement and in 1998. Black areas are
deepwater habitats; gray areas are wetlands (including ponds).
nontidal wetlands associated with impoundments total
only 105 ha (0.2%).
Since pre-settlement, terrene wetlands experienced
the greatest loss, decreasing by nearly 44%, with ter-rene
interfluve wetlands being most adversely affected.
Habitat fragmentation was significant, with the mean
size of interfluve wetlands dropping to a third of their
original size (i.e., from 33.8 ha at pre-settlement to
10.6 ha in 1998). By 1998, the mean size of the most
abundant wetlands—terrene outflow wetlands—had
decreased from 175 ha to 18 ha, while their number
increased nearly 6-fold (from 380 to 2120). Only 63%
of the pre-settlement lotic wetland area remained in
1998; lotic river wetlands experienced the greatest loss
( 70%) from about 4,100 ha to roughly 1,200 ha. The
area of estuarine wetlands dropped by an estimated
4%, and lentic wetlands became established in river
impoundments and diked former terrene wetlands.
Ponds were created from both wetlands and uplands.
The proportion of wetland area represented by differ-ent
landforms changed slightly, with a drop in inter-fluve
wetlands (77 to 71%) and an increase in fringe
and floodplain types (11 to 17% and 10 to11%, re-spectively).
The percent of outflow wetland area fell
from 73 to 68%, whereas the percent represented by
throughflow and bidirectional-tidal flow rose from 7 to
10% and 15 to 18%, respectively.
Causes of Wetland Trends
Both natural processes and human activities were
responsible for the predicted wetland losses. The chief
natural process was sea-level rise, which affected both
estuarine and palustrine wetlands. Most of the estua-rine
wetlands lost became shallow estuarine water due
to increased erosion and submergence. This hydrologic
change also moved the salt wedge further upstream
and inland converting many areas of freshwater wet-
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 413
Table 2. Historic trends in the Nanticoke’s wetland area (in hectares) by generalized NWI types: pre-settlement vs. 1998. (Note: 1998
data are more detailed than presented here – see Table 3; 1998 types have been aggregated for comparison with pre-settlement types;
totals may be slightly different than sum of numbers duo to round-off procedures.)
Wetland Type
Pre-settlement Area
(% of Wetlands)
1998 Area
(% of Wetlands)
Net Area Change
(% of Pre-settlement)
Estuarine Intertidal 9,569.6 (10.3) 6,849.4 (11.9)  2,720.2 ( 28.4%)
Palustrine Emergent
Tidal
Nontidal
1,091.7 (1.2)
25.7 ( 0.1)
273.2* (0.5)
2,169.3 (3.7)
 818.5 ( 75.0%)
 2,143.6 ( 8341%)**
Total 1,117.4 (1.2) 2,442.5 (4.2)  1,325.1 ( 118.6%)
Palustrine Forested
Tidal
Nontidal
2,926.4 (3.1)
79,511.8 (85.4)
3,307.9*** (5.8)
42,975.5*** (74.8)
 381.5 ( 13.0%)
 36,536.3 ( 46.0%)
Total 82,438.2 (88.5) 46,283.4 (80.5)  36,154.8 ( 43.9%)
Other Palustrine
Farmed
Ponds
-0-
-0-
1,428.3 (2.5)
488.0 (0.8)
 1,428.3 (NA%)
 488.0 (NA%)
Total -0- 1,916.3 (3.3)  1,916.3 (NA%)
Grand Total 93,125.2 57,491.6  35,633.6 ( 38.3%)
* Includes 153.3 ha of riverine tidal wetlands, mostly marshes.
** This increase is an artifact, since the pre-settlement extent of non-tidal emergents could not be accurately established.
*** Includes scrub-shrub wetlands and mixed communities where forested or scrub-shrub wetland was the dominant class.
lands (lotic river wetlands) to estuarine wetlands in the
lower portion of the watershed. This process continues
today, as witnessed by stumps and dead trees in es-tuarine
marshes and dying or salt-stressed trees in
neighboring landward areas. Rising sea level undoubt-edly
had the effect of extending tidal influence up-stream,
thereby changing the hydrology of former non-tidal
wetlands to a freshwater tidal regime. With Eu-ropean
settlement and subsequent population growth,
drainage and conversion of much palustrine forested
wetland to farmland took place for more than 200
years. In 1998, over 1,400 ha were classified as farmed
wetlands, while the bulk of the drained wetlands were
converted to non-wetland agricultural fields. Farming
is the predominant land use in the watershed today.
Many of the remaining palustrine wetlands are frag-mented
by roads and cropland. Human activities also
resulted in the creation of about 500 ha of ponds built
in both wetlands and uplands.
Trends by Wetland Function
Two comparisons of changes in functions were
made, one showing changes in wetland area providing
functions at significant levels (Table 5) and the other
depicting changes in functional units (Table 6). From
an area standpoint, substantial losses of wetlands pro-viding
all functions ranged from an over 50% area loss
in wetlands important for sediment retention to a 23%
loss of wetlands stabilizing shorelines and those stor-ing
coastal storm surge. More than 30% of the wetland
area rated as significant for performing half of the wet-land
functions evaluated was lost. Wetlands that
served as sources of streams (headwater wetlands)
were most negatively impacted; 87% of the pre-settle-ment
wetland area predicted as having high potential
for this function was altered. Ditching of terrene in-terfluve
wetlands either effectively drained many of
these headwater wetlands, converting them to cropland
(upland), or diminished the duration of their seasonal
wetness (reducing their contribution to streamflow).
When functional units were evaluated, the change
in the watershed’s ‘‘functional capacity’’ may be bet-ter
realized (Table 6). For most of the wetland func-tions
evaluated, the Nanticoke watershed is predicted
to be operating at 50 to 77% of its original capacity.
The streamflow maintenance function supported by
wetlands is operating at only 36% of its original ca-pacity;
this has undoubtedly had significant adverse
impacts on aquatic biota. The watershed’s capacity for
providing six other functions decreased by more than
28% (i.e., surface-water detention, nutrient transfor-mation,
sediment and other particulate retention, fish
and shellfish habitat, waterfowl and waterbird habitat,
and other wildlife habitat). The two remaining func-
414 WETLANDS, Volume 25, No. 2, 2005
Table 3. Wetlands in the Nanticoke watershed in 1998 classified by NWI wetland type to the class level (Cowardin et al. 1979).
NWI Wetland Type Area (ha)
Estuarine Wetlands
Emergent (Regularly Flooded)
Emergent (Irregularly Flooded)
Scrub-Shrub (Irregularly Flooded)
Forested (Irregularly Flooded)
Unconsolidated Shore (Irregularly Exposed)
Unconsolidated Shore (Regularly Flooded)
259.2 (96.9   oligohaline)
6,203.8 (2,469.7   oligohaline)
56.4 (34.5   oligohaline)
97.6
15.7
216.7 (111.1   oligohaline)
Total 6,849.4 (2,712.2   oligohaline)
Palustrine Wetlands (nontidal, except where noted)
Aquatic Bed
Emergent
Emergent (Tidal)
Mixed Emergent/Scrub-Shrub (Deciduous)
0.3
590.2 (3.4   Emergent/Forested)
119.9
1,260.6
Mixed Emergent/Scrub-Shrub (Evergreen)
Farmed
Needle-leaved Deciduous Forested
318.1
1,428.3
32.3
Evergreen Forested
Evergreen Forested (Tidal)
Scrub-Shrub/Emergent
Broad-leaved Deciduous Forested
Broad-leaved Deciduous Forested (Tidal)
3,350.0 (27.2   Atlantic White Cedar)
43.7
1,032.6
15,587.9 (76.0   w/Bald Cypress)
2,902.8 (10.5   w/Bald Cypress)
Mixed Forested
Mixed Forested (Tidal)
Deciduous Forested/Emergent
Forested/Scrub-Shrub and Forested/Scrub-Shrub
Deciduous Scrub-Shrub
12,228.6
231.8
166.1 (9.5   tidal)
5,665.0 (43.5   tidal)
856.5
Evergreen Scrub-Shrub
Mixed Scrub-Shrub
Scrub-Shrub (Tidal)
Unconsolidated Bottom/Vegetated
Unconsolidated Bottom
Unconsolidated Shore
2,475.9
1,633.5
76.7
16.4 (14.1   w/Bald Cypress)
468.4
3.2
Total 50,488.8
Riverine Wetlands
Emergent (Tidal)
Unconsolidated Shore (Tidal)
134.4
18.9
Total 153.3
GRAND TOTAL 57,491.5
tions (shoreline stabilization and coastal storm-surge
detention) lost nearly one-quarter of their pre-settle-ment
capacity. No function experienced an increase in
capacity.
DISCUSSION
Extensive wetlands have always been recognized on
the Delmarva Peninsula. Interpretation of the 1920s
soil survey data predicted that the percent of the coun-ty
represented by wetlands ranged from 32% for Car-oline
County to a high of 75% for Dorchester County
(Table 7). The latter county had extensive tidal wet-lands
bordering Chesapeake Bay and much flatwood
soil area (e.g., Elkton series). The extent of potential
wetlands in the five-county area was roughly 50%. For
the Nanticoke River watershed, pre-settlement wet-lands
were estimated to occupy 44% of the watershed.
This figure is consistent with the five-county wetland
total for the 1920s, especially considering that the
Nanticoke watershed did not have as high a percent of
tidal wetlands as the five-county area (13% vs. 21%
of the wetlands). Today, only 28% of the watershed is
wetland.
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 415
Table 4. Historic trends in the Nanticoke’s wetland area (in hectares) by landscape position, landform, and water flow path: pre-settlement
(YR 1400) vs. 1998. Codes for water flow path: BT   bidirectional-tidal; TH   throughflow; BI   bidirectional-nontidal; IS   isolated;
OU   outflow. Number of wetlands is approximate due to GIS processing.
Landscape Landform
Water Flow
Path 1400 # 1400 Area (ha) 1998 # 1998 Area (ha)
% Change
in Area
Estuarine Fringe*
Island
BT
BT
83
1
9,228.2
341.3
143
2
9,062.6
100.6
 1.8
 70.5
Total 84 9,569.5 145 9,163.2  4.2
Lentic Basin
Flat
Fringe
Island
BI
BI
BI
BI
—
—
—
—
—
—
—
—
26
8
14
4
44.4
8.7
50.0
2.0
 
 
 
 
Total 0 0 52 105.1  
Lotic River Floodplain BT
TH
102
10
2,907.3
66.5
151
6
957.2
11.3
 67.1
 83.0
Fringe
Island
BT
TH
BT
105
2
—
1,091.7
25.7
—
104
—
1
248.7
—
0.1
 77.2
—
 
Total 219 4,091.2 262 1,217.3  70.2
Lotic Stream Basin
Flat
Floodplain
Fringe
TH
TH
TH
BT
TH
BT
12
13
130
2
—
—
29.6
68.2
6,670.6
19.1
—
—
52
95
385
25
29
13
142.4
315.6
5,018.6
56.2
99.5
8.5
 381.1
 362.8
 24.8
 194.2
 
 
Total 157 6,787.5 599 5,640.8  16.9
Terrene Basin
Flat
IS
OU
IS
OU
TH
—
79
—
162
—
—
330.2
—
1,047.9
—
7
14
10
47
1
6.0
101.7
33.5
292.1
0.4
 
 69.2
 
 72.1
 
Fringe
Interfluve
OU
IS
OU
—
1723
380
—
4,616.2
66,655.3
1
1551
2120
0.4
1,989.2
38,144.3
 
 56.9
 42.8
TH 5 27.2 111 329.2  1110.3
Total 2,349 72,676.8 3,862 40,896.8  43.7
Grand Total 2,809 93,125.0 4,920 57,023.2**  38.8
* Includes tidal freshwater wetlands contiguous with estuarine wetlands and along estuarine waters.
** Excludes ponds.
General Limitations of the Study
Historic wetland data compiled from contemporary
soil surveys have obvious limitations. Translating this
information to historic wetland extent for the Nanti-coke
required making certain assumptions: 1) hydric
soil mapping units represent a reasonable approxima-tion
of historic wetlands, 2) areas of the Sunken series
were freshwater forested wetlands at pre-settlement, 3)
areas of typical freshwater wetland soils that are now
mapped as estuarine wetlands were also freshwater
forested wetlands at pre-settlement, 4) areas of Honga
series were estuarine wetlands at this time, although
they were forested wetlands at least 700 years ago
(Brewer et al. 1998), and 5) areas within non-hydric
soil map units that were mapped as forested wetlands
in 1998 by NWI represent hydric inclusions that were
forested wetlands at pre-settlement.
The 1998 database should adequately reflect current
conditions due to strengthened federal and state regu-lations
in the 1980s and 1990s. One must, however,
recognize the limitations of any wetland mapping ef-fort
derived mainly through photointerpretation tech-niques
(Tiner 1997, 1999). Photo quality, scale, and
environmental conditions at the time of acquisition are
416 WETLANDS, Volume 25, No. 2, 2005
Table 5. Comparison of preliminary functional assessment results for Nanticoke wetlands at pre-settlement versus 1998. Area (in hectares)
and percentage of the wetland area total are given for each function. Total wetland area for 1998 (57,543.2 ha) includes 520 ha of ponds.
Function
Potential
Significance
Pre-settlement Area
(% of total area) 1998 Area (% of total) % Change in Area
Surface-water Detention High
Moderate
Total
20,380.5 (21.9)
70,814.5 (76.0)
91,195.0 (97.9)
15,870.7 (27.6)
39,847.7 (69.2)
55,718.4 (96.8)
 22.1
 43.7
 38.9
Streamflow Maintenance High
Moderate
Total
72,971.2 (78.4)
546.4 (0.6)
73,517.6 (79.0)
9,586.2 (16.7)
33,332.4 (57.9)
42,918.6 (74.6)
 86.9
 600.0
 41.6
Nutrient Transformation High
Moderate
Total
39,009.7 (41.9)
54,115.5 (58.1)
93,125.2 (100.0)
14,476.2 (25.2)
40,864.3 (71.0)
55,340.5 (96.2)
 62.9
 24.5
 40.6
Retention of Sediments and Other
Particulates
High
Moderate
Total
20,380.1 (21.9)
20,365.2 (21.9)
40,745.3 (43.8)
15,627.2 (27.2)
1,920.1 (3.3)
17,547.3 (30.5)
 23.3
 90.6
 56.9
Shoreline Stabilization High
Moderate
Total
20,448.2 (22.0)
-0-
20,448.2 (22.0)
15,798.1 (27.5)
0.4 ( )
15,798.5 (27.5)
 22.7
 negligible
 22.7
Coastal Storm-surge Detention High 13,587.7 (14.6) 10,415.1 (18.1)  23.3
Fish/Shellfish Habitat High
Moderate
Shading*
Total
10,670.0 (11.5)
-0-
6,787.6 (7.3)
17,457.6 (18.8)
7,133.4 (12.4)
572.3 (1.0)
5,349.1 (9.3)
13,054.8 (22.7)
 33.1
 significant
 21.2
 25.2
Waterfowl/Waterbird Habitat High
Moderate
Wood Duck
Total
10,686.9 (11.5)
-0-
8,025.7 (8.6)
18,712.6 (20.1)
7,337.0 (12.8)
486.4 (0.8)
5,453.0 (9.5)
13,276.4 (23.1)
 31.3
 significant
 32.1
 29.1
Other Wildlife Habitat High
Moderate
Total
90,559.4 (97.2)
2,565.8 (2.8)
93,125.2 (100.0)
52,648.5 (91.5)
2,699.1 (4.7)
55,347.6 (96.2)
 41.9
 5.2
 40.6
Table 6. Predicted change in the Nanticoke watershed’s capacity to perform nine wetland functions from pre-settlement to 1998. Func-tional
units were derived from predictive values for each time period by applying a weighting scheme (2 for high; 1 for moderate; and
1 for other significant features, e.g., stream shading). The conservation of biodiversity function was not compared since original data
lacked sufficient detail for such comparison.
Function
Pre-settlement
Functional
Units
1998
Functional Units
Predicted %
of Original
Capacity Left
Predicted %
Change in
Functional Capacity
Surface-water detention 111,575.5 71,589.1 64.2  35.8
Streamflow Maintenance 146,488.8 52,504.8 35.8  64.2
Nutrient Transformation 132,134.9 69,816.7 52.8  47.2
Sediment and Other Particulate Retention 61,125.4 33,174.5 54.3  45.7
Shoreline Stabilization 40,896.4 31,596.6 77.3  22.7
Coastal Storm-surge Detention 27,175.4 20,830.2 76.7  23.3
Fish and Shellfish Habitat 28,127.6 20,188.2 71.8  28.2
Waterfowl and Waterbird Habitat 29,399.5 20,613.4 70.1  29.9
Other Wildlife Habitat 183,684.6 107,996.1 58.8  41.2
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 417
Table 7. Area (in heactares) of wetland soil mapping units in each county falling within the Nanticoke River watershed based on 1920s
soil surveys (Dunn et al. 1920, Snyder et al. 1924, Snyder and Gillett 1925, Snyder et al. 1926, and Winant and Bacon 1929). Statistics
are for entire county; percent of county represented by each mapping unit is given in parentheses.
Wetland
Mapping Unit
County
Caroline Area (%) Dorchester Area (%) Wicomico Area (%) Kent Area (%) Sussex Area (%)
Elkton
Plummer
Portsmouth
St. Johns
13,215 (16.6)
933 (1.1)
6,193 (7.5)
-0-
73,743 (49.4)
-0-
544 (0.4)
-0-
26,896 (27.4)
-0-
10,105 (10.3)
2,539 (2.6)
32,233 (21.0)
-0-
15,728 (10.2)
-0-
41,172 (16.8)
-0-
28,295 (11.7)
389 (0.1)
Coastal Beach
Meadow
Swamp
Tidal Marsh
-0-
4,172 (3.0)
-0-
1,788 (2.2)
-0-
2,047 (1.4)
-0-
35,679 (23.9)
-0-
1,788 (1.8)
2,747 (2.8)
6,141 (6.3)
285 (0.2)
3,346 (2.2)
4,327 (2.8)
18,449 (12.0)
1,710 (0.7)
1,373 (0.6)
10,701 (4.4)
14,225 (5.8)
Total 26,321 (31.8) 112,013 (75.1) 50,216 (51.2) 74,468 (48.4) 97,865 (40.1)
major limiting factors. Moreover, drier-end wetlands,
such as seasonally saturated and temporarily flooded
palustrine wetlands, are often difficult to separate from
non-wetlands on-the-ground, thereby complicating
their detection through photointerpretation.
It is important to re-emphasize that this type of func-tional
assessment is a preliminary one based on wet-land
characteristics interpreted through remote sensing
and using the best professional judgment of numerous
wetland specialists. Wetlands believed to be providing
potentially high or other significant levels of perfor-mance
for a particular function were highlighted. No
attempt was made to produce a more qualitative rank-ing
for each function or for each wetland based on
multiple functions, as this would require more input
from others and more data, well beyond the scope of
this study. Field checking of seasonally flooded and
seasonally flooded/saturated emergent wetlands should
be done to determine if they are marshes or wet mead-ows.
If the former, they will likely have high potential
as both fish and shellfish habitat and waterfowl habitat
rather than the moderate rating given in this analysis.
The functional assessment used (W-PAWF) does
not consider the condition of the adjacent upland (e.g.,
level of disturbance) or the actual water quality of the
associated waterbody, which may be regarded as im-portant
metrics for assessing the health of individual
wetlands (not part of this study). Collection and anal-ysis
of some of these data were done in related studies
(Tiner et al. 2000, 2001, Tiner 2004) and were not part
of the present study.
Appropriate Use of this Type of Analysis
Keeping in mind the limitations mentioned above,
this analysis is a first-cut or initial screening of the
watershed’s wetlands and an assessment of the poten-tial
impact of cumulative losses on wetland functions.
It highlights wetlands that may have a significant po-tential
to perform each of ten functions. While the
analysis provides perspective on the ability of the wa-tershed’s
wetlands to perform these functions, it does
not evaluate differences among wetlands of similar
type and function. The latter information is often im-portant
for making decisions about wetland acquisition
and designating certain wetlands as more worthy of
preservation versus others with the same categoriza-tion.
Such information can be collected through field
investigations and/or by consulting agencies having
specific expertise in a subject area.
The analysis for the Nanticoke watershed is a wa-tershed-
based wetland characterization and a historical
assessment of changes in wetland extent and function.
It can serve as an initial screening for prioritization of
wetlands for acquisition, restoration, or strengthened
protection, as an educational tool for improving the
public’s understanding of wetland functions and
trends, and as a baseline assessment of how wetlands
and functions have changed since pre-settlement. For
more than two decades, NWI maps have been used by
local governments in compiling natural resource in-ventories.
Now, by enhancing NWI data and using it
for wetland functional assessment, local planners have
a valuable tool for preparing ecologically based mu-nicipal
master plans (Honachefsky 1999).
CONCLUSIONS
Wetlands in the Nanticoke River watershed have
undergone significant changes since pre-settlement.
Prior to European colonization, about 45% of the wa-tershed
(roughly 93,000 ha) was wetland, with exten-sive
headwater wetlands supporting streamflow. By
1998, about 57,000 ha of wetlands (62% of the orig-inal
area) remained and much of this area has been
ditched, excavated, or impounded. Conversion of wet-
418 WETLANDS, Volume 25, No. 2, 2005
lands to agricultural lands was the predominant cause
of freshwater wetland change; sea-level rise was the
main agent of estuarine wetland change.
Cumulative wetland losses have led to significant
reductions of many wetland functions. Since colonial
times, it was estimated that the Nanticoke watershed
lost over 60% of its predicted capacity for streamflow
maintenance and over one-third of its capacity for four
other functions: surface-water detention, nutrient trans-formation,
sediment and other particulate retention,
and provision of other wildlife habitat. No function
experienced an increase in capacity.
The findings of this study provide an overview of
the predicted changes in wetland extent and function
for the Nanticoke River watershed since European set-tlement.
The comparison of changes in wetland func-tion
watershed-wide should be considered approximate
due to the nature of this type of analysis. As with any
remotely-sensed analysis, field checking should be
conducted to validate the interpretations regarding
functions of individual wetlands, since this type of as-sessment
is a coarse-filter approach. Despite these lim-itations,
the assessment serves as a foundation for un-derstanding
the extent to which wetlands have changed
in general form and function, and as such, it provides
a valuable tool for resource planning. It should be used
with other tools to help devise a watershed-wide strat-egy
for wetland conservation and restoration.
This pilot study demonstrated that it is possible to
produce historic assessments of wetlands and functions
through analysis of existing information and enhance-ment
of NWI data. Depending on the nature of wetland
development and the information available, many as-sumptions
have to be made. Nonetheless, this ap-proach
provides a consistent method for evaluating
wetland status and trends from a functional perspective
while helping increase our understanding of how much
historic wetland losses have impacted a watershed’s
ability to perform numerous functions.
The NWI Program in the Northeast plans to add
LLWW descriptors to the NWI digital database as
maps are updated. This will increase the value of the
NWI database and facilitate its use for preparing pre-liminary
watershed assessments of wetland functions
throughout the region. This type of assessment will
also be incorporated into localized wetland trends
studies to demonstrate how wetland losses are im-pacting
specific functions. There is also interest in this
applying these procedures to other regions. Such work
will require review of the wetland function-character-istic
correlations; minor modifications will undoubt-edly
be needed to address regional differences in fish
and wildlife habitat.
ACKNOWLEDGMENTS
This study was funded by the Kent Conservation
District and the Maryland Eastern Shore Resource
Conservation and Development Council. Herbert
Bergquist (FWS) constructed the digital database for
historic wetlands, performed GIS analyses to produce
data for this report, and prepared Figure 1. Bobbi Jo
McClain assisted in digital database construction dur-ing
the early phase of this work. Correlations between
wetland characteristics and wetland functions used to
produce the preliminary assessment of wetland func-tions
were prepared jointly by the FWS, wetland spe-cialists
from Maryland and Delaware, and other sci-entists.
Amy Jacobs (DNREC) and the Nanticoke wet-land
group she assembled reviewed the draft protocols
for correlating wetland characteristics with wetland
functions and provided recommendations to modify
the selection criteria. Participants included David
Bleil, Katheleen Freeman, Cathy Wazniak, Mitch Keil-er,
and Bill Jenkins (Maryland Department of Natural
Resource); Julie LaBranche (Maryland Department of
the Environment); Marcia Snyder, Dennis Whigham,
and Don Weller (Smithsonian Environmental Research
Center); Matt Perry and Jon Willow (U.S. Geological
Survey); Mark Biddle (DNREC); and Peter Bowman
(Delaware Natural Heritage Program). Abby Rokosch
(DNREC) provided copies of the texts of 1920s soil
survey reports for Kent and Sussex Counties and Cur-tis
Larsen (U.S. Geological Survey) contributed infor-mation
on sea-level rise for Chesapeake Bay.
LITERATURE CITED
Brewer, J. E., G. P. Demas, and D. Holbrook. 1998. Soil survey of
Dorchester County, Maryland. U.S.D.A. Natural Resources Con-servation
Service, Washington, DC, USA.
Cowardin, L. M., V. Carter, F. C. Golet, and E. T. LaRoe. 1979.
Classification of wetlands and deepwater habitats of the United
States. U.S. Fish and Wildlife Service, Washington, DC, USA.
FWS/OBS-79/31.
Dahl, T. E. 2000. Status and trends of wetlands in the conterminous
United States 1986 to 1997. U.S. Fish and Wildlife Service, Wash-ington,
DC, USA.
Dahl, T. E. and C. E. Johnson. 1991. Status and trends in the con-terminous
United States, mid-1970’s to mid-1980’s. U.S. Fish and
Wildlife Service, Washington, DC, USA.
Dunn, J. E., J. M. Snyder, and E. Hoffecker. 1920. Soil survey of
Kent County, Delaware. U.S. Department of Agriculture. Govern-ment
Printing Office, Washington, DC, USA.
Frayer, W. E., T. J. Monahan, D. C. Bowden, and F. A. Graybill.
1983. Status and trends of wetlands and deepwater habitats in the
conterminous United States, 1950’s to 1970’s. Colorado State
University, Fort Collins, CO, USA.
Hall, R. L. 1970. Soil survey Wicomico County, Maryland.
U.S.D.A. Soil Conservation Service, Washington, DC, USA.
Hefner, J. M., B. O. Wilen, T. E. Dahl, and W. E. Frayer. 1994.
Southeast wetlands: status and trends, mid-1970’s to mid-1980’s.
U.S. Fish and Wildlife Service, Region 4, Atlanta, GA, USA.
Honachefsky, W. B. 1999. Ecologically Based Municipal Land Use
Planning. Lewis Publishers, CRC Press, Boca Raton, FL, USA.
Larsen, C. E. 1998. The Chesapeake Bay: geological product of
Tiner, CUMULATIVE LOSS OF WETLAND FUNCTIONS—NANTICOKE RIVER WATERSHED 419
rising sea level. U.S. Geological Survey, Reston, VA. Fact sheet
102–98.
Matthews, E. D. 1964. Soil survey Caroline County, Maryland.
U.S.D.A. Soil Conservation Service, Washington, DC, USA.
Matthews, E. D. and W. Ireland, Jr. 1971. Soil survey Kent County,
Delaware. U.S.D.A. Soil Conservation Service, Washington, DC,
USA.
Snyder, J. M., J. H. Barton, J. E. Dunn, J. Gum, and W. A. Gum.
1924. Soil survey of Sussex County, Delaware. U.S. Department
of Agriculture, Bureau of Soils. Government Printing Office,
Washington, DC, USA.
Snyder, J. M. and R. L. Gillett. 1925. Soil survey of Wicomico
County, Maryland. U.S. Department of Agriculture, Bureau of
Soils. Government Printing Office, Washington, DC, USA.
Snyder, J. M., W. C. Jester, and O. C. Bruce. 1926. Soil survey of
Dorchester County, Maryland. U.S. Department of Agriculture,
Bureau of Soils. Government Printing Office, Washington, DC,
USA.
Tiner, R. W. 1997. NWI maps: what they tell us. National Wetlands
Newsletter 19(2):7–12.
Tiner, R. W. 1999. Wetland Indicators: A Guide to Wetland Iden-tification,
Delineation, Classification, and Mapping. Lewis Pub-lishers,
CRC Press, Boca Raton, FL, USA.
Tiner, R. W. 2002. Enhancing wetlands inventory data for water-shed-
based wetland characterizations and preliminary assessment
of wetland functions. p. 17–39. In R. Tiner (compiler) Watershed-based
Wetland Planning and Evaluation: a Collection of Papers
from the Wetland Millennium Event (August 6–12, 2000; Quebec
City, Quebec, Canada). Association of State Wetland Managers,
Inc., Berne, NY, USA. Available online at: http://www.aswm.org.
Tiner, R. W. 2003a. Dichotomous keys and mapping codes for wet-land
landscape position, landform, water flow path, and waterbody
type descriptors. U.S. Fish and Wildlife Service, National Wet-lands
Inventory Program, Northeast Region, Hadley, MA, USA.
Tiner, R. W. 2003b. Correlating enhanced National Wetlands In-ventory
data with wetland functions for watershed assessments: a
rationale for northeastern U.S. wetlands. U.S. Fish and Wildlife
Service, Northeast Region, Hadley, MA, USA.
Tiner, R. W. 2004. Remotely-sensed indicators for monitoring the
general condition of ‘‘natural habitat’’ in watersheds: an appli-cation
for Delaware’s Nanticoke River watershed. Ecological In-dicators
4:227–243.
Tiner, R. W. and H. C. Bergquist. 2003. Historical analysis of wet-lands
and their functions for the Nanticoke River watershed: a
comparison between pre-settlement and 1998 conditions. U.S.
Fish & Wildlife Service, Northeast Region, Hadley, MA, USA..
National Wetlands Inventory technical report.
Tiner, R. W., H. C. Bergquist, and B. J. McClain. 2002. Wetland
characteristics and preliminary assessment of wetland functions
for the Neversink Reservoir and Cannonsville Reservoir basins of
the New York City water supply watershed. U.S. Fish and Wild-life
Service, Northeast Region, Hadley, MA, USA. National Wet-lands
Inventory report.
Tiner, R. W., H. C. Bergquist, J. Q. Swords, and B. J. McClain.
2001. Watershed-based wetland characterization for Delaware’s
Nanticoke River watershed: a preliminary assessment report. U.S.
Fish and Wildlife Service, Northeast Region, Hadley, MA, USA.
National Wetlands Inventory report.
Tiner, R. W., and G. DeAlessio. 2002. Wetlands of Pennsylvania’s
coastal zone: wetland status, preliminary functional assessment,
and recent trends (1986–1999). U.S. Fish and Wildlife Service,
Northeast Region, Hadley, MA, USA. National Wetlands Inven-tory
report.
Tiner, R. W. and J. T. Finn. 1986. Status and recent trends of wet-lands
in five mid-Atlantic states: Delaware, Maryland, Pennsyl-vania,
Virginia, and West Virginia. U.S. Fish and Wildlife Ser-vice,
Newton Corner, MA and U.S. Environmental Protection
Agency, Region III, Philadelphia, PA, USA. Cooperative technical
report.
Tiner, R. W., I. Kenenski, T. Nuerminger, J. Eaton, D. B. Foulis,
G. S. Smith, and W. E. Frayer. 1994. Recent wetland status and
trends in the Chesapeake watershed (1982 to 1989). U.S. Envi-ronmental
Protection Agency, Annapolis, MD, USA. Chesapeake
Bay Program Technical Report.
Tiner, R. W., C. W. Polzen, and B. J. McClain. 2004. Wetland
characterization and preliminary assessment of wetland functions
for the Croton watershed of the New York City water supply
watershed. U.S. Fish and Wildlife Service, Northeast Region,
Hadley, MA, USA. National Wetlands Inventory report.
Tiner, R., S. Schaller, D. Petersen, K. Snider, K. Ruhlman, and J.
Swords. 1999. Wetland characterization study and preliminary as-sessment
of wetland functions for the Casco Bay watershed,
southern Maine. U.S. Fish & Wildlife Service, Northeast Region,
Hadley, MA, USA. National Wetlands Inventory technical report.
Tiner, R., M. Starr, H. Bergquist, and J. Swords. 2000. Watershed-based
wetland characterization for Maryland’s Nanticoke River
and Coastal Bays watersheds: a preliminary assessment. U.S. Fish
and Wildlife Service, Northeast Region, Hadley, MA, USA. Na-tional
Wetlands Inventory report.
Tiner, R. W. and J. Stewart. 2004. Wetland characterization and
preliminary assessment of wetland functions for the Delaware and
Catskill watersheds of the New York City water supply system.
U.S. Fish and Wildlife Service, Northeast Region, Hadley, MA,
USA. National Wetlands Inventory report.
Winant, H. B. and S. R. Bacon. 1929. Soil survey of Caroline Coun-ty,
Maryland. U.S. Department of Agriculture, Bureau of Chem-istry
and Soils. Government Printing Office, Washington, DC,
USA.
Manuscript received 2 February 2004; revisions received 27 January
2005; accepted 14 February 2005.