Daniel E. Kroes. OCCURENCE OF JURISDICTIONAL WETLANDS ON
RIVERINE FLOODPLAINS ALONG A CLIMATIC GRADIENT. (Under the
direction ofDr. Mark M. Brinson) Department ofBiology, February 2001.
The purpose of this study is to examine the relationship between the occurrence of
jurisdictional wetlands in floodplains and the water balance of climate along a humid to
arid continuum. Two regions were studied. The first included 36 mid-reach streams
encompassing an area from Cope, Colorado to Chariton, Iowa. This region was chosen
because of the broad range in potential évapotranspiration (PET) ratio, from 0.70 to 1.75.
The second included 16 headwater streams in eastern North Carolina with PET ratios
ranging from 0.83 to 0.67. PET ratios were estimated using Holdridge’s life zone
formula based on precipitation and temperature data at nearby sites. Wetland boundaries
were determined using field delineation along transects perpendicular to the direction of
stream flow. The width ofjurisdictional wetlands was compared with flood-prone width
(FPW) and expressed as a percent. Streams with a PET ratio greater than 0.98 did not
have wetlands associated with them. PET ratio was the most important factor in wetland
and vegetation occurrence. Soil texture, duration of overbank flow, and stream order did
not correlate with percentage ofFPW that was wetland. An increase in PET ratio
resulted in an exponential decrease in the percentage of the FPW that is wetland. Greater
channel cross-sectional areas correlated with greater wetland widths in both study
regions. As the precipitation in the central plains decreased, the number of prevalent
woody species decreased. In North Carolina the number of prevalent woody species
decreased as the PET ratio decreased. Hydrologic sources differ between the central
plain floodplains and those of the North Carolina floodplains. North Carolina coastal
plain riparian wetlands are formed and maintained by groundwater. Increasing stream
order corresponds to wetter conditions on the floodplain. In the central plains, the
exponential trend in the percentage of the FPW that is wetland indicates positive
feedback between groundwater and streamflow. Wetland status of riparian floodplains
the central plains relies upon this feedback.
OCCURRENCE OF JURISDICTIONALWETLANDS ON RIVERINE
FLOODPLAINS ALONG A CLIMATIC GRADIENT
A Thesis
Presented to
the Faculty of the Department ofBiology
East Carolina University
In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Biology
by
Daniel E. Kroes
February 2001
OCCURRENCE OF JURISDICTIONAL WETLANDS ON RIVERINE
FLOODPLAINS ALONG A CLIMATIC GRADIENT
by
Daniel Kroes
APPROVED BY: ^
DIRECTOR OF THESIS
Mark M. Brinson, Ph. D.
^
COMMITTEE MEMBER
^
Lisa M. Clough, Ph. D.
COMMITTEE MEMBER
Richard D. Rheinhardt, Ph. D.
COMMITTEE MEMBER ' KfrAa^i^
Richard K. S .. D.
CHAIR OF THE DEPARTMENT OF BIOLOGY
onald J. Newton, Ph. D.
DEAN OF THE GRADUATE SCHOOL
Thomas L. Feldbush, Ph. D.
Acknowledgements
This project has been a very challenging undertaking that has also been
exceedingly rewarding. Streams, their functions and biota, have been a main focus ofmy
interest for as long as I can remember. First and foremost I need to thank my brothers;
My older brother, Calvin, for introducing me to stream habitats, and my younger brother,
Greg, for his assistance in the field evaluation ofmy sites, observational skills, ability to
keep morale high, and most importantly, his ability to voice his opinions. I owe Mark
Brinson a dept of gratitude for saving me from the hell my life would have become iff
had decided to go to medical school. Mark gave me an option that ended up being one of
the most rewarding experiences ofmy life. My parents, Roger and Margaret, assisted in
making me who I am, providing me with an intellectual down to earth upbringing, and
sometimes, more time outdoors than I wanted, especially during duck season. My
committee members, Lisa Clough, Rick Rheinhardt, and Richard Spruill provided me
with much needed viewpoints and manuscript reviews.
I am grateful to the individuals, fiiends and colleagues, who made my college life
enjoyable. I am also grateful to Katie Haven, Chris Jones, Mark Keusenkothen, Mindi
May, and Carol Phillips for the endless reviews ofmy manuscript.
This research was supported by Roger and Margaret Kroes and by the East
Carolina University Department of Biology.
Table of Contents
List ofTables v
List of Figures vi
Introduction 1
Geomorphic Features and Their Relationship to Discharge 6
Wetlands on Riverine Floodplains 10
Proposed Relationship 14
Methods 18
Study Sites 18
Determination ofPET Ratio 23
Determination of Bankfull Depth 24
Delineation ofWetlands 27
Field Analysis ofPercentage ofWetlands 28
Determination ofOverbank and Above Average Flow Duration 29
Hydrograph Separation 29
Map Analysis 30
Analysis ofData 30
Results 32
PET Ratio Effects 32
Stream Order 40
Drainage Basin Area 44
Soil 49
IV
Overbank and Above Average Flow Duration 49
Groundwater Contribution to Discharge 51
Discussion and Conclusions 56
Maintenance ofWetlands 56
Vegetation 64
Conclusions 66
References 70
Appendix A. Floodplain Vegetation on the Central Plain Study Sites 73
Appendix B. Floodplain Vegetation on the North Carolina Study Sites 83
List of Tables
1. Study site name, location, PET ratio, USGS gaging station, and National
Climatic Data Center weather station for the central plain study sites 20
2. Study site name, location, PET ratio and National Climatic Data Center
weather station for the North Carolina study sites 22
3. Factors influencing the occurrence ofvegetation at the central plain study sites 35
4. Wetland widths, flood-prone widths and percentage ofwidths that is wetland
for the central plain and North Carolina sites 37
5. The factors influencing the occurrence of vegetation at the North Carolina
study sites 41
6. Primary and secondary wetland indicators for the central plain and North
Carolina sites 42
7. The geomorphic parameters of the central plain and North Carolina sites 46
8. The possible factors influencing the average annual bankfull flow duration
for the central plain sites 53
List of Figures
1. The vegetation regions of the United States as compared with the PET ratio 4
2. Mean annual runoff as compared with the PET ratio 5
3. Bankfull discharge of a stream 7
4. The process of entrenchment 9
5. Cross-sections of streams contrasting mechanisms of floodplain hydrologic
regimes 11
6. An excerpt from the Old Ford, North Carolina NWI quadrangle 13
7. A dimensionless, conceptual model showing the hypothetical relationship
between the PET ratio and the percent of the FPW that is wetland 16
8. Study sites in the central plain region 19
9. North Carolina study sites 21
10. Bankfull depth 25
11. The relationship between the PET ratio and four variables 33
12. The relationship between precipitation and five variables 34
13. Two factors possibly influencing North Carolina floodplains 39
14. The relationship between stream order and five variables 43
15. The relationship between channel cross-sectional area and five variables 45
16. The relationship between drainage basin area and three variables 48
17. The relationship between soil texture and four variables 50
18. The relationship between average annual overbank flow duration and four
variables 52
19. The relationship between the duration of flow exceeding average discharge
and five variables 54
vii
20. Hydrographs for selected sites from a synchronous time period 55
21. Distribution of study sites 57
22. The exponential decrease in the percentage of the FPW that is wetland
as compared with the conceptual model shows a similar relationship 58
23. The cyclical pattern of rain, high storm flow, and groundwater discharge into
the stream 61
24. Cross-sections of streams contrasting hydrologic regimes of floodplains 62
Introduction
It is important for us to expand our knowledge of the factors that control and
maintain riverine wetlands in order for us to more fully understand the impact of our
activities upon them. The climatic control of riverine wetlands has largely been
overlooked, particularly in subhumid regions. The vast majority of floodplain wetland
studies have been conducted either on streams located in humid regions that are fed by
local sources or in arid regions receiving flows from the mountains (Auble et al. 1994,
Shaffoth et al. 1998). The factors (overbank flow duration, stream order, drainage basin
area, and soil) believed to control riverine wetland development and maintenance have
been conceived at the extremes of the climatic gradient without considering the
intermediate subhumid to humid condition. Most of these factors have not been studied
along a climatic gradient from regions where the streambed itself lacks the hydrologic
regime of a wetland, to regions where the majority of the floodplain is wetland. This
study examines the occurrence ofwetlands along a humid to subhumid gradient.
Along any latitudinal cross section of the United States, there is a distinct change
in vegetation and runoff caused by the interaction ofprecipitation and évapotranspiration
(ET). From the naturally forested Atlantic Coast, the climatic potential to support forests
on uplands continues westward until approximately 95° to 97° west longitude. Between
95° and 97° west longitude natural upland vegetation undergoes a gradual change from
upland forest to tall grass prairie due to a reduction in precipitation (Kuchler 1967). The
zone of change from forest to tall grass prairie has been historically modified and
maintained by fire. Wet and dry periods result in decreased and increased frequency of
2
fire causing the forest / tall grass prairie transition zone to move west and east.
Gallery forests along high order streams in the central plain states ofNorth
America represent the western-most extent of eastern deciduous forests (Abrams 1985,
Rice 1965). Gallery forests develop in the parts of a drainage basin where fire becomes
infrequent due to increased water supplies (Abrams and Gibson 1991). Low order stream
channels in subhumid grasslands do not originate under forested canopies. In prairie, the
support of canopy trees along higher-ordered streams is associated with the decreased fire
frequency in the riparian corridor (Reichman 1987).
Just east of the Rocky Mountains, tall grass prairie is replaced by short grass
prairie. Arid and semiarid regions generally occur at lower elevations among mountain
ranges separated by higher altitude montane forests, although there are many exceptions
to this generalized vegetation pattern. Exceptions include more moist climates of higher
altitudes and windward sides of the Rockies and Sierras, along the Pacific Coast, and the
Pacific Northwest (Kuchler 1967, USDC 1968, Bailey 1976). The longitudinal pattern of
vegetation described above is greatly influenced by the relationship between actual ET
and precipitation (Kuchler 1967, Holdridge et al. 1971).
Actual ET is the amount ofwater that is evaporated and transpired from all
surfaces (plant, soil, water, etc.). Potential évapotranspiration (PET) is the maximum
amount ofwater that could potentially be evaporated and transpired from plants and
surfaces at a given temperature and / or degree of solar radiation regime if there were no
limitation in water supply. The relationship between PET and precipitation can be best
3
visualized utilizing the PET ratio (PET divided by precipitation). A PET ratio of 1.0
indicates a balance between PET and precipitation. Subhumid climates have a PET ratio
greater than 1.0, grading to arid as the ratio increases. Vegetation reflects this change by
a species change toward increasing tolerance of dry conditions. Humid climates have a
PET ratio less than 1.0, grading to superhumid as the ratio decreases with vegetation
species changing towards a decreasing tolerance of dry conditions (Eigure 1). Likewise,
flow in streams can be linked to climate (rainfall and ET) assuming that there are no
overriding effects caused by geomorphology or geology (Figure 2) (Riggs and Harvey
1990, Mitsch and Gosselink 1993). In order to evaluate the relationship between
precipitation and streamflow, consider a hypothetical watershed that is homogenous in
precipitation and ET. The soils of the watershed act as a reservoir for storage ofwater
and subsequent release as ET, and depending on certain conditions, may discharge
groundwater to the stream.
Augmentation or reduction ofbaseflow (the discharge in a stream not related to
storm activity) or the duration of overbank flow by any of several conditions can cause
streamflow to depart from the ideal, hypothetical, climatically controlled condition. This
departure may allow riverine wetlands to be maintained in a climate or hydrologic setting
where they might not otherwise exist, or wetlands might not exist in areas where they
could normally. Conditions that may cause departure from normal include watersheds
with low permeability surfaces and supplementation of streamflow. Watersheds with low
permeability materials have a higher percentage of runoff resulting in lessened
groundwater recharge and ET (Riggs and Harvey 1990). Watersheds with areas of low
4
Mediterranean, western arid, and semi- Northern forest
arid
Pacific Northwest and Rocky Mountain Eastern deciduous forest
Plains grasslands Southern forest
Central forest I
40° N
124° W 97° 30” W 75° W
Figure 1. The vegetation regions of the United States as compared with
the PET ratio along 40° N latitude, modified from Bailey (1976). PET
ratios remain fairly constant between the east coast and about 97° W until
precipitation begins to decrease in the rain shadow of the Rockies. The
western limit of upland forest coincides with the PET ratio of 1.0.
Mean annual runoff, in millimeters
jgg 0-10 ^.\w, 300 - 1000-'¿¿V.
^ 600- 1000
> 1000
40° N
a 1 Cone. CO
u
H Region of studyA\ \
Cl, / \ T —/^^,,^;-^Promise City, lA
0
124° W 97° 30”W 75° W
Figure 2. Mean annual runoff as compared with the PET ratio, modified
from Riggs and Harvey (1990). Runoff decreases with increasing PET ratios.
When compared with Figure 1, the transition from forested to grass
dominated uplands occurs at slightly less than 100 mm of runoff per year.
6
permeability tend to have increased magnitude of flood discharge and shorter durations of
stormflows when compared to watersheds with highly permeable soils. An example of
low permeability watersheds can be seen along streams in areas with clay soils, such as
the Brazos River, Texas (NWS 1995).
Deep aquifer discharge can supplement streamflows through springs.
Supplemented streamflow can also be found at any stream receiving significant return
flow from irrigation. If a stream has parts of its drainage basin in the mountains, glacial
melting and montane snowmelt can represent water sources originating from a different
physiographic region. Water from these sources increases baseflow and the duration of
overbank flow beyond what downstream segments could otherwise support.
Supplementation of flow may support areas otherwise not hydrologically suited for
riverine wetlands. The Platte River, receiving discharge from the RockyMountains,
supports riverine wetlands in regions where few would be without supplemented flow
(Johnson 1998, USFWS 1998a).
Geomorphic Features and Their Relationship to Discharge
Bankfull discharge is the maximum water transporting capacity of a stream
channel, beyond which inundation of some portion of the flood-prone area (FPA) occurs
(Figure 3). The FPA is the area contained within twice the vertical distance from the
deepest point of the channel (thalweg) to the top of the stream bank (bankfull). On
average, this inundation of the FPA occurs at an interval of 1.5 years in the United
States, with a variable duration (Leopold et al. 1964).
The return interval of bankfull stage is influenced by many factors. For some
1 Flood-prone width
X Bankfull depth
Bankfull depth
Bankfull stage Thalweg
Figure 3. Bankfull discharge of a stream is the discharge required to fill the active channel. Discharges
greater than bankfull result in inundation of some portion of the floodplain.
8
streams in arid regions, the return interval may be greater than 30 years due to extreme
variability in precipitation events (Leopold et al. 1964, Williams 1978, Leopold 1994,
Rosgen 1996). Often, a longer return interval of bankfull stage is a result of a
disequilibrium in sediment load or discharge resulting from an entrenchment of the
stream into the alluvium (Leopold 1994, Burke and Nutter 1995). For example, as soil is
eroded from uplands it may become deposited on floodplains and streambeds resulting in
a higher relative elevation for both the floodplain and streambed, initially resulting in
greater and more frequent flooding (Trimble 1970). When the upland soils are stabilized,
the output of sediment from the stream system exceeds the input. This disequilibrium
causes the stream to erode the streambed first and the floodplain later. Until a new
floodplain is formed at a level supported by floodplain forming discharges, there is an
abnormally deep stream channel that requires a greater discharge to reach bankfull stage.
This abnormally deep channel results in longer return intervals and greater elevational
differences between the water surface and floodplain surface (Figure 4). The reduced
duration and frequency of flooding and elevational differences may negatively impact
wetlands as the hydroperiod (duration of saturation) of the floodplain is reduced.
The duration of overbank flow is influenced bymany factors including drainage
basin area and configuration. Drainage basin area greatly influences the magnitude and
duration of flooding (Leopold 1994). Larger drainage basins tend to have greater flood
duration and magnitude than smaller drainage basins with the same overbank return
interval. Somewhat related to drainage basin area is stream order. High order streams in
a drainage basin generally have greater flow year round than lower order streams with
sediment sources
Deposited sediment on the
floodplain and stream bed
Stabilized u^pstream soils -I,nci?si?on
(d) (c)
Figure 4. The process of entrenchment can be triggered in many ways, one ofwhich is destabilization of
uplands, (a) An undisturbed stream and floodplain, (b) When upland soils are destabilized, the rate of
sedimentation is increased into the stream system resulting in an increased base level and flooding, (c)
When uplands are stabilized, or the rate of sediment output is decreased, the stream erodes the bed. This
results in an abnormally deep bankfull stage (entrenchment), (d) Over time the stream will form a new
floodplain, leaving the old floodplain as a terrace (after Leopold 1994).
VO
10
relatively comparable basin characteristics. The gaining (groundwater contribution to
streamflow) or losing (streamflow contribution to groundwater) nature of the stream may
also influence flood duration. Losing streams tend to have a shorter duration of flooding
than gaining streams due to the loss of stream discharge to the groundwater system
(Figure 5) (Fetter 1980, Heath 1983). These factors (drainage basin area, stream order,
and the gaining / losing nature of the stream) may have an influence on the formation and
maintenance of floodplain wetlands.
Wetlands on Riverine Floodplains
Fluvial processes create floodplains that are transitional between uplands and the
stream. Floodplains tend to be wetter than the surrounding uplands due to their lower
geomorphic position and their proximity to the water table and stream channel. Wetland
portions of floodplains are regulated under the Clean Water Act as waters of the United
States. However, portions ofmany floodplains do not always fall into the jurisdictional
wetland (hereafter referred to as wetlands) category as defined by the US Army Corp of
Engineers (USAGE) (Environmental Laboratory 1987).
Riverine wetlands differ from other types ofwetlands in that they occur in
flood-prone areas (FPA) and riparian corridors associated with stream channels (Brinson
1993a). Because of the association of riverine wetlands with stream channels, the
throughput and transformations of nutrients, sediments, and water are generally much
greater than most other wetland types. The dominant sources ofwater for riverine
wetlands are overbank flow and groundwater discharge. Additional water sources may
include precipitation, interflow (unsaturated groundwater flow), and overland flow from
Gaining stream Losing stream (b)
Figure 5. Cross-sections of streams contrasting mechanisms of floodplain hydrologic regimes, (a) In regions where
groundwater levels are high, streamflow is maintained by groundwater (gaining), (b) In regions where groundwater
levels are low, streamflow is lost to the groundwater table (losing). Vertical arrows depict the range in water table
levels. ^ = water table level.
12
adjacent uplands (Brinson 1993b). The hydrologic regime of floodplains are partially
determined by the period of time at which the discharge of the stream exceeds bankfull,
with flooding ranging from being nearly permanent to intermittently inundated or
saturated (Larson et al. 1981). In arid climates, overbank flow and groundwater recharge
from the stream channel may support wetlands in FPAs along perennial streams. Rare,
perennial streams in arid regions are usually fed by water from springs or from areas
outside the arid area (mountains, humid regions, etc.) (Agnew and Anderson 1992).
The National Wetlands Inventory (NWI) maps (Figure 6) can be used to remotely
estimate the extent and area of riverine wetlands at study sites similar to the approach
used by Cashin (1990) and Stolt and Baker (1995). However, NWI maps may not be
accurate due to inaccuracies created during the mapping process of remotely sensed data
and the inability to distinguish small wetland areas (Stolt and Baker 1995). Riverine
wetlands are especially subject to this inaccuracy due to their narrow width.
Wetlands are jurisdictionally delineated by evaluating vegetation, soil, and
hydrology. To be considered hydrophytic, vegetation in wetlands must be adapted to
saturated or inundated conditions (Tiner 1991). The indicator status of hydrophytes
adapted to these conditions is listed in regional publications of the US Fish and Wildlife
Service (Reed 1988). Hydric soils develop under conditions where soil oxygen is
depleted in response to saturated conditions over long periods of time during the growing
season. In order for an area to meet the hydrologic requirement for a wetland, soil must
be continuously saturated within 30 cm of the surface for 14 days of the growing season,
for greater than 50 percent of years (Environmental Laboratory 1987).
13
Figure 6. An excerpt from the Old Ford, North Carolina NWI quadrangle in
Arcdata format as viewed through ArcView 3.0 featuring Big Swamp and
tributaries arcing across the excerpt. Different wetland types are represented by
different shades and colors. In this figure upland is represented by the shade
covering the majority of the figure. U = uplands, PFOlA = palustrine forest,
broad-leaved deciduous, temporarily flooded, PFOIC = palustrine forest, broad-
leaved deciduous, seasonally flooded, PF01/4B = palustrine forest, broad-
leaved deciduous / needle-leaved evergreen, saturated, PF04/1B = palustrine
forest, needle-leaved evergreen / broad-leaved deciduous, saturated, PSSlA =
palustrine scrub-shrub, broad-leaved deciduous, temporarily flooded (USFWS
1998b).
14
Due to the logistical problems and cost associated with continuous measurement
of the water table, hydric soil indicators, rather than water table measurements, are
usually used to indicate hydroperiod. The depth to the water table and its associated
capillary fringe along intermittent and ephemeral streams prevent the formation of hydric
soil indicators within 30 cm of the soil surface. In such cases the upper 30 cm of soil is
seldom saturated. This aerated condition prevents the accumulation of organic carbon and
the corresponding development and maintenance of low redox potentials (Gambrell and
Patrick 1978). During overbank flow events in arid climates, duration of inundation may
be insufficient to maintain anoxic conditions for a period of time sufficient to produce
gleying, mottling, and a buildup of organic material in the soil, thus, preventing the
formation and maintenance ofjurisdictional wetlands (Diers and Anderson 1984).
Proposed Relationship
In the United States, along a gradient from the humid east to the arid west,
precipitation decreases and the PET remains relatively constant (within a range of 200
mm, except in montane regions) along any one latitude. Therefore, within any defined
drainage basin size, the prevalence ofwetlands in floodplains (measured as a percent of
FPA) should gradually decrease with increasing PET ratio, if all other factors remain
constant.
I hypothesize that there is an ecotone where several factors (PET ratio, flood
duration, groundwater discharge, and others) change and converge to cause riverine
wetlands to be a significant percent of the FPA. I expect this percentage change to be
sharp as the PET ratio decreases because: (1) precipitation not used by ET contributes to
15
soil moisture in uplands leading to a groundwater contribution to the FPA, and (2)
groundwater input not used by ET in the FPA results in groundwater contribution to the
stream. Variable source area (Hewlett 1961) input increases downstream baseflow and
the duration of overbank flow creating a positive feedback effect as the water table rises
in the FPA. The combination of these factors may interact exponentially with decreases
in the PET ratio until the entire FPA is wetland, which may extend beyond the FPA
(Figure 7). This feedback hypothesis would be invalid if the correlation ofPET ratio
with the percent of the FPW that is wetland does not show an exponential curve. Some
exceptions to the feedback hypothesis might occur in situations where conditions other
than overbank flow and normal groundwater discharges dominate the FPA.
In areas where the PET ratio exceeds 1.0, a moisture deficit occurs. This deficit
affects climatically controlled streams through a reduction in the duration of flooding and
the amount of soil moisture. There is probably a threshold value of overbank flood
duration in riparian areas that must occur for wetlands to be maintained in these areas. In
the humid east, riverine wetlands are partially maintained through saturation near the soil
surface from a high water table and capillary fringe (Patterson et al. 1985, Cole et al.
1997). The relative contributions ofwater sources in the humid climates of the east differ
from those of the subhumid central plains. Runoff decreases in the central plains causing
streams and floodplains to rely on upstream sources ofwater in the form of overbank
flow if they are to maintain wetland status (Riggs and Harvey 1990).
At the longitude where PET exceeds precipitation (PET ratio > 1) (97° W), the
zone of soil saturation along many streams may seldom meet the jurisdictional
16
Ifmultiple factors interact to result in a
Flood Hypothetical Groundwater
duration wetland discharge
response
Figure 7. A dimensionless, conceptual model showing the hypothetical
relationship between the PET ratio and the percent of the FPW that is wetland. If
the PET ratio increased at a uniform rate from wet to dry, then the percentage of
the FPW that is wetland would decrease as the PET ratio increased to 1.0,
assuming that no other factors are influencing the occurrence of riverine
wetlands. Ifmultiple factors interact in a positive feedback marmer, then
wetlands would decrease exponentially as the PET ratio increases. Increased
flood duration and locally high groundwater tables would shift the PET ratio
where wetlands occur.
17
requirement of 30 cm for 14 days of the growing season. Differing hydraulic
conductivities of various soils and hydraulic gradients through the floodplain may
drastically influence the duration of soil anoxia. The intermittent to ephemeral nature of
losing streams in subhumid regions may not support groundwater levels for a duration
sufficient to support wetland conditions. Without sufficient hydroperiod from overbank
flooding, I expect that hydrophytes may be missing or confined to the channel banks.
Exactly how much flooding is required to support wetlands in subhumid regions
has yet to be verified. If overbank flow duration and the zone of saturation levels are a
function of the PET ratio, then at some PET ratio, overbank flooding and zone of
saturation levels are reduced to the point at which jurisdictional wetlands do not exist
within a given drainage basin area. It is my intent to characterize the controlling factors
in the development and maintenance of riparian wetlands. Characterization was done by
examining several factors that may influence the development and maintenance of
riparian wetlands; stream order, drainage basin area, soil, and overbank flow duration.
These factors were then compared with wetland and vegetation measurements and
occurrences along a PET ratio gradient from humid to subhumid.
Methods
Study Sites
Central plain study sites were selected along 40° N latitude to include the climate
change from humid to subhumid centering on the region where potential vegetation
changes from forest to grassland between 94° W and 103° W longitude (Kuchler 1967)
(Figure 8, Table 1). This transect was selected as representative, coinciding with the
western limit of upland forests. This region was also selected due to the large number of
currently operating USGS stream monitoring stations.
The region of study included 36 streams encompassing an area from Chariton,
Iowa, to Cope, Colorado (38° 30' to 41° 30' N latitude). A variety of drainage basins
ranging in size from 10.6 km^ to 1260 km^ were selected in order to compare between
different watershed areas from humid to subfrumid. Criteria for selecting the 23 USGS
gaged sites were based on sites having a synchronous period of record (October 1983-
September 1993) for mean daily discharge and minor to no alteration to the stream
channel. This period of time was chosen in order to estimate non-El Niño, climatic
conditions. The 13 ungaged sites were selected using USGS topographic maps and field
reconnaissance. Based on field visits, sites were rejected that were charmelized, received
back flooding from dams, received significant augmentation of flow as return flow from
municipalities or agriculture, or had active beaver dams. Preference was given to non-
agriculturally altered reaches of streams.
Sixteen streams in North Carolina were selected from headwater streams sampled
by Rheinhardt et al. (1998) (Figure 9, Table 2). These streams were selected in order to
19
Degrees West Longitude
Figure 8. Study sites in the central plain region. PET ratios along 40° N latitude.
PET ratios greater than 1.0 indicate a water deficit. PET ratios of individual sites vary
from this plot due to differing locations of the closest station to the individual study
sites.
20
Table 1. Study site name, location, PET ratio, USGS gaging station, and National Climatic Data
Center weather station for the central plain study sites. Latitude (N), longitude (W).
USGS NCDC
Site# stream latitude longitude PET ratio gaging stn. weather stn.
1 White Breast Ck. 41.06.886 93.22.109 0.71 — 1394
2 Blue R. @ Grandview 38.59.874 94.31.777 0.72 06893500 4359
3 Little Platte R. 39.23.393 94.37.392 0.72 — 7862
4 Indian Ck. 38.56.331 94.40.537 0.72 06893300 5972
5 Blue R., upstr. 38.48.750 94.40.517 0.72 06893080 5972
6 Trib to Little Blue R. 39.06.366 94.19.380 0.73 — 4850
7 Elk Ck. 40.43.300 93.56.317 0.73 — 536
8 102 River 40.35.202 94.47.017 0.73 06819185 576
9 Soldier Ck. 39.26.929 95.57.122 0.78 06889160 1529
10 Badger Ck. 41.27.178 93.46.321 0.78 — 2203
11 Soldier Ck., upstr. 39.33.950 95.57.750 0.78 06889140 1529
12 Turkey Ck. 39.56.867 96.06..500 0.78 06814000 1408
13 Salt Ck. 38.36.533 95.38.283 0.79 06911500 4912
14 Weeping Water Ck. 40.47.583 95.54.667 0.80 06806500 5810
15 no Mile Ck. 38.38.404 95.32.726 0.80 — 6498
16 Mill Ck. 39.03.790 96.10.574 0.80 06888500 8563
17 Soldier Ck. @ Grove 39.12.133 95.52.417 0.80 06889200 7007
18 Little Timber Ck. 39.43.424 96.24.233 0.85 — 5063
19 Chapman Ck. 39.01.867 97.02.400 0.87 06878000 5306
20 Rock Ck. 41.01.775 96.33.171 0.88 06803530 5362
21 Wildcat Ck. 41.00.173 97.21.654 0.89 — 8328
22 Big Blue R. 41.06.211 97.20.819 0.89 - 8328
23 Mill Ck. 39.48.833 97.02.033 0.90 06884200 8578
24 White Rock Ck. 39.53.917 98.15.083 0.93 06854000 4982
25 Stevens Ck. 40.51.417 96.35.700 0.93 06803520 4815
26 Kings Ck. 39.06.117 96.35.700 0.95 06879650 8259
27 Turkey Ck. 41.09.400 98.33.367 0.98 06784800 7515
28 Elm Ck. 40.05.333 98.26.117 1.04 06852000 3395
29 Salt Ck. 39.08.500 97.50.167 1.05 06876700 5363
30 Thompson Ck. 40.05.350 98.45.633 1.13 06851500 7070
31 Coates Ck. 40.06.132 98.54.295 1.17 — 3595
32 Driftwood Ck. 40.08.827 100.40.489 1.21 06836500 5310
33 S Fk Sappa Ck. 39.40.370 100.43.300 1.25 06844900 5906
34 Bow Ck. 39.33.767 99.17.067 1.31 06871500 6374
35 Arikaree R. 39.39.961 102.50.687 1.35 — 4380
36 N Fk Smokv Hill R. 39.19.818 102.16.528 1.75 — 1121
21
Figure 9. North Carolina study sites are located in the coastal plain region. The
PET ratios for these sites ranged from 0.83 to 0.67.
22
Table 2. Study site name, location, PET ratio and National Climatic Data Center
weather station for the North Carolina study sites. Latitude (N), Longitude (W).
NCDC
Site id. stream latitude longitude PET ratio weather stn.
A Trib to Corduroy Swamp 36.28.583 77.15.967 0.67 5996
B Bluewater Branch 36.20.700 77.03.917 0.67 5996
C Lobelia Run 36.25.050 77.25.533 0.75 4456
D Collie Creek 36.22.750 77.27.733 0.75 4456
E Trib to Sandy Run 36.12.600 77.15.150 0.76 4962
F Pecan Grove Slough 36.06.833 77.29.533 0.76 4962
G Trib to Wildcat 36.10.266 76.58.667 0.76 4962
H Big Swamp 35.59.017 77.01.683 0.78 9440
I Etheridge Swamp 35.58.383 77.18.250 0.78 9440
J Phillipi Branch 35.35.250 77.15.667 0.78 3638
K Otter Creek 35.43.033 77.31.083 0.78 3638
L Trib to Six Runs 34.51.717 78.10.367 0.81 1881
M Bulltail Creek 34.44.483 78.12.150 0.81 9423
N Trib to Crane Creek 34.55.367 78.16.567 0.81 9423
0 Elm City 35.48.950 77.47.367 0.82 9476
P Spicer Preserve Creek 34.58.967 77.58.467 0.83 9081
23
test and perfect the methods of analysis with previously studied sites. When compared
with the region where 40° N latitude intersects the East Coast, North Carolina streams
have a lower stream gradient with a warmer climate. PET ratios are comparable. These
first through fourth order streams were all ungaged and located on the humid inner
coastal plain ofNorth Carolina. The same criteria utilized in the central plains for
rejection or acceptance of sites were applied in the selection ofNorth Carolina sites
except for the discharge data, because none were available.
Determination of PET Ratio
Potential évapotranspiration is the amount ofwater returned to the atmosphere as
vapor through the combined effects of evaporation and transpiration, if the water supply
is unlimited. PET can be measured by several direct methods, or estimated using one of
several equations (Gray et al. 1970). The most accurate way to estimate PET would be to
measure the factors influencing PET (solar radiation, wind velocity, vegetation, surface
correction, temperature, water availability, etc.) on site for a period of time sufficient to
determine an average year. For the purpose of this study, PET was estimated as
described by Holdridge et al. (1971) from the mean annual biotemperature multiplied by
58.93. The Holdridge equation gives values similar to the Thomthwaite and Holzman
(1942) equation without the calculation of sunshine duration and annual heat index. Mean
annual biotemperature was calculated from the weighted monthly averages ofminimum
and maximum daily temperatures. For months averaging less than 0° C, a value of 0° C
was inserted in place of a negative number. These monthly values were then totaled for
the years 1983 - 1993 and averaged to determine the annual PET. Average annual
24
precipitation was calculated from average daily rainfall. Potential évapotranspiration was
then divided by average precipitation to determine PET ratio (Holdridge et al. 1971).
Data for these calculations were obtained from a CD-ROM database containing climatic
data taken by the National Climatic Data Center (NCDC) (Earthinfo 1995a) from the
closest monitored weather station to each study site (< 35 km). Sites were assigned
numbers (central plains) and letters (North Carolina) based upon increasing PET ratios
(Figures 8 and 9).
Determination of Bankfull Depth
Flood stages for many streams are determined from data reported by the National
Weather Service (NWS) (1997). However, the NWS estimates for flood stage are, in
many cases, the stage at which flood damage occurs, which is not necessarily directly
related to bankfull stage. Flood stage is also listed for some streams by the USGS station
records. However, USGS stages are subject to vagaries in the definition of “flooding”
and bankfull (Williams 1978; Burke and Nutter 1995; P. Tumipseed personal
communication 1997). For more accurate data on flood stage, bankfull must be
determined in the field (Rosgen 1996).
Average bankfull depth was determined by analysis of individual stream cross-
sections within a reach of channel equal in length to 20 bankfull channel widths (Rosgen
1996). Obvious visual indicators of bankfull (top of the point bar, vegetation change,
topographic break, change in size distribution of surface materials, and change in debris
deposited between rocks) were measured along 3 cross-sections at intervals of 10
bankfull widths (Figure 10) (Leopold 1994, Rosgen 1996). The bankfull depth was
Surveying Flood-prone Laser level
Figure 10. Bankfull depth is the depth from bankfull stage to the thalweg of the stream.
Bankfull depth was determined from bankfull indicators at three cross sections within the reach
in order to determine average bankfull depth. In this study, an interval of 10 bankfull widths
was used between cross-sections. The flood-prone width is the width of floodplain at twice
bankfull depth. Modified from Rosgen (1996). K)
26
averaged for the 3 cross-sections to determine the average bankfull depth. This average
bankfull depth was used to calculate bankfull discharge from stage / discharge curves and
tables created by USGS state representatives J. E. Putnam (KS), D.A. Eash (lA), G.B.
Engel (NE), and L.A. Waite (MO). Bankfull discharge was then compared with USGS
stream discharge records (Earthinfo 1995b).
Determination of bankfull depth proved to be difficult after heavy rainfalls and for
higher order streams. The large amounts of rainfall in the Great Plains during early
summer 1998 due to El Niño caused frequent flooding at many study sites. Most sites
had been flooded within a week prior to my visits. Analysis of floodplains was not
prevented unless the rainfall had occurred within a couple days and the flow still
exceeded bankfull. One site had to be visited three times in order to complete the
analysis. When stream movement was slow, with a depth greater than 1.85 m from the
thalweg to surface, depth was measured by marking the water surface on a surveying rod
after finding the thalweg.
Difficulties were also encountered while entering and exiting the stream. Steep
banks of clay or loam were extremely slippery because they were often saturated with
water from precipitation, ground water discharge, high streamflows, or from my dripping
clothing. Two methods of access were used: (1) Steps were cut into the banks using a
shovel and on streams with high banks, and (2) Rappelling rope was tied to trees in order
to give a handhold and increase traction. Rappelling rope was used at one site in order to
secure myself to a tree so that the equipment and I would not be washed downstream
while finding the thalweg and measuring bankfull depth. These methods greatly
27
increased the speed and efficiency of the surveying and measurement of bankfull depth
and wetland width.
Delineation ofWetlands
Typically, the procedure of reach selection, vegetation analysis, and soil sampling
were accomplished simultaneously. The initial site examination involved a geomorphic
survey either from the channel or along the bank in order to determine if alteration had
occurred and to select a representative reach of stream. Streams were examined for a
distance ranging from 500 m to 2 km depending on stream width and the amount of
variation present in the geomorphology and vegetation. During this initial examination,
species that were prevalent were noted and soil samples were examined periodically. If
the site was suitable, closer examination of the vegetation and soil was done along
transects during the measurement of flood-prone width and wetland width.
Prevalent and dominant plant species were identified and then tallied. Species
that represented greater than 10 percent of the community were considered to be
prevalent. Species that comprised greater than 20 percent of the community, covered the
greatest percentage of area, or constituted the perceived majority of basal area were
considered to be dominant. Species were given indicator status (Obligate upland (UPL)
found in uplands 99% of time; Facultative upland (FACU) in uplands 67 - 99%;
Facultative (FAC) found in wetlands 34 - 66%; Facultative wetland (FACW) found in
wetlands 67 - 99%; Obligate wetland (OBL) found in wetlands 99%) (nomenclature used
for following listings) based on their frequency of occurrence in wetlands in the specified
region as listed in the national list of plant species that occur in wetlands (Reed 1988).
28
These indicator statuses were then given ecological index (El) values (UPL = 5, FACU =
4, FAC = 3, FACW = 2, OBL = 1) and averaged (Wentworth et al. 1988). The number of
species that were prevalent and dominant determined species richness.
Soil samples were taken from several locations within the reach of FPA. Soils
were examined for hydric conditions (mottles, oxidized pore linings, organic streaking,
sulfidic odor, and the current state of hydration) within 30 cm of the soil surface. Matrix
and mottle color were determined using Munsell soil color charts (GretagMacbeth 1996).
The texture of soils was determined in the field at several locations in the FPA according
to the Thein (1979) method. This method of determining texture utilizes the
cohesiveness and grain size of the soil to categorize it into a texture type. For
regressional analysis, the soil textures were given numerical values from fine to coarse
(Taylor et al. 1999).
Field Analysis of Percentage ofWetlands
The edge of the flood-prone width (FPW) was surveyed, marked, and measured.
On both sides of the stream, measurements were taken of the wetland width along 3 or
more transects perpendicular to the floodplain direction at intervals of 10 bankfull widths
or less. Delineation was conducted similarly to the United States Army Corps of
Engineers 1987 protocol (Environmental Laboratory 1987). At all streams, vegetation,
soil, primary, and secondary wetland indicators were taken into consideration while
making the wetland determination. For large floodplains, the same method of
determination for bankfull depth and flood-prone depth was used but measurements of
the FPW were taken from the roadway. In all cases, wetland widths were measured
29
along at least 3 transects, averaged and divided by the average FPW and expressed as a
percent.
Determination of Overbank and Above Average Flow Duration
Bankfull discharges were estimated using tables and curves created for the gaged
streams by the state USGS offices. Mean daily discharges for the period of October 1983
- September 1993 (Earthinfo 1995b) were then compared with this bankfull discharge.
The total number of days during this period that average daily discharge exceeded
bankfull discharge was tallied for the 10-year period and averaged.
Elk Creek near Decatur City, Iowa (site 7) will be used as an example of this
procedure. Site 7 has an average bankfull depth of 3.04 m. In the expanded-precision
rating table, the discharge required to meet bankfull depth is 109.2 m^/s. In the Earthinfo
(1995b) CD-ROM listing ofUSGS stream data, mean daily discharges were compared
with the bankfull discharge. Days that average discharge equaled or exceeded bankfull
discharge were counted. During the period of October 1983 - September 1993, 3 days
exceeded 109.2 mVs for an average of 0.3 days/year for the 10-year period.
Duration of above average flow was determined by comparison of the average
daily discharge for the period ofOctober 1983 - September 1993 with the individual days
(Earthinfo 1995b). Days that exceeded the average were tallied and averaged to
determine the average number of days exceeding average discharge.
Hydrograph Separation
Six gaged streams were selected in order to estimate and compare the
groundwater input to stream discharge across the climatie gradient. Three of these
30
streams (sites 2, 12, and 16) were located in humid regions and three (sites 30, 33, and
34) were in subhumid regions. Hydrographs were created for these six streams from a
synchronous time period from March - September 1993. Baseflow separations were
calculated using the fixed base method (Fetter 1994). Days after peak (N) = 0.8(A)°^
were determined
where N = days from peak discharge
A = drainage basin area (km^).
The average baseflow was calculated by removing the ascending and descending (days
after peak) limbs of stormflow events and averaging the remaining discharges.
Map Analysis
Stream order was determined by analysis ofUSGS 1:24000 topographic maps
using Strahler’s ordering system (1957). Maps were in digital raster graphic (DRG)
format obtained on-line for Missouri, Kansas, and Nebraska. Maps for Colorado and
Iowa were obtained from the USGS in DRG format on CD-ROM. Maps were analyzed
using ArcView 3.0 sofrware. Streams, represented by dashed or solid blue lines, with no
upstream bifurcation, were designated as first order. Stream orders were then compared
with other quantitative data.
Analysis of Data
Data were analyzed using SPSS 8.0. Percentages were converted to proportions
and transformed using an arcsine transformation in SPSS. Scatter plots were created for
all quantitative data and regression lines were drawn with the best fit using Microsoft
31
Excel 97. Outliers were removed in the analysis ofpercentage wetland based on the
presence of non-climatic influences limiting or enhancing wetland status. Sites without
trees were removed in the El analysis. In the wetland width analysis for the central
plains, one outlier was removed that had exceptional wetland width for cross-sectional
width and one that had exceptionally low wetland width to cross-sectional area. In
wetland width analysis for North Carolina, Otter Creek was removed due to atypical
stream characteristics. P - values were determined using SPSS and listed for significant
correlations. Vegetation was analyzed by El and comparison of Jaccard similarity
(Odum 1971). Jaccard similarity (S) = 2 (C) / (A-i- B)
where A = number of species at site A
B = number of species at site B
C — number of species common to both sites
Ecological index and Jaccard similarity ofprevalent species were compared between
adjacent sites when ordered by increasing values ofPET ratios, latitude, longitude,
stream order, soil texture, cross-sectional area, drainage basin area, percent wetland, and
wetland width (m).
Results
PET Ratio Effects
Potential évapotranspiration (PET) ratios for the central plain streams ranged from
0.71 to 1.75. Streamflow patterns varied with PET ratios from perennial (lower PET
ratio) to ephemeral (higher PET ratios). PET ratios explained variations better than
simply precipitation in 3 of 4 comparisons (Figures 11 and 12). Vegetation in the
floodplains ranged from forest dominated by silver maple {Acer saccharinum) and
stinging nettle {Urtica gracilis) to upland grasses, cottonwood {Populus sargenta), and
honey locust (Gleditsia triacanthus).
PET ratios did not significantly correlate with the ecological index (El) of sites (r^
= 0.03) (Figure 1 la). No sites had an El average of less than 2.6 above the PET ratio of
0.98 (Table 3). As PET ratios increased forbs and, on some floodplains, shrubs were
replaced in the flood-prone width (FPW) by upland grasses and prickley pear cacti
{Opuntia sp.) as the dominant forms of understory vegetation. The number ofwoody
species that were prevalent in the flood-prone area (FPA) decreased as precipitation
decreased (r^ = 0.55) (Figure 12a). Tree species that are considered to be wetland
vegetation (FAC+ to OBL) became uncommon at streams with a PET ratio above 0.90.
It was observed that these species also tended to be found closer to the stream as the PET
ratio increased. Most wetland vegetation above the PET ratio of 1.04 was located within
close proximity of the charmel.
Urtica gracilis (FACW) and Laportea canadensis (FACW) dominated most
floodplains with PET ratios up to 0.80. Trees that were FACW {Acer saccharinum,
33
PET ratio PET ratio
Figure 11. The relationship between the PET ratio and four variables at the central
plain sites: (a) ecological indices, (b) the number ofprevalent woody species, (c)
the percent of the FPW that is wetland, and (d) the duration of flow exceeding
average discharge. Outliers were removed as indicated in methods.
34
precipitation (mm)
precipitation (mm) precipitation (mm)
0 500 1000 1500
precipitation (mm)
Figure 12. The relationship between precipitation and five variables at the
central plain sites: (a) the richness ofwoody species, (b) the percentage of the
floodplain that is wetland, (c) wetland width, (d) the duration of above average
flow, and (e) PET ratios. Although PET ratios and precipitation were closely
related, PET ratios do a better job of explaining variations than precipitation
alone.
Table 3. Factors influencing the occurrence of vegetation at the central plain study sites, n/a = not available.
Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78 0.78 0.78 0.78 0.79 0.80 0.80 0.80 0.80 0.85
Bkfl fl duration/yr n/a 4.2 n/a 0.2 0.3 0.3 n/a n/a n/a n/a 0.1 2.5 0.9 1.7 n/a 2.0 0.9 n/a
Above avg days n/a 88 n/a 57 62 88 n/a 54 n/a n/a 40 64 63 87 n/a 97 60 n/a
Wetland width (m) 50 63 11 2 0 29 7 16 149 8 3 20 30 22 10 40 61 73
Percent wetland 11 78 71 3 0 62 100 42 27 3 2 2 13 72 69 3 43 22
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5 5 5 6 4
# species 15 20 10 14 17 16 17 19 11 19 17 12 15 10 17 9 8 12
El 2.58 2.88 2.45 2.65 2.61 2.71 3.05 2.59 2.36 2.96 2.92 2.78 2.71 3.10 2.68 2.67 2.88 3.17
Precipitation (mm) 946 928 928 1060 1060 1147 1050 1050 957 950 957 918 958 879 941 900 982 904
Woody species 6 8 3 9 9 10 10 8 5 10 6 4 9 7 8 5 5 7
Drainage 407 n/aarea (km ) n/a 487 606 119 68.9 136 n/a 221 128 n/a 43.8 715 287 624 n/a 824
Site number 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
PET ratio 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98 1.04 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Bkfl fl duration/yr 3.7 1.8 n/a n/a 3.8 1.5 2.4 0.0 2.3 2.5 2.5 1.3 n/a 3.2 0.2 0.4 n/a n/a
Above avg days 58 55 n/a n/a 68 43 56 94 70 28 54 37 n/a 44 22 43 n/a n/a
Wetland width (m) 24 9 0 43 42 6 2 0 3 0 0 0 0 1 0 0 0 0
Percent wetland 2 38 0 27 15 8 1 0 28 0 0 0 0 16 0 0 0 0
Strahler order 5 5 3 4 5 5 5 3 4 4 6 5 3 4 4 4 5 4
# species 8 10 10 4 12 5 7 13 8 12 11 14 9 4 4 6 5 3
El 2.96 2.94 2.79 2.57 2.98 2.67 2.89 3.00 2.09 2.61 3.30 2.79 2.93 2.68 2.93 3.12 2.68 4.00
Precipitation (mm) 857 746 779 779 917 721 813 835 690 687 768 678 616 546 620 646 437 427
Woody species 5 5 6 1 7 2 2 5 1 6 4 7 5 2 2 5 2 0
Drainage area (km ) 777 308 n/a n/a 891 894 124 10.6 171 102 995 723 n/a 932 1160 883 n/a n/a
U)
36
Platanus occidentalis, and Celtis laevigata) and OBL (Salix exiqua) were dominant at
sites with PET ratios of less than 0.89. Acer negando (FAC) was common on many
floodplains with PET ratios between 0.72 and 1.31. Populas sargenta (FACU), Ulmus
rubra (FAC), and Gleditsia triacanthus (FAC) dominated floodplains with PET ratios
above 0.90. Floodplain vegetation showed the greatest average site to site Jaccard
similarity (54%) when arranged by PET ratio. Equivalent averages for other variables
were percent wetland: 30%, wetland width: 30%, latitude: 33%, longitude: 42%, stream
order: 47%, drainage basin area: 43%, overbank flow duration: 40%, cross-sectional area:
43% (Appendix A).
The percentage ofFPW that is wetland decreased sharply as the PET ratio
increased to 0.98 (r^ = 0.34) (Figures 1 Ic). The highest PET ratio at which wetlands
occurred was 1.21 (16 %) along Driftwood Creek in Nebraska. The wetlands along this
stream appeared to have been supported by agricultural return flows. During the 4 hours
it took to conduct the FPA analysis, the stream stage rose 15 cm, coinciding with the
afternoon flood irrigation of the surrounding fields. The additional input brought the
stream stage to within 20 cm ofbankfull. This site would have been eliminated from
study if the existing conditions had been known prior to the completion of sampling. The
next driest site to have wetlands was at a PET ratio of 0.98.
All central plain sites had wetlands associated with them until a PET ratio of 0.93
except for two streams (Table 4). The Blue River upstream site (PET ratio = 0.72), in
Kansas lacked wetlands because the stream was cut into limestone and did not have soil
on one side of the stream. On the other side of the stream, the soils in the FPA were
37
Table 4. Wetland widths, flood-prone widths and percentage ofwidth that is wetland for
the central plain and North Carolina sites, n/a = not available.
Site number 1 2 3 4 5 6 7 8 9
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78
Flood-prone width (m) 459 81 15 81 19 47 7 38 550
Wetland width (m) 50 63 11 2 0 29 7 16 149
Percent wetland 11 78 71 3 0 62 100 42 27
Site number 10 11 12 13 14 15 16 17 18
PET ratio 0.78 0.78 0.78 0.79 0.80 0.80 0.80 0.80 0.85
Flood-prone width (m) 280 156 1007 231 31 15 1340 143 330
Wetland width (m) 8 3 20 30 22 10 40 61 73
Percent wetland 3 2 2 13 72 69 3 43 22
Site number 19 20 21 22 23 24 25 26 27
PET ratio 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98
Flood-prone width (m) 1206 24 121 158 280 75 194 67 10
Wetland width (m) 24 9 0 43 42 6 2 0 3
Percent wetland 2 38 0 27 15 8 1 0 28
Site number 28 29 30 31 32 33 34 35 36
PET ratio 1.04 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Flood-prone width (m) 155 458 331 6 5 350 41 49 10
Wetland width (m) 0 0 0 0 1 0 0 0 0
Percent wetland 0 0 0 0 16 0 0 0 0
Site id. A B c D E F G H I
PET ratio 0.67 0.67 0.75 0.75 0.76 0.76 0.76 0.78 0.78
Flood-prone width (m) 85 74 73 56 18 31 131 413 216
Wetland width (m) 85 74 73 56 18 31 131 413 216
Percent wetland 100 100 100 100 100 100 100 100 100
Site id. J K L M N O P
PET ratio 0.78 0.78 0.81 0.81 0.81 0.82 0.83
Flood-prone width (m) 30 104 44 117 44 74 13
Wetland width (m) 29 87 44 117 44 74 13
Percent wetland 96 84 100 100 100 100 100
38
isolated from normal streamflow due to the horizontal layering of the bedrock. Wildcat
Creek (PET ratio = 0.89) in Nebraska for unknown reasons did not have wetlands
associated with it within the study reach. Wildcat Creek was ephemeral and had a FPA
composed of silty clay. The majority of this stream's watershed had been converted to
agricultural fields with plow lines running through the channel.
PET ratios for the North Carolina sites ranged from 0.67 to 0.83. Upland
vegetation was composed of oak - hickory - pine (Quercus - Carya - Pinus) forest.
Streamflow patterns varied with stream order and drainage basin area from intermittent to
perennial. Floodplain forests were composed primarily of red maple {Acer rubrum),
water túpelo {Nyssa aquatica) (in wettest sites), and swamp túpelo {Nyssa biflora).
All sites in North Carolina had an El average of less than 2.5 with a slight
decrease as the PET ratio decreased (r =0.15) (Figure 13a). Herbaceous vegetation
remained similar in indicator status at most sites. However, trees were more sensitive to
the decrease in the PET ratio. At streams with higher PET ratios (0.78-0.83), trees such
as loblolly pine {Pinus taeda) (FAC), sweetgum {Liquidambar styraciflua) (FAC+), and
sourwood {Oxydendron arboreum) (FACU) were common in the FPA. As the PET ratio
decreased below 0.78, these species became less common and were replaced by red
maple {Acer rubrum) (OBL), water túpelo {Nyssa aquatica) (OBL), and swamp túpelo
{Nyssa biflora) (OBL). These three OBL trees were common on most FPAs with
increasing dominance as PET ratios decreased. In North Carolina the number of
prevalent woody species in the FPA increased with increasing PET ratios (r^ = 0.19)
(Figure 13b). The number of prevalent herbaceous species remained fairly constant
39
Eaveralge
T3
'î
x-section (m^)
Figure 13. Two factors possibly influencing North Carolina floodplains: (a)
ecological indices and PET ratio, (b) the number of prevalent woody species and
PET ratio, (c) cross-sectional area and PET ratio, (d) wetland width and the cross-
sectional area of the stream, and (e) ecological indices and cross-sectional area.
40
(Table 5, Appendix B).
Most sites in North Carolina had flood-prone widths (FPW) that were 100 percent
wetland. Two sites had less than 100 percent: Otter Creek (85 %) and Phillipi Branch (97
%) (Table 4). Floodplains in North Carolina exhibited more primary and secondary
wetland indicators than the central plain FPAs, even for PET ratios that were higher than
several of the central plain sites. Consistently, more North Carolina floodplains had
buttressed vegetation and water stained leaves. In contrast, the central plain sites
exhibited more sediment and wrack deposition (Table 6). North Carolina exhibited a
higher percentage of FPAs that featured depressional areas, and a lower gradient across
FPWs that were capable ofponding water.
Stream Order
Central plain streams ranged in order from third to sixth with the majority of
streams being fourth and fifth order. Stream order increased slightly with drainage basin
area (r^ = 0.11) (Figure 14a) and cross-sectional area increased with stream order (r^ =
0.42) (Figure 14b). However, wetland width was not directly related to stream order (r^ =
0.02) (Figure 14c). These streams also did not show an increase in the percentage of
FPW that is wetland (r = 0.02) with increasing stream order (Figure 14d). Furthermore,
El averages did not correlate with stream order (r^ = 0.00) (Figure 14e) (Table 3).
Herbaceous vegetation showed no trend in response to stream order. However,
trees did show a trend in response to stream order. Ulmus rubra (FAC) showed
dominance on fourth order streams and some fifth order streams in drier areas. Acer
saccharinum (FACW) showed dominance on fifth order streams and on some fourth and
Table 5. The factors influencing the occurrence of vegetation at the North Carolina study sites.
Site id. A B C D E F G H I J K L
PET ratio 0.67 0.67 0.75 0.75 0.76 0.76 0.76 0.78 0.78 0.78 0.78 0.81
Wetland width (m) 85 74 73 56 18 31 131 413 216 29 87 44
Percent wetland 100 100 100 100 100 100 100 100 100 96 84 100
Strahler order 2 2 1 1 1 1 1 3 4 2 4 1
# species 5 7 6 7 7 2 6 7 9 8 11 8
El 2.06 1.43 1.67 2.04 1.86 1.00 2.60 1.29 1.52 1.63 2.09 2.00
Precipitation (mm) 1234 1234 1143 1143 1171 1171 1171 1136 1136 1185 1185 1190
Woody species 2 3 2 2 5 2 4 4 5 3 7 4
Drainage area (km ) 3.09 13.74 1.40 3.27 0.17 0.46 3.55 42.11 59.21 1.84 127.49 0.92
Site id. M N O P
PET ratio 0.81 0.81 0.82 0.83
Wetland width (m) 117 44 74 13
Percent wetland 100 100 100 100
Strahler order 2 1 2 1
# species 14 8 6 10
El 2.50 2.25 1.50 2.24
Precipitation (mm) 1284 1284 1150 1163
Woody species 6 5 1 7
Drainage area (km^) 5.68 1.31 6.54 0.44
Table 6. Primary and secondary wetland indicators for the central plain and North Carolina sites.
+ = presence of condition, — = absence of condition.
Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Soil listed as hydric + -- + +
Buttressing + + + + -- + +
Sediment deposition + + + + - + + + + + + + + + + + + - -- + + -- + + + -- + + +
Water marks + +
Saturated @ 12" + + + - — + + + + — + + - - + + + + + + + + + + - -
Ponded water + + + + +
Wrack deposition + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Gleyed/ mottled + — + — — + + — + — + + + + — + + + + + — + + — + ~ + - -
Stained leaves
Coated soil grains
Streaking
Site number 32 33 34 35 36 A B C D E F G H I J K L M N 0 P
Soil listed as hydric + + + +
Buttressing + + + + + + + + + + + + + + + +
Sediment deposition + - + - + + - +
Water marks -- + + + + -- +
Saturated @ 12" + + + + + + + + + + - + + + + -
Ponded water + •f + + + + +
Wrack deposition + + + + + -f- + +
Gleyed/ mottled + - + + + + + -- + + + - + +
Stained leaves + + + + + + + + +
Coated soil grains + -- --
Streaking +
ro
43
0)
-O
L.
O
E
CQ
0^
L.
35
Figure 14. The relationship between stream order and five variables at central plain
sites: (a) drainage basin area, (b) channel cross-sectional area, (c) wetland width,
(d) the percent of the FPW that is wetland, and (e) the ecological indices.
44
third order streams in areas with lower PET ratios. In North Carolina Taxodium
distichum (OBL) was only present on third and greater ordered streams (Appendix A).
The cross-sectional area of the studied streams ranged in the central plains from
2.8 m^ to 144.5 m^. In North Carolina the stream cross-sections ranged from 0.004 m^ to
21.6 m . Central plain stream cross-sectional areas decreased in response to a rise m the
PET ratio (r = 0.32), while North Carolina streams did not (Figures 15a and 13c). In
North Carolina and in the central plains streams, greater cross-sectional areas were
associated with greater wetland widths (Table 7). With the removal of two outliers (sites
9 and 18) the r^ rises to 0.70 at the central plain sites (Figure 15b). With the removal of
Otter Creek (site K), the North Carolina r^ rises to 0.89 (Figure 13d). This relationship
does not exist between cross-sectional area and percent of FPW that is wetland at the
central plain sites (r^ = 0.06) (Figure 15c), or El averages for the central plain (r^ = 0.00)
or North Carolina (r^ = 0.00) (Figures 15d and 13e).
Drainage Basin Area
Drainage basin areas did not correlate with the percentage of the FPW that is
wetland (r^ = 0.11) (Figure 16a). El averages showed slight increases with drainage basin
area (r^ = 0.19) (Figure 16b) (Tables 3 and 5). However, the El average / drainage basin
area comparison may be skewed due to the increase in gaged drainage basin areas with
PET ratio (r^ = 0.24) (Figure 16c). Gaged drainage basin size increased with PET ratio
due to selection criteria utilizing USGS gaging stations that monitor mean daily
discharge. Most streams that have no discharge for the majority of the year are not
monitored for mean daily discharge.
45
?a
%PtranrsEafortmTieod
x-section (m') x-section (m )
x-section (m^)
Figure 15. The relationship between cross-sectional area and five variables at the
central plain sites: (a) the PET ratio, (b) wetland width, (c) the percent of the FPW
that is wetland, (d) the ecological indices, and (e) the duration of flow exceeding
average discharge.
Table 7. The geomorphic parameters of the central plain and North Carolina sites.
Qbkfl = bankfull discharge, n/a = not available.
Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Drainage area (km"^) n/a 487 606 119 69 136 n/a 221 128 n/a 44 715 287 624
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5
Bankfull depth (m) 4.58 3.20 2.36 1.97 1.56 3.04 0.95 2.76 2.34 2.75 1.84 3.13 1.99 3.39
Bankfull width (m) 22.0 18.5 13.5 n/a 15.5 21.0 6.5 15.2 7.4 16.9 12.3 17.6 18.9 19.0
Cross-sectional area (m^) 100.8 59.2 31.9 n/a 24.2 63.8 6.2 42.0 17.3 46.5 22.5 55.2 37.6 64.4
Flood-prone width (m) 459 81 15 81 19 47 7 38 550 280 156 1007 231 31
Qbkfl (m^/s) n/a 80 n/a 91 33 109 n/a n/a n/a n/a 37 64 85 79
Site number 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Drainage area (km^) n/a 824 407 n/a 777 308 n/a n/a 891 894 124 11 171 102
Strahler order 5 5 6 4 5 5 3 4 5 5 5 3 4 4
Bankfull depth (m) 3.24 4.12 5.16 2.56 3.77 2.61 0.39 3.09 3.69 2.38 3.01 1.58 2.18 1.11
Bankfull width (m) 13.0 17.0 28.0 11.6 23.0 10.9 7.3 13.1 30.2 8.5 11.5 8.9 8.8 7.6
Cross-sectional area (m^) 42.1 70.0 144.5 29.7 86.7 28.4 2.8 40.5 111.4 20.2 34.6 14.1 19.1 8.4
Flood-prone width (m) 15 1340 143 330 1206 24 121 158 280 75 194 67 10 155
Qbkfl (m^/s) n/a 73 297 n/a 45 37 n/a n/a 57 27 17 41 8 7
Site number 29 30 31 32 33 34 35 36
Drainage area (km ) 995 723 n/a 932 1160 883 n/a n/a
Strahler order 6 5 3 4 4 4 5 4
Bankfull depth (m) 3.34 1.32 0.87 0.86 1.39 0.96 0.75 0.94
Bankfull width (m) 10.8 16.2 5.0 5.2 6.6 9.6 9.1 5.1
Cross-sectional area (m^) 36.1 21.4 4.4 4.5 9.2 9.2 6.8 4.8
Flood-prone width (m) 458 331 6 5 350 41 49 10
Qbkfl (m^/s) 27 32 n/a 1 11 20 n/a n/a
4:^
ON
Table 7 cont'd. The geomorphic parameters of sites.
Site id. A B C D E F G H I J K
Drainage area (km ) 3.1 13.7 1.4 3.3 0.2 0.5 3.6 42.1 59.2 1.8 127.5
Strahler order 2 2 1 1 1 1 1 3 4 2 4
Bankfull depth (m) 0.44 0.59 0.21 0.34 0.17 0.02 0.31 1.12 0.47 0.35 2.00
Bankfull width (m) 3.3 4.8 3.5 2.8 1.9 0.2 3.1 7.5 8.8 3.1 10.8
Cross-sectional area (m^) 1.5 2.8 0.7 1.0 0.3 0.004 1.0 8.3 4.1 1.1 21.6
Flood-prone width (m) 85 74 73 56 18 31 131 413 216 30 104
Site id. L M N 0 P
2
Drainage area (km ) 0.9 5.7 1.3 6.5 0.4
Strahler order 1 2 1 2 1
Bankfull depth (m) 0.44 0.45 0.48 0.39 0.22
Bankfull width (m) 1.5 2.2 1.0 2.3 1.3
Cross-sectional area (m^) 0.7 1.0 0.5 0.9 0.3
Flood-prone width (m) 44 117 44 74 13
4:^
48
area (km^)
area (km^)
Figure 16. The relationship between the drainage basin area and three variables at
the central plain sites: (a) the percent of the FPW that is wetland, (b) the El
average, and (c) PET ratio.
49
Soil
Soils ranged from clay to coarse sandy loam. The majority of soils along the
central plain streams were clay based or had clay constituents. Soil color varied at
several streams depending upon the location of sampling, wetland status, and geomorphic
setting. Soil texture showed no relationship with the percentage of FPW that is wetland
(r^ = 0.02) (Figure 17a). Soil texture varied only slightly with the duration of overbank
flow (r^ = 0.07) (Figure 17b). Fine-grained soils were associated with a slightly longer
duration of flooding than coarse-grained soils. Bankfull depth also varied slightly with
soil texture (r =0.15) (Figure 17c). Fine-grained soils generally had steeper banks with
greater bankfull depths than the coarse-grained soils (Table 7).
Overbank and Above Average Flow Duration
Most of the floodplains of the central plain streams studied were not
geomorphically capable of ponding water in the area of study. This was due mainly to a
lack of depressions and impoundments by natural levees in the floodplain. Most
inundation was a result of direct, overbank flow from the stream. Those floodplains that
were geomorphically capable ofponding water on the floodplain met wetland criteria for
the majority of the ponded depression area. The average annual duration of overbank
flow for central plain streams ranged from 0 to 4.2 days per year. Generally streams
would not exceed bankfull discharge for several years and then, during a wet year,
bankfull would be exceeded several times. During the time period in which analysis was
done, bankfull discharge was exceeded several times on many streams in the central
plains.
50
5.0
y = -0.011x + 0.253
* R^= 0.021 u 4.0«
? * (a) ^ 3.0
?
^
1.0
%transformed —? r—?- ? ? ? 1 ? ? 0.00 5 10 15 0 5 10 15soil texture soil texture(dempth)
soil texture soil texture
Figure 17. The relationship between soil texture and four variables at the central
plain sites: (a) the percent of the FPW that is wetland, (b) duration of average
annual overbank flow, (c) bankfull depth, and (d) the duration of above average
discharge. 1 = clay, 2 = silty clay, 3 = sandy clay, 4 = silty clay loam, 5 = clay
loam, 6 = silty loam, 7 = loam, 8 = sandy clay loam, 9 = sandy loam, 10 = loamy
sand, 11= medium sandy clay loam, 12 = coarse sandy loam.
51
The best correlations with flood duration for the central plain streams were
drainage basin area (r^ = 0.14) (Figure 18a) and stream order (r^ = 0.13) (Figure 18b).
For the 23 gaged sites, duration of overbank flow had no effect on the percent of the FPW
that is wetland (r^ = 0.03) (Figure 18c) or on the El average (r^ = 0.01) (Figure 18d)
(Table 8).
Longer durations of flow exceeding average were associated with greater wetland
width (r^ = 0.48) (Figure 19a), cross-sectional area (r^ = 0.30) (Figure 15e), and the
percent of the FPW that is wetland (r^ = 0.29) (Figure 19b). Longer durations of above
average flow did not correlate with lower El average (r^ = 0.00) (Figure 19c). The
duration of time that stream flow exceeded average responded linearly to increases in
PET ratio (r^ = 0.41) (Figure lid). Greater durations were also associated with finer soil
textures (r =0.15) (Figure 17d). Duration was not a function of drainage basin area (r =
0.04) or of stream order (r^ = 0.02) (Figure 19d and e).
Groundwater Contribution to Discharge
Average baseflow for the three streams with PET ratios less than 1.0 were as
follows site 2 was 3.43 m /sec, site 12 was 3.44 m /sec, and site 16 was 5.72 m /sec
(Figure 20a, b, and c). Average baseflow for the three streams with PET ratios greater
?5 -5
than 1.0 were as follows: site 30 was 0.76 m /sec, site 33 was 0.03 m /sec, and site 34
was 0.65 m^/sec (Figure 20d, e, and f). When normalized, sites 2 (7.04 m^/sec/1000
km^), 12 (4.81 mVsec/1000 km^), and 16 (9.17 mVsec/1000 km^) had greater average
baseflow contribution than sites 30 (1.05 mVsec/1000 km^), 33 (0.03 m^/sec/1000 km^),
and 34 (0.74 m^/sec/1000 km^).
52
%trans(forkmaemdrea^) days / yr.
days / yr. days / yr.
Figure 18. The relationship between average annual overbank flow duration and
four variables at the central plain sites: (a) drainage basin area, (b) stream order,
(c) the percent of the FPW that is wetland, and (d) the ecological indices.
Table 8. The possible factors influencing the average annual bankfull flow duration for the central plain sites.
Qbkfl = bankfull discharge, bkfl fl duration = bankfull flow duration, soil codes: 1 = clay, 2 = silty clay, 3 = sandy clay,
4 = silty clay loam, 5 = clay loam, 6 = silty loam, 7 = loam, 8 = sandy clay loam, 9 = sandy loam, 10 = loamy sand,
11= medium sandy clay loam, 12 = coarse sandy loam, n/a = not available.
Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78 0.78 0.78 0.78 0.79 0.80 0.80 0.80 0.80 0.85
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5 5 5 6 4
Drainage area (km^) n/a 487 606 119 68.9 136 n/a 221 128 n/a 44 715 287 624 n/a 824 407 n/a
Qbkfl (m^/s) n/a 80 n/a 91 33 109 n/a n/a n/a n/a 37 64 85 79 n/a 73 297 n/a
Bkfl fl duration/yr n/a 4.2 n/a 0.2 0.3 0.3 n/a n/a n/a n/a 0.1 2.5 0.9 1.7 n/a 2.0 0.9 n/a
Soil code 8 2 3 3 1 7 6 7 1 7 1 4 7 2 8 7 6 1
Site number 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
PET ratio 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98 1.04 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Strahler order 5 5 3 4 5 5 5 3 4 4 6 5 3 4 4 4 5 4
Drainage area (km ) 777 308 n/a n/a 891 894 124 11 171 102 995 723 n/a 932 1160 883 n/a n/a
Qbkfl (m^/s) 45 37 n/a n/a 57 27 17 41 8 7 27 32 n/a 1 11 20 n/a n/a
Bkfl fl duration/yr 3.7 1.8 n/a n/a 3.8 1.5 2.4 0.0 2.3 2.5 2.5 1.3 n/a 3.2 0.2 0.4 n/a n/a
Soil code 1 1 2 2 1 7 2 6 2 11 2 9 9 6 n/a 8 12 3
54
Ea(vemwralgideth)
3 4 5 6 7
stream order
Figure 19. The relationship between the duration of flow exceeding average
discharge and five variables at central plain sites: (a) wetland width, (b) the percent
of the FPW that is wetland, (c) ecological indices, (d) drainage basin area, and (e)
stream order.
55
Figure 20. Hydrographs for selected sites from a synchronous time period from
March - September 1993. Bkfl = bankfrill discharge. Avg = average discharge. Avg
base = average baseflow. Hydrographs a, b, and c have wetlands associated with
them. Average baseflows were: (a) 3.43 mVsec, (b) 3.44 m^/sec, and (c) 5.72 m^/sec.
Hydrographs d, e, and f did not have wetlands associated with them. Average
baseflows were: (d) 0.76 m^/sec, (e) 0.03 m^/sec, and (f) 0.654 m^/sec.
Discussion and Conclusions
The purpose of this study was to determine the controlling factors in the
development and maintenance of riparian wetlands. Characterization was done by
examining several factors that may influence the development and maintenance of
riparian wetlands: stream order, drainage basin area, soil, and overbank flow duration.
These factors were then compared with wetland and vegetation measurements and
occurrences along a PET ratio gradient from humid to subhumid.
Maintenance ofWetlands
The greatest factor in determining the percentage of the FPW that is wetland was
the PET ratio. Without augmentation of discharge, it would appear that there must be a
surplus ofprecipitation over PET for wetlands to occupy a significant percentage of a
floodplain. The majority of floodplains with a PET ratio below 0.98 had wetlands
associated with them (Figure 21). This ratio appears to be the threshold for wetland
occurrence along the studied streams. In order for wetlands to be maintained, the supply
ofwater to floodplains must exceed the water demand of vegetation and the drainage
caused by the presence of the stream channel. However, this supply does not appear to
be solely driven by the PET ratio; other modifiers appear to be present. The presence of
these modifiers is shown by the exponential decrease in the percent of the floodplain that
is wetland as PET ratios increase (Figure 22a). There were large variations in the
percentage ofFPW consisting ofwetland that are only partially explained by the other
data collected through this study. One remaining possibility is the position of the
groundwater table.
57
Figure 21. Distribution of sites with (triangles) and without (circles) wetlands.
58
Oí
U
I
CJ
es
-o
a
J2
o»
PET ratio
Multiple factors interact to result in a (b)
exponential positive feedback situation.
If wetland occurrence were simply
• i a function ofPET ratio.
1 Wr • »
1
4-» 1^ ••X
D
Wet^- PET ratio ? Dry
Climate 1-0 Climate
Flood Hypothetical Groundwater
duration wetland discharge
response
Figure 22. The exponential decrease in (a) the percentage of the FPW that
is wetland as compared with (b) the conceptual model shows a similar
relationship. As climate becomes drier, multiple hydrologic sources also
decrease. As climates become wetter these factors interact in a positive
feedback manner. The feedback situation may result in wetland extension
beyond the FPW.
59
Groundwater inputs to stream channels were estimated for some sites in this
study. Of the selected streams, streams with associated wetlands had higher baseflow
discharges than those without. The increased durations of high baseflow and above
average discharge are, in part, the result of high groundwater levels in the drainage basin.
The more moist soil conditions could result in greater wetland widths along streams due
to greater throughputs ofwater in the stream and the floodplain soils from groundwater
discharge and bank storage.
Although precipitation is the ultimate source ofwater for the studied riparian
systems, there appears to be differences in the delivery (timing, intensity, volume) of
water between floodplains. North Carolina floodplains were consistently wetter than the
central plain floodplains even at similar PET ratios. This would suggest a greater
contribution of groundwater to the studied North Carolina floodplains. The wide
variations in the percent of the FPW that is wetland, at the central plain sites, could also
be attributed to variations in groundwater contributions. There is the possibility of a
positive feedback situation between the groundwater flow originating from precipitation
on uplands and streamflow resulting in stockpiling of groundwater in the floodplain.
During and following precipitation events, groundwater is recharged and
streamflow is increased by overland flow and groundwater discharge. As the stormflow
recedes, upstream variable source areas contribute groundwater discharge maintaining the
baseflow of the stream (Hewlett 1961a, Hewlett 1961b). The greater the baseflow is (and
thus the stream stage) the lower the hydraulic gradient is between the channel and the
floodplain, resulting in slower groundwater discharge. This desynchronous drainage
60
increases the duration of saturation in the floodplain at the point of study by decreasing
the rate of drainage of the floodplain soils and backing up the flow of groundwater
through the floodplain. Backed up groundwater flow results in a rise in the water table
elevation in the floodplain (Figures 23 and 24). Backed up groundwater flow in an
alluvial aquifer was shown to exist on Little Stony Creek, California in relation to
increased reservoir stage (M.C. Rains, presentation at Society ofWetland Scientists
annual meeting 1999). If the water table is high enough, then every addition ofwater not
utilized by vegetation or drained by the stream channel would contribute directly to the
water table elevation and could result in an exponential increase in the percentage of
FPW until wetlands may even extend beyond the FPW, as in North Carolina.
In order for this situation to function, yearly precipitation on the uplands and
floodplain must regularly exceed the ET of the vegetation and all other losses
(agricultural and domestic wells and regional groundwater discharge). This excess must
occur for a period of time during the growing season, sufficient to raise the groundwater
table to the level that the FPW remains saturated within 30 cm of the surface for 14 days
after the precipitation or overbank event. On the studied floodplains, the pattern of the
percentage of the FPW that is wetland decreased exponentially as the PET ratio
increased, indicative ofmultiple, decreasing water sources.
Greater channel depths did not negatively affect the occurrence ofwetlands
in North Carolina or in the central plains. In contrast, streams showed increased wetland
width with greater channel cross-sections, regardless of drainage basin area, stream order
or PET ratio. Natural channel cross-sections are formed by the discharges moving
Figure 23. The cyclical pattern of rain, high streamflow, and groundwater discharge into the stream. This
pattern is typical in humid region streams, (a) During baseflow conditions, streamflow is maintained by
groundwater discharge, (b) If groundwater levels become low enough (drought), streamflow may become
interrupted, (c) During storm events, streamflow and precipitation recharge the groundwater system, (d) The
greater the upstream hydrologic sources the higher the stream stage resulting in decreased groundwater
discharge and a saturated floodplain, (e) As stream stage decreases groundwater discharge increases,
draining the floodplain. Higher PET ratios result in a shortened period of time between steps (c) and (a).
This leads to a negative-feedback situation where the percentage of the FPW that is wetland decreases
exponentially with watertable decreases until floodplains do not meet wetland criteria.
groundwater profile during base flow
Interaction between streamflow and the water table
Figure 24. Cross-sections of streams contrasting hydrologic regimes of floodplains, (a) In regions where
groundwater supports streamflow and floodplain hydrology, overbank flow is not necessary to maintain wetlands,
(b) In regions where streamflow is lost to groundwater, wetlands can be maintained by supplemental sources
(irrigation retumflow, streamflow from outside the region), (c) However, in humid regions where groundwater
levels are normally too deep to support high percentages of the FPW as wetland, but high enough to support
streamflow (high hydraulic gradient), there may be an interaction between streamflow and the water table that may
support wetlands. There is a positive feedback interaction between groundwater levels and streamflow driven by
climate. This is expressed as a exponential decrease in the percent of the FPW that is wetland as PET ratios
increase. This interaction is a consequence of a reduced hydraulic gradient through the floodplain and an input of
water from the stream into the groundwater table. This results in elevated groundwater levels in the floodplain, and
thus greater percentages of the FPW as wetland. V = water table level.
ON
K)
63
through them (Leopold et al. 1964, Knighton 1998). Although the presence of a deeper
channel suggests that there may be more rapid drainage due to an increased hydraulic
gradient, apparently this is not necessarily the case. Most of the channels that were deep
tended to be perennial with greater average daily discharges ofwater in them, effectively
reducing a possible hydraulic gradient that would otherwise augment floodplain drainage.
Deeper channels also tended to be cut into fine-grained soils whose hydraulic
conductivities are less than those of coarse-grained soils.
On North Carolina FPAs, with the exception ofOtter Creek, there was a very
strong correlation ofwetland width with channel cross-section. Otter Creek is located on
the steeper gradient of the south side of Tar River, resulting in a greater stream gradient
when compared with other coastal plain streams. Otter Creek had wetlands comprising
85 percent of its FPW. This is indicative of the inability of the groundwater table to
sustain wetlands for the entire FPW of this stream.
Interestingly the duration of overbank flow did not relate to the percentage of the
FPW that is wetland. The lack of relation may have been because the difference between
the smallest and the greatest duration (0.0 and 4.2 days, respectively) was not large
enough to show the effect of overbank flow duration. The duration of overbank flow is
probably conservative, given that in order for a day to be counted as an overbank flow
day, daily discharge had to average greater than or equal to bankfull discharge. The use
of average daily discharge probably underestimated overbank flow duration by 0.5 - 1
day per overbank event. Some short overbank events (less than 1 day), which probably
exceeded bankfull, were not counted due to their low average daily discharge.
64
The analysis of the drainage basin area effects was biased by the original site
selection, utilizing mean daily discharge records as a requirement for selection. Very few
ephemeral streams are monitored for mean daily discharge, simply because they lack
flow most of the time. In order for most streams to be monitored for mean daily
discharge, they must be at least intermittent to perennial. As climate becomes drier
(higher PET ratio), it takes larger watersheds to provide intermittent or perermial flow.
Due to this lack of consistent flow in small sub-humid watersheds, the driest study sites
had the largest watersheds. Comparable watershed areas on the wet end of the gradient
created streams that were too deep, wide, and fast flowing for the methods used in this
study. As a result of this, El averages increased to drier conditions as drainage basins
increased in area.
Vegetation
Precipitation had the largest effect on vegetation of the parameters examined.
Drier climate resulted in lower groundwater levels, resulting in the contraction and
eventual exclusion of the proportion of flood-prone width that was forested. As
precipitation decreased, the number of non-graminoid species increased. Vegetation in
FPAs changed from woody and forb dominated to those dominated by graminoids. This
transition is at least partially influenced by current and historical agricultural /
silvicultural practices. The majority of the central plains has historically been grazed
and, in some areas, overgrazed by cattle (Donahue 1999). This grazing may have
reduced the diversity of the FPA vegetation, especially on floodplains on the drier end of
65
the gradient where grazing pressure may have eliminated or reduced forbs and woody
plants.
Ofnote, the encroachment of agriculture into the FPW decreased the width of
forested floodplain along some streams. This encroachment undoubtedly changed the
outcome of the vegetation analysis through alteration in species composition and their
distribution in the natural gradient from the streambank to the edge of the FPW.
Encroachment by cropland reduced the species richness of vegetation for large portions
of some FPAs. This encroachment has also changed the FPA average ecological index of
vegetation, depending on the amount of area under cultivation. In addition to this, during
the period of cultivation, the soils may be dry enough to allow some crops (all UPL
species) to survive for at least their growing period, seriously influencing the
interpretation of vegetation analysis during this period.
Neither the duration of average annual overbank flow nor the duration of above
average discharge affected the distribution ofwoody vegetation. At the maximum
observed overbank flow duration of 2.2 percent of the growing season, duration of
inundation does not limit the distribution ofwoody vegetation (Bedinger 1979). The
occurrence of two species that were affected by stream order {Ulmus rubra and Acer
saccharinum) showed a greater sensitivity to stream discharge regimes than the duration
of inundation. Woody species on the North Carolina floodplains showed greater
sensitivity to PET ratio differences than the herbaceous species. Most annual herbaceous
species respond to recent hydrologic regimes, whereas woody species respond to longer-
term regimes due to their perennial life cycles.
66
Of the studied central plain floodplains and several others visited, very few had
shrubs on them. The few that did were mainly Ligustrum spp. Cattle grazing may have
had a negative impact on shrubs due to high seedling herbivory throughout the central
plains. Saplings and tall herbaceous species have in many cases filled the niche of the
shrub layer. The FPAs ofmany observed streams in the central plains are quite open
after the herbaceous layer dies offduring the winter.
Relative densities of vegetation would have been useful in the analysis of the
vegetation data. However, time constraints precluded the collection of these data. This
became evident as analysis was done and there were unanswered questions about the
trends in vegetation dynamics over the climatic gradient. A prevalence index (weighted
averages, Sharmon Index, or Simpson Index) may have shown this better than the average
El.
Ifpossible, the effect and occurrence ofbeaver damming action should be studied
on a PET ratio gradient. This would lead to an understanding of the historic and current
effects ofbeaver on wetland occurrence and stream stability.
Conclusions
• On the central plain sites the exponential trend in the percentage of the FPW that is
wetland indicates a positive feedback situation between the groundwater system and
streamflow.
• Greater discharges from upstream variable source areas, resulting in higher stream
stages, decrease the hydraulic gradient through the floodplain. The reduction in
hydraulic gradient increases the duration of saturation in the floodplain soils.
67
• Along North Carolina streams, high groundwater levels maymaintain riparian
wetlands.
• Precipitation was more important than PET ratio in determining woody species
richness in the central plains.
• Drainage basin area, stream order, overbank flow duration, and soil texture did not
significantly correlate with the maintenance or development of riverine wetlands on
the central plain study sites.
References
Abrams, M. D. 1985. Age diameter relationships of Ouercus species in relation to
edaphic factors in gallery forests in Northeast Kansas. Forest Ecology and
Management 13:181-193.
Abrams, M. D., and D. J. Gibson. 1991. Effects of fire exclusion on tallgrass prairie and
gallery forest communities in eastern Kansas, pp. 3-10. In S.C. Nodvin and T.A.
Waldrop (eds). Fire and the environment: Ecological and cultural perspectives,
proceedings of and international symposium. Southeastern Forest Experiment
Station, Asheville, NC.
Auble, G.T., J.M. Friedman, and M.L. Scott. 1994. Relating Riparian vegetation to
present and future streamflows. Ecological Applications 4:544-554.
Agnew, C. and E. Anderson. 1992. Water resources in the arid realm. Routledge,
London.
Bailey, R.G. 1976. Ecoregions of the United States. USDA Forest Service,
Intermountain Region, Ogden, Utah. U.S. GPO 777-152.
Bedinger, M.S. 1979. Forests and flooding with special reference to the White River and
Ouachita River basins, Arkansas. U.S. Geological Survey Open File Report 79-
68.
Brinson, M.M. 1993a. A hydrogeomorphic classification for wetlands. Technical
Report WRP-DE-4, U.S. Army Engineer Waterway Experiment Station,
Vicksburg, MS.
Brinson, M.M. 1993b. Changes in the functioning ofwetlands along environmental
gradients. Wetlands 13:65-74.
Burke, M.A., and W.L. Nutter. 1995. Channel morphology evolution and riverine
wetland hydrology in the Georgia piedmont, pp. 463-475. In K.L. Campbell
(ed). Versatility ofwetlands in the agricultural landscape. American Society of
Agricultural Engineers. Tampa, FL.
Cashin, G.E. 1990. Wetland development in the North Carolina coastal plain:
presettlement to 1980’s. Masters Thesis. Duke University, Durham, NC.
Cole, C.A., R.P. Brooks, and D.H. Wardrop. 1997. Wetland hydrology as a function of
hydrogeomorphic subclass. Wetlands 17:456-467.
69
Diers, R. and J.L. Anderson. 1984. Development of soil mottling. Soil Survey
Horizons, winter 1984, pp. 9-12.
Donahue, D.L. 1999. The western range revisited. University of Oklahoma Press,
Norman, OK. p. 64.
Earthinfo. 1995a. NCDC Summary of the Day. Boulder, Colorado, CD-ROM.
Earthinfo. 1995h. USGS Daily Values. Boulder, Colorado, CD-ROM.
Environmental Laboratory. 1987. Corps of Engineers wetlands delineation manual.
Technical report Y-87-1, US Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Fetter, C.W. Jr. 1980. Applied hydrogeology. Charles E. Merril, Columbus, OH. pp.
42-44.
Fetter, C.W. Jr. 1994. Applied hydrogeology, third edition. Prentice Hall, Upper Saddle
River, NJ. p. 57.
Gambrell, R.P. and W.H. Patrick. 1978. Chemical and microbial properties of anaerobic
soils and sediments, pp. 375-423. In Hook, D.D. and R.M.M. Crawford (eds).
Plant life in anaerobic environments. Ann Arbor Sci. Pub., Ann Arbor, MI.
Gray, D.M., G.A. McKay, J.M. Wigham. 1970. Energy, evaporation, and
évapotranspiration, pp. 3.1-3.66. In Gray, D.M. Handbook on the principles of
hydrology. Water Information Center, Port Washington, NY.
GretagMacbeth. 1996. Munsell soil color charts. GretagMacbeth, New Windsor, NY.
Heath, R.C. 1983. Basic ground-water hydrology. USGS water-supply paper; 2220,
Denver, CO.
Hewlett, J.D. 1961a. Soil moisture as a source ofbaseflow. USDA Forest Service
Station paper no. 132, Asheville, NC.
Hewlett, J.D. 1961b. Watershed Management, pp. 61-66. /« Annual report for 1961.
USDA Forest Service. Southeastern For. Exp. Sta. Asheville, NC.
Holdridge, L.R., W.C. Grenke, W.H. Hatheway, T. Liang, and J.A. Tosi, Jr.. 1971.
Forest environments in tropical life zones: a pilot study. Pergamon Press,
Elmsford, NY. p. 9.
70
Johnson, W.C. 1998. Adjustment of riparian vegetation to river regulation in the Great
Plains, USA. Wetlands 18:608-618.
Knighton. D. 1998. Fluvial forms and proeesses: a new perspeetive. John Wiley and
sons. New York, NY. pp. 68-71.
Kuchler, A.W. 1967. Vegetation Mapping. Ronald Press, New York, NY.
Larson, J.S., M.S. Bedinger, C.F. Bryan, S. Brown, R.T. Huffman, E.L. Miller, D.G.
Rhodes, and B.A. Touchet. 1981. Transition from wetlands to uplands in
southeastern bottomland hardwood forests, pp. 225-273. In Clark, J.R. and J.
Benforado (eds). Wetlands of bottomland hardwood forests. Elsevier Scientific
Publishing, NY.
Leopold, L.B., M.G. Wolman, and J.P. Miller. 1964. Fluvial Processes in
Geomorphology. W.H. Freeman, San Francisco, CA.
Leopold, L. B. 1994. A View of the River. Harvard University Press, Cambridge,
MA.
Mitsch, W. J. and J.G. Gosselink. 1993. Wetlands. Van Nostrand Reinhold, New York,
NY.
NWS (National Weather Service). 1995. Disasterous floods on the Trinity, Brazos,
Colorado, and Guadalupe rivers in Texas, December 1991 - January 1992. U.S.
Dept ofCommerce, NOAA, National Weather Service, Silver Springs, MD.
NWS. 1997. Flood Forecasting, www.nws.noaa.gov.
Odum, E.P. 1971. Fundamentals of ecology, p. 144. W.B. Saunders, Philadelphia,
PA.
Patterson, G.G., G.K. Sperry, and B.J. Whetstone. 1985. Hydrology and its effect on
distribution of vegetation in Congaree Swamp National Monument, South
Carolina. U.S. Geological Survey Water-Resources Investigations Report 85-
4256. Columbia, SC.
Reed, P.B. 1988. National list ofplant species that occur in wetlands: 1988 national
summary. Biological Report 88(24), US Fish and Wildlife Service, Washington,
D.C.
Reichman, O.J. 1987. Konza prairie: a tallgrass natural history. University Press of
Kansas. Lawrence, KS. p 121.
71
Rheinhardt, R. D., M. C. Rheinhardt, M. M. Brinson, and K. Fraser. 1998. Forested
wetlands of low order streams in the inner coastal plain ofNorth Carolina, USA.
Wetlands 18:365-378.
Rice, E.L. 1965. Bottomland forests of north-central Oklahoma. Ecology 46:708-714.
Riggs, H.C. and K.D. Harvey. 1990. Temporal and spatial variability of streamflow. pp.
81-96. /« Wolman, M.G. and H.C. Riggs (eds). 1990. Surface water hydrology.
Geological Society ofAmerica, Inc., Boulder, CO.
Rosgen, D. 1996. Applied river hydrology. Wildland Hydrology, Minneapolis, MN.
Shaffoth, P.B., G.T. Auble, J.C. Stromberg, and D.T. Patten. 1998. Establishment of
woody riparian vegetation in relation to annual patterns of streamflow. Bill
Williams River, Arizona. Wetlands 18:577-590.
Stolt, M.H., and J.C. Baker. 1995. Evaluation ofNational Wetlands Inventory maps to
inventory wetlands in the southern Blue Ridge ofVirginia. Wetlands 15:346-353.
Strahler, A.N. 1957. Quantitative analysis ofwatershed geomorphology. American
Geophysical Union Transactions 38:913-920.
Taylor, J.P., D.B. Webster, and L.M. Smith 1999. Soil disturbance, flood management,
and riparian woody plant establishment in the Rio Grande floodplain. Wetlands
19:372-382.
Thein, S.S. 1979. A flow diagram for teaching texture-by-feel analysis. Journal of
Agronomic Education 40:54-55.
Thomthwaite, C.W. and B. Holzman 1942. Measurement of évapotranspiration from
land and water surfaces. U.S. Dept, ofAgriculture, Washington, D.C.
Tiner, R.W. 1991. The concept of a hydrophyte for wetland identification. BioScience
41:236-247.
Trimble, S.W. 1970. The Alcovy River swamps: the result of culturally accelerated
sedimentation. Bulletin of the Georgia Academy of Science 28:131-141.
USDC. 1968. Climatic Atlas of the United States. NOAA. ESSA, EDS. U.S. Govt
Print. Off, Washington, D.C.
U.S. Fish and Wildlife Service. 1998a. National Wetland Inventory maps. FWS,
ftp://’www.nwi.fws.gov/arcdata/.
72
U.S. Fish and Wildlife Service. 1998b. National Wetland Inventory map, Oldford.eOO.
FWS, ftp://www.nwi.fws.gov/arcdata/raleigh/.
Wentworth, T.R., G.P. Johnson, and R.I. Kologiski. 1988. Designation ofwetlands by
weighted averages of vegetation data: a preliminary evaluation. Water Resources
Bulletin, 24:389-396.
Williams, G.P. 1978. Bank-full discharge of rivers. Water Resources Research,
14:1141-1154.
Appendix A. Floodplain Vegetation on the Central Plain Study Sites
Appendix A. Floodplain vegetation on the central plain study sites. + = prevalent, X = dominant, H = herb, V = vine, S = shrub.
T = tree, n/a = not available. Note: some species are listed more than once with different wetland indicator status due to differing
regional indicator status. Jaccard similarity calculated by comparison with the next higher PET ratio site (1 vs. 2, 2 vs. 3, etc).
Site id. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78 0.78 0.78 0.78 0.79 0.80
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5
Jaccard similarity 0.56 0.45 0.55 0.63 0.74 0.67 0.54 0.69 0.57 0.50 0.69 0.76 0.67 0.56
2
Drainage area (km ) n/a 487 606 119 69 136 n/a 221 128 n/a 44 715 287 624
Species Indicator status
Cannabis sativa FACU- H
Agropyron smithii FACU H
Amaranthus albus FACU H
Ambrosia artemisiifolia FACU H + + + + + + + + + + + -1- -1- -1-
Elymus canadensis FACU H
Helianthus annuus FACU H
Rudbeckia hirta FACU H
Solidago nuttallii FACU H
Solidago nuttallii FACU+ H + - - ~
Elymus canadensis FAC- H - - - + - - + -1-
Phytolacca americana FAC- H - - + ~ - - + - - + -t- - - -
Asclepias speciosa FAC H -t- + - -
Cannabis sativa FAC H + + - - -
Commelina communis FAC H
Elymus virginicus FAC H
Panicum rigidulum FAC H
Phytolacca americana FAC H
Rhus toxicodendron FAC H ~ +
Verbesina alternifolia FAC H
Hordeum jubatum FAC + H +
Appendix A. Continued.
Site id. 15 16 17 18 19 20 21 22 23 24 25 26 27 28
PET ratio 0.80 0.80 0.80 0.85 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98 1
Strahler order 5 5 6 4 5 5 3 4 5 5 5 3 4 4
Jaccard similarity 0.57 0.30 0.48 0.64 0.56 0.72 0.42 0.29 0.61 0.44 0.38 0.38 0.52 0.4
Drainage area (km ) n/a 824 407 n/a 777 308 n/a n/a 891 894 124 11 171 102
Species Indicator status
Cannabis sativa FACU- H -f- + + 4- 4-
Agropyron smithii FACU H X ~ —
Amaranthus albus FACU H
Ambrosia artemisiifolia FACU H + X ~ + — — 4- — ~ -
Elymus canadensis FACU H ~ — + + 4- ~ -
Helianthus annuus FACU H
Rudbeckia hirta FACU H
Solidago nuttallii FACU H
Solidago nuttallii FACU+ H
Elymus canadensis FAC- H +
Phytolacca americana FAC- H +
Asclepias speciosa FAC H 4- 4-
Cannabis sativa FAC H +
Commelina communis FAC H
Elymus virginicus FAC H 4- ~ + X
Panicum rigidulum FAC H ~ + 4- 4- — —
Phytolacca americana FAC H
Rhus toxicodendron FAC H
Verbesina alternifolia FAC H — — — X — X ~ — — 4- 4- X — —
Hordeum jubatum FAC + H
Appendix A. Continued.
Site id. 29 30 31 32 33 34 35 36
PET ratio 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Strahler order 6 5 3 4 4 4 5 4
Jaccard similarity 0.58 0.55 0.38 0.67 0.75 0.27 0.20 <—
Drainage area (km ) 995 723 n/a 932 1160 883 n/a n/a
Species Indicator status
Cannabis sativa FACU- H
Agropyron smithii FACU H X
Amaranthus albus FACU H +
Ambrosia artemisiifolia FACU H
Elymus canadensis FACU H
Helianthus annuus FACU H -1- -1-
Rudbeckia hirta FACU H -1- -
Solidago nuttallii FACU H
Solidago nuttallii FACU+ H
Elymus canadensis FAC- H
Phytolacca americana FAC- H
Asclepias speciosa FAC H + -
Cannabis sativa FAC H
Commelina communis FAC H
Elymus virginicus FAC H + - - -1- -1- -1- - -
Panicum rigidulum FAC H
Phytolacca americana FAC H 4-
Rhus toxicodendron FAC H ~ - +
Verbesina alternifolia FAC H
Hordeum jubatum FAC + H
Appendix A. Continued.
Site id. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78 0.78 0.78 0.78 0.79 0.80
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5
Jaccard similarity 0.56 0.45 0.55 0.63 0.74 0.67 0.54 0.69 0.57 0.50 0.69 0.76 0.67 0.56
Drainage area (km ) n/a 487 606 119 69 136 n/a 221 128 n/a 44 715 287 624
Species Indicator status
Rhus radicans FAC + H + + + + + + + ..
Urtica gracilis FAC + H X + X + X X + + X + + + +
-
FACW- H + + + -Elymus virginicus
Ipomoea lacunosa FACW- H
Hordeum jubatum FACW H - - +
Impatiens sp. FACW H + + + X + + - -- + -
- -
- -
Ipomoea lacunosa FACW H + + -
Laportea canadensis FACW H + X + - - + + X + X - - X +
Pilea pumila FACW H - X + + - X + + + +
Polyganum sp. FACW H + - - + +
Urtica gracilis FACW H
Verbesina alternifolia FACW H - - - + - - - +
Phalaris arundinacea FACW+ H + ~ + + - + + -- - X + - -
Alisima plantago-aquatica OBL H + -
Boehmeria cylindrica OBL H - - - X +
Calamagrostis canadensis OBL H "
Parthenocissus quinquefolia FAC V + + —„
Smilax rotundifolia FAC V + + +
Smilax hispida FAC V + + +
Appendix A. Continued.
Site id. 15 16 17 18 19 20 21 22 23 24 25 26 27 28
PET ratio 0.80 0.80 0.80 0.85 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98 1
Strahler order 5 5 6 4 5 5 3 4 5 5 5 3 4 4
Jaccard similarity 0.57 0.30 0.48 0.64 0.56 0.72 0.42 0.29 0.61 0.44 0.38 0.38 0.52 0.4
Drainage area (W) n/a 824 407 n/a 777 308 n/a n/a 891 894 124 11 171 102
Species Indicator status
Rhus radicans FAC + H + + - -
Urtica gracilis FAC + H +
Elymus virginicus FACW- H
Ipomoea lacunosa FACW- H
Hordeum -jubatum FACW H + +
Impatiens sp. FACW H + -
Ipomoea lacunosa FACW H
-
Laportea canadensis FACW H X - + X + - - - + + X - -
Pilea pumila FACW H + +
-
Polyganum sp. FACW H + - -
Urtica gracilis FACW H - X X X + + + - + + - - + +
Verbesina alternifolia FACW H -
Phalaris arundinacea FACW+ H + + X X X - -- ~ - + - + +
Alisima plantago-aquatica OBL H +
Boehmeria cylindrica OBL H + + X -
Calamagrostis canadensis OBL H — +
Parthenocissus quinquefolia FAC V
Smilax rotundifolia FAC V + - - - + +
Smilax hispida FAC V +
00
Appendix A. Continued.
Site id. 29 30 31 32 33 34 35 36
PET ratio 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Strahler order 6 5 3 4 4 4 5 4
Jaccard similarity 0.58 0.55 0.38 0.67 0.75 0.27 0.20
Drainage area (km ) 995 723 n/a 932 1160 883 n/a n/a
Species Indicator status
Rhus radicans FAC + H + + +
Urtica gracilis FAC + H
Elymus virginicus FACW- H
Ipomoea lacunosa FACW- H
Hordeum jubatum FACW H - - +
Impatiens ~sp. FACW H +
Ipomoea lacunosa FACW H
Laportea canadensis FACW H
Pilea pumila FACW H
Polyganum sp. FACW H - +
Urtica gracilis FACW H + X + - - -- ~
Verbesina alternifolia FACW H
Phalaris arundinacea FACW+ H ~ X
Alisima plantago-aquatica OBL H ~ +
Boehmeria cylindrica OBL H
Calamagrostis canadensis OBL H
Parthenocissus quinquefolia FAC V
Smilax rotundifolia FAC V
Smilax hispida FAC V
Appendix A. Continued.
Site id. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
PET ratio 0.71 0.72 0.72 0.72 0.72 0.73 0.73 0.73 0.78 0.78 0.78 0.78 0.79 0.80
Strahler order 5 5 5 3 5 5 3 4 4 5 4 5 4 5
Jaccard similarity 0.56 0.45 0.55 0.63 0.74 0.67 0.54 0.69 0.57 0.50 0.69 0.76 0.67 0.56
2
Drainage area (km ) n/a 487 606 119 69 136 n/a 221 128 n/a 44 715 287 624
Species Indicator status
Aesculus glabra FAC S + 4- - - 4-
Ligustrum -sp. FAC s - - - + + 4-
Staphylea trifolia FAC s — + 4-
Celtis occidentalis FACU T 4-
Juglans nigra FACU T + - - -t- -1- -1- 4- 4- 4- 4- - +
-
-
Morus rubra FACU T 4-
Quercus - -macrocarpa FACU T - + -1- -1- 4- 4- - 4- 4-
- + -
Celtis occidentalis FAC- T - + - -- + - 4- 4- - - 4- - 4-
Morus rubra FAC- T + 4- 4- 4- 4- 4- + 4- --
Quercus +macrocarpa FAC- T
Acer negando FAC T +
Gleditsia triacanthos FAC T 4- 4- - -
Platanus occidentalis FAC T + + X -t- -- + - - - -— 4- 4-
Populas sargenta FAC T - -1- - -f- + 4- + -h - 4- - 4- ~ -
Ulmus americanas FAC T
Ulmus rubra FAC T + -t - -1- + 4- + + X 4- X - X 4-
Populas sargenta FAC+ T 4-
Acer negando FACW- T -1- X + -1- + X -h + 4- - - 4- -
Acer saccharinum FACW T X X X 4- X 4- - X 4- X - 4- + 4-
Platanus occidentalis FACW T
Salix exigua OBL T -1- - - -1- + 4- 4- 4- - X
-
- + -
Salix rígida OBL T 00o
Appendix A. Continued.
Site id. 15 16 17 18 19 20 21 22 23 24 25 26 27 28
PET ratio 0.80 0.80 0.80 0.85 0.87 0.88 0.89 0.89 0.90 0.93 0.93 0.95 0.98 1
Strahler order 5 5 6 4 5 5 3 4 5 5 5 3 4 4
Jaccard similarity 0.57 0.30 0.48 0.64 0.56 0.72 0.42 0.29 0.61 0.44 0.38 0.38 0.52 0.4
Drainage area (km ) n/a 824 407 n/a 777 308 n/a n/a 891 894 124 11 171 102
Species Indicator status
Aesculus glabra FAC S
-
Ligustrum - - - +sp. FAC S - - + + +
Staphylea trifolia FAC S
Celtis occidentalis FACU T X X + + -1- .. X
FACU T + + X - - -Juglans nigra - - + +
Morus rubra FACU T - + - + + + + - + - + - - -
Quercus macrocarpa FACU T
Celtis occidentalis FAC- T +
Morus rubra FAC- T +
- -
Quercus +macrocarpa FAC- T
Acer negando FAC T -- + + - + - + -1-
Gleditsia triacanthos FAC T ~ - + + X
Platanus occidentalis FAC T +
Populas sargenta FAC T
Ulmus americanas FAC T - +
Ulmus rubra FAC T + - - + + + + - + + X -1- X +
-
Populas - -sargenta FAC+ T - + + + + -- - + X +
Acer negando FACW- T +
Acer saccharinum FACW T X X ~ + - X + X +
Platanus occidentalis FACW T ~ — +
Salix exigua OBL T + -I- X
Salix rigida OBL T 00
Appendix A. Concluded.
Site id. 29 30 31 32 33 34 35 36
PET ratio 1.05 1.13 1.17 1.21 1.25 1.31 1.35 1.75
Strahler order 6 5 3 4 4 4 5 4
Jaccard similarity 0.58 0.55 0.38 0.67 0.75 0.27 0.20 ?—
Drainage area (km^) 995 723 n/a 932 1160 883 n/a n/a
Species Indicator status
Aesculus glabra FAC S
Ligustrum sp. FAC s - + + - - + - -
Staphylea trifolia FAC s
Celtis occidentalis FACU T — -t- — — + + — —
Juglans nigra FACU T + + +
Moms mbra FACU T X +
Quercus macrocarpa FACU T
Celtis occidentalis FAC- T
Moms mbra FAC- T
Quercus macrocarpa FAC- T
Acer negando FAC T - + +
Gleditsia triacanthos FAC T X - -
Platanus occidentalis FAC T
Populas sargenta FAC T
Ulmus americanas FAC T
Ulmus mbra FAC T + X + X - + - -
Populas sargenta FAC+ T + + X + X + X -
Acer negando FACW- T
Acer saccharinum FACW T
Platanus occidentalis FACW T
Salix exigua OBL T
Salix risida OBL T + — 00
to
Appendix B. Floodplain Vegetation on the North Carolina Study Sites
Appendix B. Floodplain vegetation on the North Carolina study sites. + = prevalent, T = tree, S = shrub, H = herb, F = fern
— = not prevalent.
Site id. A B C D E F G H I J K L
PET ratio 0.67 0.67 0.75 0.75 0.76 0.76 0.76 0.78 0.78 0.78 0.78 0.81
Stream order 2 2 1 1 1 1 1 3 4 2 4 1
Percent wetland 100 100 100 100 100 100 100 100 100 96 84 100
Drainage area (km ) 0.31 1.37 0.14 0.33 0.02 0.05 0.36 4.21 5.92 0.18 12.75 0.92
Species Indicator status
Acer saccharum FACU- T +
Oxydetidrum arboreum FACU T
Quercus alba FACU T -
Ilex opaca FAC- T + - - + + +
+
Carpinus caroliniana FAC T + - - - - -
Liriodendron tulipifera FAC T
Pinus taeda FAC T ~
Liquidambar styraciflua FAC+ T + - + +
Magnolia grandifolia FAC+ T - - - - +
Quercus mitchauxii FACW- T + - - - - -
Betula nigra FACW T + -
Magnolia virginiana FACW+ T -
Acer rubrum OBL T + + - + + - - + + + + -
Nyssa aquatica OBL T - + + - + + - + + + + +
Nyssa biflora OBL T - + + - + + + + + + + +
Taxodium distichum OBL T + + - + —
Clethra alnifolia FACW S — „ — + +
Ligustrum sinense FAC S + -
Persea borbonea FACW s
00
Appendix B. Continued.
Site id. M N O P
PET ratio 0.81 0.81 0.82 0.83
Stream order 2 1 2 1
Percent wetland 100 100 100 100
Drainage area (km ) 0.57 0.13 0.65 0.04
Species Indicator status
Acersaccharum FACU- T
Oxydendrum arboreum FACU T - -1- - +
Quercus alba FACU T - ~ - +
Ilex -opaca FAC- T -t- -1- -
Carpinus caroliniana FAC T - - - -
Liriodendron tulipifera FAC T -1- - - -
Pinus taeda FAC T + - - -1-
Liquidambar styraciflua FAC+ T + -1- - +
Magnolia grandifolia FAC+ T - - - -
Quercus mitchauxii FACW- T - - - -
Betula nigra FACW T
Magnolia virginiana FACW+ T - - - +
Acer rubrum OBL T + - + +
Nyssa aquatica OBL T + -1- - -
Nyssa biflora OBL T - -1- - +
Taxodium distichum OBL T — — — —
Clethra alnifolia FACW S — — — +
Ligustrum sinense FAC S + - - -
Persea borbonea FACW s + — — —
00
LAI
Appendix B. Continued.
Site id. A B C D E F G H I J K F
PET ratio 0.67 0.67 0.75 0.75 0.76 0.76 0.76 0.78 0.78 0.78 0.78 0.81
Stream order 2 2 1 1 1 1 1 3 4 2 4 1
Percent wetland 100 100 100 100 100 100 100 100 100 96 84 100
2
Drainage area (km ) 0.31 1.37 0.14 0.33 0.02 0.05 0.36 4.21 5.92 0.18 12.75 0.92
Species Indicator status
Phytolaca americana FACU+ H -
Eupatorium compositifolium FAC- H
Cypripedium acaule FAC H - - - +
Euphorbia heterophylla FAC H - + - -
Rhus radicans FAC H + + + + -
Arundinaria gigantea FACW H + - + + + - + - ~ - + +
Boehmeria cylindrica FACW H ~ + + + + - +
Impatiens sp. FACW H - + -
Saururus cernuus OBF H — + + + " ” — + + + — —
Smilax hispida FAC V —
Smilax rotundifolia FAC V -h + + + + +
Woodwardia aereolata OBL F + + + + __ +
00
ON
Appendix B. Concluded.
Site id. M N O P
PET ratio 0.81 0.81 0.82 0.83
Stream order 2 1 2 1
Percent wetland 100 100 100 100
2
Drainage area (km ) 0.57 0.13 0.65 0.04
Species Indicator status
Phytolaca americana FACU+ H + - - -
Eupaíorium compositifolium FAC- H -1- - - -
Cypripedium acaule FAC H - - - -
Euphorbia heterophylla FAC H - - - -
Rhus radicans FAC H + - - -
Arundinaria gigantea FACW H + + -1- +
Boehmeria cylindrica FACW H - - - -
Impatiens sp. FACW H - - + -
Saururus cernuus OBL H — — + "
Smilax hispida FAC V -t- — „ —
Smilax rotundifolia FAC V - -1- -1- —
Woodwardia aereolata OBL F -1- -1- -1- -t-
00
-J