Flood tolerance of plant species in bottomland forests of the southeastern United States

MISSING IMAGE

Material Information

Title:
Flood tolerance of plant species in bottomland forests of the southeastern United States
Physical Description:
xiii, 208 leaves : ill., maps ; 29 cm.
Language:
English
Creator:
Theriot, Russell F
Publication Date:

Subjects

Subjects / Keywords:
Forests and forestry -- South Atlantic States   ( lcsh )
Plants -- Effect of water levels on -- South Atlantic States   ( lcsh )
Forest Resources and Conservation thesis Ph. D
Dissertations, Academic -- Forest Resources and Conservation -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 202-206)
Statement of Responsibility:
by Russell Francis Theriot.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001757974
oclc - 26699662
notis - AJH1021
System ID:
AA00003281:00001

Full Text
















FLOOD TOLERANCE OF PLANT SPECIES
IN BOTTOMLAND FORESTS OF THE
SOUTHEASTERN UNITED STATES













BY

RUSSELL FRANCIS THERIOT


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1992


UNIVERSITY OF FLORInA 11".-1-




































Digitized by the Internet Archive
in 2011 with funding from
University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation

To my parents


http://www.archive.org/details/floodtoleranceof00ther















ACKNOWLEDGMENTS


This research effort was greatly aided by innumerable people.

Foremost among them is Dr. Dana Sanders, who assisted me in all phases

of the study. His advice and encouragement throughout this effort are

greatly appreciated. I thank Drs. Ken Rodgers, Dan Evans, and Tom Hein-

eke for their assistance in identifying the plant species in this study.

Blake Parker was invaluable in helping me interpret the soils in the

study. Don Hill conducted the geodetic surveys for all of the sites.

Phil Jones and Jeff Irvin taught me all I know about hydrology. Phil

was especially helpful in correcting the water surface elevations

between the study sites and the gauging stations, and Jeff interpreted

the hydrologic program into FORTRAN language. I am also grateful to

Dr. Dara Wilber, who assisted me in the statistical analyses of the

data.

Many people provided helpful discussion and advice on various

aspects of the study, including Drs. Donal Hook, Bill Patrick,

Helen Leitman, Sandra Brown, and Bill Mitsch.

I gratefully acknowledge the support and understanding of Dr. Bob

Engler, who allowed a work schedule flexible enough to finish the

manuscript.

I especially want to thank my wife and children, who encouraged me

to go on when it would have been easy to quit.








Finally, I would like to thank my major professor, Dr. Jerome

Shireman, for his advice, encouragement, and support in guiding my

degree program, and also my committee for their helpful suggestions and

patience.















TABLE OF CONTENTS


Page


ACKNOWLEDGMENTS . . .

LIST OF TABLES . . .

LIST OF FIGURES . . .

ABSTRACT . . .

INTRODUCTION .........

Plant Community Organization ...
Bottomland Forest Community Organization .
Zonation of Bottomland Forests .
Purpose and Objectives . .

METHODS . . .

Study Area . . .
Site Selection .. .
Determining Hydrologic Zone Elevations .
Site Preparation and Data Collection .
Analyzing Vegetation Data . .
Calculating Species FTI Numbers .

RESULTS AND DISCUSSION . .

Flood Analysis of Study Sites . .
Vegetation Data . .
Weighted Averaging . .
Statistical Analysis of the Vegetation Data .
Cluster Analyses . .
Discriminant Function Analysis .
Regional Variation in Species FTI Numbers .

SUMMARY AND CONCLUSIONS . .

APPENDIX A SITE DESCRIPTIONS AND MAP LOCATIONS .

APPENDIX B GUIDE FOR COMPUTER PROGRAM FOR ANALYZING
HYDROLOGIC DATA . .

APPENDIX C HYDROGRAPH FOR STEELE BAYOU (SITE 3) .

v








APPENDIX D IMPORTANCE VALUES FOR SPECIES BY ZONE AND
VEGETATION LAYER ................ 102

APPENDIX E FTI PLANT LIST AND COMPARISON WITH TWO OTHER
WATER-TOLERANCE RATING SYSTEMS ... 186

REFERENCES . . ... .. .202

BIOGRAPHICAL SKETCH . . ... 207















LIST OF TABLES


Page


Table


1 Hydrologic Zones Occurring in Bottomland Forests of the
Southeastern United States ... .... .12

2 Annual Flood Frequency (Percent of Years in Which
Boundary Is Exceeded at Least Once during Growing
Season for More than 7 days) for Zone Boundaries .. 21

3 Average Annual Duration of Flood Events (days)
for Zone Boundaries . . ... .23

4 FTI Numbers of Commonly Occurring Species in this
Study .. .. . .... 25

5 Variations in Species Flood Tolerance Index
Numbers According to Life Stage . ... 30

6 Comparison of Three Water-Tolerance Ratings for
Selected Bottomland Forest Tree Species ... 34

7 Relative Frequencies in Each Hydrologic Zone of Tree
Species Used in the Statistical Analyses; Groupings
of Species Correspond to Cluster Membership ..... .40

8 Relative Frequency of Occurrence of Each Sapling
Species in the Hydrologic Zones along with Their
Cluster Memberships . ... ... 45

9 Relative Frequencies in Each Hydrologic Zone of the
Vine Species Used in Statistical Analyses ... .. 47

10 Mean Importance Values for Species in Each Cluster Used
in the DFA, Arranged by Zone/Sample . .. 50

11 Predicted Hydrologic Zones (Columns) and Actual
Zones (Rows) Based on DFA Results Using Only
Tree Importance Values . ... .51

12 Predicted Hydrologic Zones (Columns) and Actual
Zones (Rows) Based on DFA Results Using Average
FTI Values for All Observed Tree Species at the Site 52










Table


13 Cross-Validation Results of Zone Membership Using


Linear Discriminant Function Analysis .

D-l Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .


. . 53


at Site 1,
. .


D-2 Importance Values for Species Occurring at Site 2,
Arranged by Zone and Vegetation Layer .

D-3 Importance Values for Species Occurring at Site 3,
Arranged by Zone and Vegetation Layer .

D-4 Importance Values for Species Occurring at Site 4,
Arranged by Zone and Vegetation Layer .


D-5 Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .

D-6 Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .


at Site 5,


at Site 6,


. 103


. 108


. 121


. 124


. 127


D-7 Importance Values for Species Occurring at Site 7,
Arranged by Zone and Vegetation Layer .

D-8 Importance Values for Species Occurring at Site 8,
Arranged by Zone and Vegetation Layer .

D-9 Importance Values for Species Occurring at Site 9,
Arranged by Zone and Vegetation Layer .


D-10 Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .

D-11 Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .

D-12 Importance Values for Species Occurring
Arranged by Zone and Vegetation Layer .


at Site 10,


at Site 11,


at Site 12,


. 137


. 141


. 147


. 152


D-13 Importance Values for Species Occurring at Site 13,
Arranged by Zone and Vegetation Layer .

D-14 Importance Values for Species Occurring at Site 14,
Arranged by Zone and Vegetation Layer .

D-15 Importance Values for Species Occurring at Site 15,
Arranged by Zone and Vegetation Layer .


viii


. 168


Page








Table Page

D-16 Importance Values for Species Occurring at Site 16,
Arranged by Zone and Vegetation Layer ... 173

D-17 Importance Values for Species Occurring at Site 17,
Arranged by Zone and Vegetation Layer ... 180

E-l FTI Plant List . . ... .. 187
















LIST OF FIGURES


Figure

1 Zonal classification of bottomland forest wetlands
(adapted from Clark and Benforado 1981) . .

2 Study area and sites in the southeastern United States

3 Representation of a typical research site .

4 Ecological amplitude of some commonly occurring species;
CAAQ: Carya aquatica; FOAC: Forestiera acuminata; FRPE:
Fraxinus pennsylvanica; LIST: Liquidambar styraciflua;
NYAQ: Nyssa aquatica; PITA: Pinus taeda; QUAL: Quercus
alba; QULY: Quercus lyrata; SAAL: Sassafras albidum; QUNI:
Quercus nigra; TADI: Taxodium distichum; ULAM: Ulmus


americana


Cluster diagram for trees. . .

Cluster diagram for saplings and shrubs.

Cluster diagram for saplings alone .

Cluster diagram for shrubs alone .

Cluster diagram for vines . .

Cluster diagram for herbs . .

Mean tree FTI numbers plotted versus observed
predicted hydrologic zones for all 55 sites .

Neches River (sites 1 and 2) .

Steele Bayou (site 3) . .

Ouachita River (sites 4 and 5) .

Yazoo River (site 6) . .

Big Black River (site 7) . .

L'Anguille River (sites 8 and 9) .


. 39

. 42

. 43

. 44

. 46

. 48

and
. 57

. 64

. 66

. 68

. 70

. 72

. 74


Page


. . 32


5

6

7

8

9

10

11



A-1

A-2

A-3

A-4

A-5

A-6








Page


A-7

A-8

A-9

A-10

A-11

A-12

A-13

A-14

B-1

C-1


Pearl River (site 10) . .

Apalachicola River (site 11) . .

Apalachicola River (site 12) . .

Ocmulgee River (site 13) . .

Altamaha River (site 14) . .

Edisto River (site 15) . .

Lynches River (site 16) . .

Waccamaw River (site 17) . .

Program logic for computation of days saturated .

Hydrograph for Steele Bayou (site 3); the shaded
areas represent the nongrowing season, a: 1961-1965,
b: 1966-1970, c: 1971-1975, d: 1976-1980 .


S. 76

S. 78

S. 79

. 81

S. 83

. 85

S. 87

. 89

. 95




. 98














Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

FLOOD TOLERANCE OF PLANT SPECIES IN BOTTOMLAND
FORESTS OF THE SOUTHEASTERN UNITED STATES

By

Russell Francis Theriot

May 1992


Chairman: Jerome V. Shireman
Major Department: School of Forest Resources and Conservation


Vegetation data on species composition along a hydrologic gradient

were collected at 17 bottomland forest sites throughout the southeastern

United States. Weighted averages based on importance values calculated

from 55 stands resulted in flood tolerance index (FTI) numbers, the

optimum position for each species along the defined hydrologic gradient,

for 312 identified species.

Commonly occurring species were evaluated using cluster analyses

and discriminant function analyses. Data on tree, sapling, and vine

species clustered into distinct groups, with tree species being the most

reliable; however, shrubs and herbaceous species did not cluster dis-

tinctly. Discriminant function analysis using FTI numbers for tree

species proved to be 82 percent reliable in predicting zones.

The accuracy of the Flood Tolerance Index (FTI) numbers did not

vary regionally in the southeastern United States. Therefore, a single


xii








FTI number calculated for each species can be used to predict hydrologic

zones for the entire study area.


xiii















INTRODUCTION


Bottomland forests are found in the floodplains of rivers in the

southeastern United States from eastern Texas to Virginia. They have

distinct topographic features that are the result of historical hydro-

logic characteristics of the rivers, including periodic fluctuations in

water levels and changes in stream course. Recognizable floodplain

topographic features include first bottoms, second bottoms or terraces,

uplands, riverfront, swamp, poorly drained flats, well-drained flats,

and sloughs (Putnam, Furnival, and McKnight 1960). These features are

characterized by different hydrologic regimes and can be identified as a

hydrologic gradient transitional between permanent water and terrestrial

uplands.

Many studies have previously described the forest communities

associated with these floodplain features (Putnam, Furnival, and

McKnight 1960; Broadfoot and Williston 1973; Chambless and Nixon 1975;

Hodges and Switzer 1979; Mohler 1979; and Hupp and Osterkamp 1985).

However, studies describing the relationship between plant species dis-

tribution and specific inundation/saturation regimes in bottomland for-

ests are rare (Bedinger 1971; Mohler 1979; Huffman 1980; and Leitman,

Sohm, and Franklin 1984). Even so, these studies all demonstrated that

frequency and duration of inundation/saturation exert a controlling

influence on the composition, structure, and distribution of wetland

plant communities. As an example, Bedinger (1971) found a definite

1








relationship between the distribution of plant species and the frequency

and duration of flooding in the lower White River Valley, Arkansas.

Using flood frequency and duration, he defined four species associations

on the White River floodplain, each of which had a distinctly different

tolerance to inundation. He concluded that based on plant species-

flooding relationships, plant communities could be used as a basis to

transfer flooding parameters to ungauged streams.


Plant Community Organization


The concept of community structure has been debated for decades.

Clements (1916) first described communities as discrete, self-organizing

entities that could be considered as discrete organisms. Gleason (1917)

disagreed with Clements' organismal concept and proposed a hypothesis

relating to the individualistic occurrence of plants. His hypothesis

has developed into the continuum concept, which indicates that plant

species distribution is determined by the species' response to its envi-

ronment. Whittaker (1967) and McIntosh (1980) later developed Gleason's

ideas, expanding on the continuum concept. They maintain that since

plant species adapt differently, no two occupy the same zone. This

results in a continuum of overlapping species associations, each

responding to subtly different environmental factors (e.g., water, soil

pH, nutrients, and solar radiation). A continuum can be described for

each factor in various increments or zones.

Zonation simply describes the different levels of an environmental

gradient to which a species is responding. The reason zonation is so

obvious in some ecosystems is that environmental gradients are








"ecologically" steep and groups of species have fairly similar tolerance

that tend to group them on these gradients (Mitsch and Gosselink 1986).

Gleason's individualistic hypothesis can be supported by several

studies (Curtis and McIntosh 1951; Brown and Curtis 1952; Bray 1956;

Whittaker 1956; Curtis 1959; Whittaker and Niering 1965; and Mohler

1979). These studies show that although species have different ecologi-

cal amplitudes and, in fact, do not occupy the same niche, they organize

as units based on similar ecological conditions. Moreover, intergrades

caused by interspecific competition occur between defined types of plant

associations. These intergrades can be attributed to continuous envi-

ronmental variability in time or space or to environmental modification.


Bottomland Forest Community Organization


Van Der Valk (1981) developed a qualitative model of succession in

freshwater wetlands based on the "individualistic" approach to vegeta-

tion proposed by Gleason. He based his approach on three key life his-

tory features of plant species: life-span, propagule longevity, and

propagule establishment requirements. These features are all directly

affected by the flooding on bottomland forests.

Brinson (1990), in discussing the "power line" designation for a

wetland classification developed by Kangas (1990), considered the power

and frequency of inundation as the way in which flood events organize

the plant communities in riverine forests. He characterized the flood

events as high, medium, and low power events with flood power and fre-

quency of inundation being inversely proportional. High power flood

events have a low frequency and determine patterns of the large flood-

plains features (e.g., oxbow lakes, relict levees, and low ridges and








swales) that persist for hundreds to thousands of years. Medium power

flood events, which occur at an intermediate frequency, affect ecosystem

structures that exist from decades to hundreds of years. He identified

tree species associations as an ecosystem component likely to be influ-

enced at this scale. The low power, high-frequency flood events occcur

annually and affect short-term patterns such as seed germination and

seedling survival. His characterization emphasized the dramatic impact

flooding has on the regeneration of vegetation in bottomland forests.

Grubb (1977) stated that scientists have failed to understand ade-

quately how plant communities maintain themselves because of a failure

to account for the phenomenon of regeneration in plant communities.

Huenneke and Sharitz (1986), in a study of microsite abundance and dis-

tribution of woody seedlings in a South Carolina cypress-tupelo swamp,

concluded that the availability and nature of microsites may affect the

distribution and composition of the seedling and sapling strata, thus

differentiating the "regeneration niche" described by Grubb.

Although plant species association is determined by a number of

interacting environmental factors, it is generally agreed that flooding

is the dominant environmental factor at work in bottomland forests,

affecting regeneration and life under saturated soil conditions. Flood-

ing persisting for more than a few days will prevent the replenishment

of soil oxygen once the soil microbes and plant roots consume the avail-

able soil oxygen in the root zone during respiration. Only those plant

species that have evolved a mechanism for living in reducing (anaerobic)

soil conditions will survive such conditions. In most instances, recur-

ring flooding provides a competitive advantage for plant species that

are adapted to saturated and reduced soils.








Chemistry of wet soils (Pearsall and Mortimer 1939; Patrick and

Mikkelsen 1971; Ponnamperuma 1972; Patrick and Delaune 1976; and

Faulkner et al. 1991), and the various physiological effects on vegeta-

tion under reducing conditions are well documented (Cannon and Free

1920; Conway 1940; Dubinina 1961; Hosner and Boyce 1962; Hook and Brown

1973; Hook and Scholtens 1978; Vester 1972; and Hook and Crawford 1980).


Zonation of Bottomland Forests


The hydrologic gradient in bottomland forests ranges from zones of

nearly continuous inundation/saturation in deep swamps to infrequent

inundation/saturation events for brief periods on upland sites. Because

different species respond to different timing and duration of inunda-

tion, a strong correlation exists between the distribution of a species

and its associated hydrologic and soil-moisture conditions (Hosner and

Boyce 1962; Dickson, Hosner, and Hosley 1965; Bedinger 1971, 1978;

Larson et al. 1981; Best, Segal, and Wolfe 1990; and Faulkner et al.

1991). The National Wetlands Technical Council (NWTC) proposed the

zonal classification of floodplain forests (Clark and Benforado 1981).

The classification system defined six hydrologic zones based on fre-

quency and duration of inundation and soil saturation (Figure 1) and

provides the basis for testing in this study.

Larson et al. (1981) summarized the works of others on the occur-

rence of plant species in the Gulf Coastal Plain from 238 belt transects

in Texas, Louisiana, Arkansas, Mississippi, Alabama, and Florida accord-

ing to their maximum tolerance to soil-moisture or hydrologic regimes.

Larson and his cohorts developed a list of 79 tree and shrub species

associated with one or more of the NWTC hydrologic zones. However, the














0




m


0


O
1-
zv


ma

i-


>-


lo
I-
I-0


wz
z


0
0
S0

a.
0 M. r"J

-J

I U
s


aw
oo
>-
JO




LL L.
(U


-4


0
0
0
~sf
on

ar
I-
I--
LLJ
3:Z

0
-1


-J
z




LLS
U,
ui
Zn


- 1-1


E-


in


>-
F-
Z




I-
z


F- i3 zo 0
Sz i--i
0Li ZL.z
|l a -, o
I g 0 Iu
|ln i^ O


w
z
o
N 2


ra
0r
F- 0

S


o F-
Zwiz<
000>-
Owtr
-i i: L Li.
. L.. 0LL
_i n; j i


a
z
Szoz
0 F- Li W: 0
OO0 <
a 0 .. 0 (n


2
w
F-


0
UJ


o
0



0






I
o
I-


0
0

sm a o
0C3 9 IQ
o >





I I
100



1"


N


4 r









0 0
4-1
0)




















U c








-1
u




a,*











ce








list identifies only presence or absence of a species in a zone and does

not identify the ecological amplitude or optimum position of each

species along the hydrologic gradient.


Purpose and Objectives


The purpose of the study was to develop flood tolerance index

(FTI) numbers that refect the optimum position for plant species occur-

ring along the hydrologic gradient in bottomland forests of the south-

eastern United States. The resulting FTI numbers can then be used to

estimate the hydrologic regimes of similar ungauged areas using vegeta-

tion. Specific objectives were to develop methods for translating

recorded hydrologic data into hydrologic zone elevations for southeast-

ern bottomland forests, calculate weighted averages of plant species

based on dominance, and determine methods for applying FTI numbers to

species occurring in bottomland forests of the southeastern United

States.















METHODS


Study Area


The study was conducted in portions of the subtropical ecoregion

of the southeastern United States (Bailey 1980), including portions of

eastern Texas and the Gulf and South Atlantic states. Northern limits

of the area extended across northern Arkansas, Mississippi, Alabama,

Georgia, and South Carolina. The study area included the states of

Louisiana, Arkansas, Mississippi, and Alabama. Georgia and South Caro-

lina were included, except for the piedmont region. Only the extreme

eastern portion of Texas was included, as was the northern portion of

Florida (Figure 2). The intent was to study natural undisturbed sites

encompassing the largest possible area where the resulting FTI numbers

would be applicable without including areas that would introduce too

many additional species or different climatic variables.

Specific sites were selected according to the following criteria:

(1) No major disturbance (e.g., timber harvesting, ditching, or diking)

had occurred during the past 20 years; (2) sufficient hydrologic data

(10 to 20 years of daily stream gauge readings) accurately portraying

water-level fluctuations on the site (considering ponding, tributary

influence between site and gauge, etc.) were available; (3) no site

changes (e.g., timber harvesting or ditching) were anticipated during

the study period; (4) soil data (e.g., soil surveys, soil series, and/or




















-N-


I


LEGEND
0 STUDY SITES


SCALE


200 0


9











16


13
14

12
1








200 400 KM


Figure 2. Study area and sites in the southeastern United States


I 1








soil phases, texture, and permeability coefficients) were available; and

(5) plant communities were characteristic (e.g., plant communities with

few rarely occurring species) of the study area.


Site Selection


Several hundred potential sites were considered, but most were

eliminated because of insufficient stream gauge data. More than 50

sites were visited, but only 17 (Figure 2) satisfied all site criteria

and were used in the study. Although all 17 sites met the selection

criteria, not all hydrologic zones in each site were suitable for study.

Some zones were too narrow and others had been disturbed recently by

agricultural or silvicultural practices.

Sites 1 and 2 were located in the Neches River basin in southeast-

ern Texas. The Steele Bayou, Yazoo River, and Big Black River basins in

Mississippi, respectively, were designated sites 3, 6, and 7. Sites 4

and 5 were located in the Ouachita River and sites 8 and 9 in the

L'Anguille River basins in Arkansas. Site 10 was located in the Pearl

River basin in Louisiana, and sites 11 and 12 in the Apalachicola River

basin in Florida. Sites 13 and 14 were located in Georgia in the

Ocmulgee River and Altamaha River basins, respectively. Sites 15, 16,

and 17 were located in South Carolina in the Edisto, Lynches, and

Waccamaw River basins, respectively.

All sites were characterized by a growing season of greater than

200 days and average annual rainfall ranging from 105 to 170 cm. The

overstory typically ranged from cypress-tupelo or willow in depressions

and low flats to white oak-hickory or pine on the high ridges. Inter-

mediate areas included overcup oak-bitter pecan, green ash, willow oak,








and American elm overstory communities. The herbaceous understory was

typically dense with diverse species of trees and shrubs, vines, and

herbs. Appendix A includes a general description of each study site.


Determining Hydrologic Zone Elevations


Hydrologic data for each site were obtained either from the

U.S. Geological Survey (flow data) or from the local Corps of Engineers

District (stage or flow data). Data were analyzed using a FORTRAN com-

puter program developed for determining hydrologic zone elevations in

study sites where flooding occurred. The program output is the duration

of inundation plus soil saturation of each hydrologic zone boundary,

expressed as flow rate or stage data. Table 1 presents inundation/

saturation frequency and duration for Zones 2 to 6.

Hydrologic zone elevations for each site were computed using the

most recent 10 to 20 years of daily stream gauge data. When gauge data

were provided as daily discharges (flow rate), a rating table (relation-

ship between stage and discharge) was obtained to determine the corre-

sponding stages (elevation).

Plant species show little or no adverse effects from flooding in

the winter (dormant) season (Hall and Smith 1955; Bruckner, Bowersox,

and Ward 1973). Therefore, hydrology during the dormant season was not

used in this study to determine zones.

The dates of the first and last day of the growing season for each

site were provided as input to the computer program. Growing season for

this study was defined as the period between the last average occurrence

of 32* F in the spring and the first average occurrence of 32" F in the

fall. The program eliminated all nongrowing season data and ranked the








Table 1

Hydrologic Zones Occurring in Bottomland Forests
of the Southeastern United States


Zone Name
2 Semipermanently to
permanently inun-
dated or saturated

3 Regularly inun-
dated or saturated

4 Seasonally inun-
dated or saturated

5 Irregularly inun-
dated or saturated






6 Intermittently
inundated or
saturated


Typical
Inundation/Saturation Frequencya
Annual (1 year frequency)
90 to 100 years/100 years


51 to 90 years/100 years
(>l-year to 2-year frequency)

51 to 90 years/100 years
(>l-year to 2-year frequency)

11 to 50 years/100 years (well
drained) (>10 years 2-year
frequency)
1 to 10 years/100 years (poorly
drained) (100 years, 10-year
frequency)

1 to 10 years/100 years
(100 years, 10-year frequency)


Duration"
(percent)
>75-100




>25-75


>12.5-25

>5-12.5







<5


Source: Adapted from Larson et al. (1981).
a Although typical inundation/saturation frequencies are provided
for each zone, almost any frequency could be associated with any
duration of inundation/saturation. Therefore, only duration of
inundation/soil saturation was used to determine hydrologic
zones.
b Duration based on the growing season.



remaining daily readings during the period of record from highest to

lowest flow (or stage). Elevations corresponding to the 75, 25, 12.5,

and 5 percent durations of inundation were computed. Because the

resulting elevations did not include the period during which the soils

remain saturated after a period of inundation, saturation effects were

integrated. A general description of the soil series occurring in each

zone of the study site was obtained from Soil Conservation Service (SCS)








county soil surveys. An estimated range of permeabilities for the top

30 cm of the soil profile (i.e., defined for this study as the effective

root zone) was determined. This range approximated the period required

for the root zone to become saturated after inundation. The slowest

value in the range of permeabilities was used to determine the minimum

duration of inundation required to saturate the soil. A second range of

soil permeabilities between the 30-cm and 90-cm depth was determined.

The slowest permeability value of the soil profile between 30 and 90 cm

was used to estimate the time required for draining of the root zone

after dewatering. A mean daily transpiration factor for floodplain

forests of 5.6 mm (Brown 1981) also was incorporated for computing

desaturation.

Permeability and transpiration coefficients were provided as pro-

gram input, and new flow (or stage) values for hydrologic zone bound-

aries were derived that reflected both inundation and soil saturation.

This iterative process required a computer search. The computer program

added the days of saturation to the days of inundation, and the output

was flow (or stage) values that represented the estimated boundary of

each hydrologic zone, based on inundation and saturation. The gauge

elevation was added to the stage for each zone to obtain the mean sea

level elevation at the gauge. When the site was not immediately adja-

cent to the gauging station, the change in water surface elevation

between the study area and the gauging station was determined using the

best available water surface profile data. Appendix B explains how the

computer program analyzes the hydrology data to produce zone boundaries.








Site Preparation and Data Collection


A temporary benchmark was established at each of the 17 sites by

surveying from a permanent benchmark. A reconnaissance of the area was

conducted for suitable sites, and mean sea level elevations for each

hydrologic zone boundary were surveyed along the topographic gradient.

The contours of each hydrologic zone boundary within the site were

marked with surveyor flags. Fifty-five hydrologic zones were estab-

lished on the 17 study sites. Sampling methods were adapted from meth-

ods described by Whittaker (1973), except where noted.

Sample plots were established parallel to the hydrologic zone

boundary (Figure 3). Plots were positioned on the downslope side of the

boundary with at least a 5-m buffer between the sample plots and the

upper and lower boundary of the hydrologic zone. A belt transect (20 m

wide by 40 m long) containing 10 sample subplots (8 m by 10 m) was

established within each zone.

Small soil pits in each sample plot were dug with a tile spade to

a depth necessary to identify the soil series. In all cases a county

soil survey was used to identify the mapped soil series, and information

was obtained to verify the soil series on site. Assistance from the

local SCS office was used to determine the correct soil series and soil

permeability coefficients for each zone sampling site.

Vegetation was sampled by vegetative layer. All trees in each

sample plot were identified by species and the diameter at breast height

(1.5 m) of individuals having a diameter of greater than or equal to

7.5 cm was measured and recorded to the nearest whole centimeter.








OBM

ZONE 6

ZONE 5
S5TBM
ZONE 4


ZONE 3


0BM BENCHMARK
<(TBM TEMPORARY BENCHMARK


SSURVEYED CONTOUR LINE
414m SAMPLE PLOTS


9T HYDROLOGY GAUGE


Figure 3. Representation of a typical research site


S40mm ZONE 2
eo I0


LEGEND








All saplings and shrubs (woody plants less than 7.5 cm in diam-

eter, but greater than 1.0 m in height, excluding vines) in each sample

plot were identified by species, and the height class of each individual

was recorded. Saplings or shrubs with more than one stem clustered from

a single root system were counted as individuals only when separation

occurred at or below ground level. The following height classes were

used: Class 1 = 1.0 to 2.0 m, Class 2 = 2.1 to 3.0 m, Class 3 = 3.1 to

4.0 m, Class 4 = 4.1 to 5.0 m, and Class 5 = >5.0 m.

All climbing woody vines greater than 1.0 m in height in each

sample plot were identified by species, the stems of each species coun-

ted, and the height class of the highest individual on each tree or

sapling/shrub recorded. The following height classes were used:

Class 1 1.0 to 3.0 m, Class 2 = 3.1 to 6.0 m, Class 3 = 6.1 to 12.0 m,

and Class 4 = >12.0 m. Vines were recorded when any portion of the

plant occurred in, or overhung, the plot. Individual stems were

recorded when separation from the root system occurred at or below

ground level.

Percent cover was estimated for each species of herb and woody

seedling (greater than 1.0 m in height) rooted in the plot in two

randomly located 1.0-m2 quadrats in each subplot using the Daubenmire

(1968) cover class method.


Analyzing Vegetation Data


Importance values for species in all vegetation layers except the

herbaceous layer were calculated by adding values for relative density,

relative frequency, and relative dominance. Importance values for her-

baceous species were calculated by summing relative frequency and








relative dominance. Importance values were used to determine the FTI

number for each species.

When species could not be positively identified in the field,

voucher specimens were collected and later identified. Species nomen-

clature was determined using the National List of Scientific Plant Names

(U.S. Department of Agriculture 1982).


Calculating Species FTI Numbers


Changes in composition of biotic communities along environmental

gradients can be addressed with several statistical techniques, the most

notable being gradient analysis (Whittaker 1978). Gradient analysis can

take several different forms depending on the objective of the analysis.

Inferring environmental values (e.g., hydrologic zones) from vegetative

species composition is called a "calibration problem" by Ter Braak and

Prentice (1988) and is the appropriate approach for this study.

One method of calibration is to use weighted averaging (WA) to

estimate environmental factors at sites based on species optima. If a

species exhibits a unimodal distribution with respect to an environmen-

tal variable, its occurrence is concentrated around the peak of this

function (Ter Braak and Prentice 1988). Species with similar optima

will naturally tend to occur together. Therefore, an intuitive estimate

of the environmental factor of a site is the average of the optima for

the species present. The FTI numbers represent weighted averages of

species occurrence.

Two additional statistical methods of calibrating an environment

hydrologicc zone) with vegetation, recommended by Ter Braak and Prentice








(1988), are cluster analysis and discriminant function analysis. These

two methods were applied to test the reliability of the FTI numbers.

FTI numbers were calculated for each species occurring in each

vegetation layer. A species could have three different FTI numbers at a

given site, depending on its growth form. For example, Quercus nigra

would have three different FTI numbers when present on a site as a tree,

sapling, and seedling. Species FTI numbers for each site were computed

by the following formula:




FTI = j(J i IVij)
i jIVij


where

i = the ith species

ja = 2.5, 3.5...6.5 hydrologicc zone)

IVij = importance value for species i in the hydrologic zone j

After species FTI numbers were computed for all species in all sites,

the average FTI number (FTI) for each species across all sites was cal-

culated using the following formula:




E =.i FTIj
FTI, = J-
ni


a Because vegetation was sampled between zone boundaries, midrange zone
numbers (e.g., 2.5 for Zone 2, 3.5 for Zone 3, etc.) for zones were
used in calculating FTI numbers.





19


where

i = the ith species

j = sites 1 to 17

FTIij = FTI number of species i at site j

ni = number of sites at which species i occurred















RESULTS AND DISCUSSION


Flood Analysis of Study Sites


Twenty years of hydrologic data were used for all sites except

sites 4, 5, and 14. Sites 4 and 5 had a 19-year hydrologic record, and

site 14 had a 12-year record. Calculations of change in water surface

elevation between the gauging station and site was necessary for sites

1, 2, 3, 5, 9, 12, and 17. All other sites were adjacent to the gauging

station and did not require adjustments. Hydrologic analyses of sites

11 and 12 at the Apalachicola River were verified using information from

another study (Leitman, Sohm, and Franklin 1984).

The hydrologic records for all sites were analyzed by season for

five-year increments. In all cases, variation in flow through time was

determined to be within normal seasonal and annual fluctuations. There-

fore, it was assumed that the hydrologic record reflected normal condi-

tions (i.e., no major drainage projects during the period of record had

significantly impacted the plant community structure).

The hydrologic data also were analyzed to determine annual flood

frequency and duration for each site. The boundaries between Zones 2

and 3 and between Zones 3 and 4 are flooded virtually every year

(Table 2). The boundary between Zones 4 and 5 is flooded at least every

other year, and the boundary between Zones 5 and 6 is flooded from once








Table 2

Annual Flood Frequency (Percent of Years in Which Boundary
Is Exceeded at Least Once during Growing Season for More
than 7 days) for Zone Boundaries


Zone Boundaries
Site 2-3 3-4 4-5 5-6

1 100 90 70 10

2 100 90 70 10

3 100 100 85 25

4 100 94 61 20

5 100 85 55 20

6 100 96 77 20

7 100 100 92 75

8 100 100 96 60

9 100 100 96 65

10 100 100 83 70

11 100 100 95 70

12 100 100 95 70

13 100 100 90 65

14 100 100 92 67

15 100 90 85 55

16 100 100 100 70

17 100 100 90 70








in 20 years to 3 out of every 4 years. Similar findings were reported

by Clark and Benforado (1981) and Roelle et al. (1987).

In general, the average duration of a flood event by site ranged

from 3 months to longer than 5 months for the Zone 2-3 boundary, from

3 weeks to greater than 2 months for the Zone 3-4 boundary, from 1 week

to 1 month for the Zone 4-5 boundary, and from less than 1 day to

10 days for the Zone 5-6 boundary (Table 3).

As an example, a hydrograph of site 3 at Steele Bayou (Appendix C)

for 20 years of data shows that flooding conditions vary greatly from

one season to another and from one year to another. Also, flooding

during the growing season usually occurs during early spring and is

usually continuous with the nongrowing season flooding. Flooding rarely

occurs late in the growing season. Unusual events such as the 1973

flood can greatly affect the calculated value of average events. For

example, if the data for 1973 were excluded, the average duration per

flood event at site 3 is reduced to 37 days, 25 days, 9 days and less

than 1 day, from 143, 30, 14, and 6 days, for Zone boundaries 2 through

5, respectively (Table 4). Such an event has an especially large effect

on average duration even over a 20-year period, especially in the higher

zones.


Vegetation Data


Vegetation data were collected at each of the 55 hydrologically

defined zones for the 17 study sites (Appendix D). Eleven stands were

sampled in Zone 2, 15 stands in Zone 3, 14 stands in Zone 4, 4 stands in

Zone 5, and 11 stands in Zone 6. The total possible number of stands








Table 3

Average Annual Duration of Flood Events (days) for Zone Boundaries


Site

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Study Area
Average


2-3

119

119

143

92

93

152

153

139

139

141

143

143

162

159

162

174


143 26 37 18


Zone Boundaries
3-4 4-5

21 12

21 12

30 14

41 19

51 24

44 22

50 25

20 15

20 15

32 14

20 16

20 18

29 15

53 24

52 25

58 8


5-6

<1

<1

6

7

8

9

10

<1

<1

3

8

6

8

5

9

<1


18 + 6 6 3








that could have been sampled for each zone was 17. Some zones were

unsuitable for analysis due to disturbance or because they were too

narrow to support sampling areas (Appendix A). Zone 5 was especially

susceptible to disturbance. In some cases Zone 5 was cleared because it

was dry enough to be farmed during the growing season. In other cases,

Zone 5 was too narrow because it was located near the toe of a slope.

Therefore, Zone 5 was sampled only at sites 3, 11, 12, and 14.

FTI numbers were calculated for 74 tree species, 118 species of

saplings and shrubs, 31 species of woody vines, and 268 species of herbs

and woody seedlings, representing 312 different plant species. Because

some species occurred in more than one vegetative layer, the total num-

ber of species FTI numbers exceeded the total plant species identified.

Appendix E contains a listing by stratum of the plant species

identified in this study and the calculated FTI numbers with standard

deviations provided for each species.

FTI numbers were calculated for all plants identified by species

in this study, regardless of how frequently they occurred in the study.

The FTI numbers calculated for species with few occurrences may be sus-

pect. FTI numbers for some of the more commonly occurring species in

the study are listed in Table 4. These species can generally be found

along the moisture gradient in any bottomland hardwood forest in the

southeastern United States in roughly the order from wettest to driest

community as presented. Some species, such as Nuttall's oak, are common

in only part of the region, and not all species can be expected on the

same site due to the species' response to other environmental gradients.








Table 4

FTI Numbers of Species Commonly Occurring in Bottomland Forests


Scientific
Name


Nyssa aquatica

Salix nigra

Fraxinus caroliniana

Taxodium distichum

Planera aquatica

Acer drummondii

Forestiera acuminata

Gleditsia aquatica

Carya aquatic

Quercus lyrata

Quercus laurifolia

Betula nigra

Acer rubrum

Ilex decidua

Fraxinus pennsylvanica

Ulmus americana

Quercus nuttallii

Quercus phellos

Acer negundo

Celtis laevigata

Carpinus caroliniana


Common
Name

Trees

Water tupelo

Black willow

Pop-ash

Bald cypress

Water elm

Drummond red maple

Swamp privet

Water locust

Water hickory

Overcup oak

Laurel oak

River birch

Red maple

Possumhaw holly

Green ash

American elm

Nuttall's oak

Willow oak

Box elder

Sugarberry

American hornbeam


FTI
Number


2.62

2.83

2.87

2.97

3.12

3.48

3.48

3.50

3.54

3.73

3.89

4.01

4.21

4.35

4.44

4.46

4.50

4.81

4.83

4.84

4.84








Table 4--continued


Scientific
Name

Liquidambar styraciflua

Platanus occidentalis

Nyssa sylvatica

Carya illinoensis

Quercus nigra

Morus rubra

Ilex opaca

Ulmus alata

Pinus taeda

Quercus alba

Sassafras albidum

Fagus grandifolia

Cornus florida

Ostrya virginiana

Quercus stellata

Quercus falcata

Carya tomentosa



Salix nigra

Itea virginica

Planera aquatica

Cephalanthus occidentalis

Styrax americana


Common
Name

Sweetgum

Sycamore

Black gum

Pecan

Water oak

Red mulberry

American holly

Winged elm

Loblolly pine

White oak

Sassafras

American Beech

Flowering dogwood

American hophornbeam

Post oak

Southern red oak

Mockernut hickory

Saplings and Shrubs

Black willow

Virginia willow

Water elm

Buttonbush

Snowbell


FTI
Number

5.03

5.18

5.27

5.57

5.73

5.75

5.79

6.43

6.41

6.50

6.50

6.50

6.50

6.50

6.50

6.50

6.50


2.83

2.83

3.01

3.13

3.41








Table 4--continued


Scientific
Name

Forestiera acuminata

Cyrilla racemiflora

Celtis laevigata

Crataegus viridis

Carya illinoensis

Platanus occidentalis

Acer negundo

Magnolia grandiflora

Liquidambar styraciflua

Cornus drummondii

Vaccinium elliottii

Quercus nigra

Sambucus canadensis

Halesia diptera

Morus rubra

Cercis canadensis

Vaccinium arboreum

Gleditsia triacanthos

Quercus alba

Cornus florida

Ilex vomitoria


Common
Name

Swamp privet

Titi

Sugar berry

Green hawthorn

Pecan

Sycamore

Box elder

Southern magnolia

Sweetgum

Rough leaf dogwood

Elliott blueberry

Water oak

Elderberry

Silverbell

Red mulberry

Redbud

Farkleberry

Honey locust

White oak

Flowering dogwood

Yaupon

Woody Vines

Morning glory


FTI
Number

3.57

3.72

4.37

4.46

5.00

5.05

5.20

5.43

5.52

5.69

5.82

5.92

5.95

6.09

6.25

6.37

6.45

6.50

6.50

6.50

6.50


Ipomoea wrightii


2.50








Table 4--continued


Scientific
Name

Smilax walteri

Brunnichia cirrhosa

Amplelopsis arborea

Campsis radicans

Vitis palmata

Trachelospermum difforme

Vitis riparia

Smilax laurifolia

Cocculus carolinus

Berchemia scandens

Similax bona-nox

Toxicodendron radicans

Smilax rotundifolia

Vitis rotundifolia

Parthenocissus quinquefolia

Lonicera japonica

Gelsimium sempervirens



Ludwigia decurrens

Rorippa islandica

Echinodorus cordifolius

Aster simplex

Boehmeria cylindrica


Common
Name

Walter's greenbriar

Ladies' eardrops

Peppervine

Trumpet creeper

Cat grape

Star jasmine

Riverbank grape

Bamboo-vine

Carolina moonseed

Rattan vine

Saw greenbriar

Poison ivy

Common greenbriar

Muscadine grape

Virginia creeper

Japanese honeysuckle

Carolina jessamine

Herbs

Primrose-willow

Yellowcress

Creeping burhead

Lowland white aster

Small-spike falsenettle


FTI
Number

3.05

3.58

3.94

4.05

4.07

4.18

4.27

4.33

4.37

4.55

4.75

4.82

5.18

5.71

5.93

6.50

6.50



2.50

2.50

3.00

3.04

3.34








Table 4--continued


Scientific
Name


Spermacoce glabra

Saururus cernuus

Leersia lenticularis

Justicia ovata

Urtica chamaedryoides

Clematis virginiana

Eupatorium rugosum

Cocculus carolinus

Viola missouriensis

Hypericum hypericoides

Mitchella repens

Arundinaria gigantea

Vitis rotundifolia

Gelsemium sempervirens

Sanicula canadensis

Galium aparine

Carex albolutescens

Cynanchum laeve

Lonicera sempervirens


Common
Name

Smooth buttonweed

Lizard-tail

Catchfly grass

Waterwillow

Nettle

Virgins bower clematis

White snakeroot

Snailseed

Missouri violet

St. Johnswort

Partridge berry

Giant cane

Muscadine

Carolina jessamine

Black snakeroot

Catchweed bedstraw

Sedge

Cynanchum

Trumpet honeysuckle


FTI
Number

3.50

3.65

3.67

3.83

4.42

4.57

4.67

4.78

4.83

5.25

5.32

5.34

5.89

6.31

6.39

6.50

6.50

6.50

6.50








Examples of FTI numbers by life stage for woody species are pre-

sented in Table 5. Although differences between FTI numbers for tree

and sapling life stages are not great, FTI numbers for saplings tend to

be slightly higher than for trees because saplings are generally more

sensitive to flooding than trees. The sapling life stage also tends to

occur in a broader range of zones than the tree life stage due to tem-

poral variations in selective pressures (e.g., competition and response

to flooding). FTI numbers of seedlings have little value in determining

hydrologic zones because they only reflect seed dispersal potential and

germination wherever seedbed conditions are favorable. For example,

seedlings of least tolerant species (e.g., Sassafras albidum)



Table 5

Variations in Species Flood Tolerance Index Numbers
According to Life Stage


Scientific Name Tree Sapling Seedling

Taxodium distichum 2.97 3.33 3.09

Gleditsia aquatic 3.50 3.15 3.27

Carya aquatica 3.54 3.70 3.69

Quercus lyrata 3.73 3.99 3.80

Fraxinus pennsylvanica 4.44 4.27 4.00

Quercus nuttallii 4.50 4.50 4.50

Acer negundo 4.83 5.20 5.58

Celtis laevigata 4.84 4.37 4.77

Liquidambar styraciflua 5.03 5.52 4.87

Quercus nigra 5.73 5.92 5.85

Sassafras albidum 6.50 6.50 6.07








occasionally occur in lower zones (Table D-3, Zone 3), but the

individuals do not survive to maturity unless the hydrologic regime is

drastically altered.

The average importance value was plotted for each species in every

zone where it occurred in the study. Although a species can be expected

to occur in a number of zones, many had a peak occurrence in a particu-

lar zone (Figure 4). However, because this study did not analyze a

continuous gradient, FTI numbers were calculated from mean importance

values across the entire study and do not necessarily represent the

maximum in ecological amplitude for a species.

Three general species distribution patterns are shown in Figure 4.

The first pattern includes species such as water tupelo (NYAQ) and bald

cypress (TADI), in which mean importance value is greatest in Zone 2 and

the species no longer occurs after either Zone 3 or 4. This pattern is

indicative of species having the strongest competitive advantage in

areas of greatest duration of inundation/soil saturation. The second

pattern is typified by species such as water oak (QUNI), loblolly pine

(PITA), sassafras (SAAL), and white oak (QUAL), in which the greatest

mean importance value occurs in Zone 6 (uplands) and decreases from Zone

5 to 2. Hence they have a stronger competitive advantage in areas where

inundation/soil saturation is less than 5 percent of the growing season.

However, some of these species (e.g., water oak and loblolly pine) may

occasionally occur as dominants in wetlands. The third pattern is typi-

fied by species having the greatest mean importance values in Zones 3,

4, and 5. Species in this group sometimes occur as dominant species in

either Zone 2 or 6, but are best adapted for occurrence at some point in

Zones 3, 4, and 5. Species having the greatest mean importance values



























60 -

U 50

40 -
o

30


20 -

10 FRP


ZONE 2 3 4 5 6
WET DRY
HYDROLOGIC ZONES
















Figure 4. Ecological amplitude of some commonly occurring species;
CAAQ: Carya aquatica; FOAC: Forestiera acuminata; FRPE: Fraxinus
pennsylvanica; LIST: Liquidambar styraciflua; NYAQ: Nyssa aquatica;
PITA: Pinus taeda; QUAL: Quercus alba; QULY: Quercus lyrata;
QUNI: Quercus nigra; SAAL: Sassafras albidum; TADI: Taxodium
distichum; and ULAM: Ulmus americana








in Zone 3 are overcup oak (QULY) and bitter pecan (CAAQ), while American

elm (ULAM) and sweetgum (LIST) develop the greatest mean importance

values in Zones 4 and 5, respectively.

Other systems have been developed to identify the degree of wet-

ness for which a species is best adapted (e.g., Hook 1984, Reed 1988).

These systems used qualitative descriptions, such as "most tolerant" or

"obligate hydrophyte," and are based primarily on the literature or

"expert evaluations" and not on a single coordinated study. Species

evaluations were also made by life forms in this study, and no such dis-

tinction was made in the other two systems. However, all three systems,

including the FTI numbers, were developed to identify a degree of wet-

ness for which a species is best adapted. They all have five categories

that vary from wettest to driest. Therefore, an obvious comparison

would be to compare species having an FTI integer of 2, with the most

tolerant and obligate designation, an FTI integer of 3, with the highly

tolerant and facultative wet description, etc.

Species identified in this study are listed along with correspond-

ing FTI numbers and ratings of those species from the systems developed

by Hook and Reed in Appendix E. Among selected tree species shown in

Table 6, all species having an FTI number from 2 to 4 are obligate

plants (OBL) in the National List of Plant Species that Occur in Wet-

lands: Southeast (Region 2) (Reed 1988) and have a water-logging-

tolerance rating of most or highly tolerant (Hook 1984). All listed

species except Pinus taeda that have an FTI number of 6 to 6.5 are fac-

ultative upland (FACU) species (Reed 1988) and are rated by Hook (1984)

as the least-tolerant species. P. taeda (FTI-6.41) has an indicator

status of facultative (FAC) and is rated by Hook as moderately tolerant








Table 6

Comparison of Three Water-Tolerance Ratings for


Selected Bottomlan t


Tree Species


FTII Standard
Deviation


Nyssa aquatic
Salix nigra
Fraxinus caroliniana
Taxodium distichum
Forestiera acuminata
Gleditsia aquatica
Carya aquatica
Quercus lyrata
Betula nigra

Diospyros virginiana

Acer rubrum

Fraxinus pennsylvanica

Ulmus americana

Quercus phellos

Acer negundo

Carpinus caroliniana
Celtis laevigata
Liquidambar styraciflua

Carya illinoensis
Pinus taeda


2.62
2.83
2.87
2.97
3.48
3.50
3.54
3.73
4.01


0.20
0.58
0.41
0.61
0.50
0.00
0.34
0.68
1.73


4.13 0.82

4.21 + 0.68

4.44 0.67

4.46 0.62

4.81 1.07

4.83 0.47


4.84
4.84
5.03


0.61
0.56
0.65


5.57 1.01
6.41 0.14


6.50
6.50
6.50
6.50


0.00
0.00
0.00
0.00


NWIb
Status
Region 2


OBL
OBL
OBL
OBL
OBL
OBL
OBL
OBL
OBL


FAC

FAC

FACW

FACW

FACW

FACW

FAC
FACW
FAC+

FAC+
FAC

FACU
FACU
FACU
FACU


Mean for all study sites.
Taken from Reed (1988) (see Appendix A).
Taken from Hook (1984) (see Appendix B).


Cornus florida
Fagus grandifolia
Quercus alba
Sassafras albidum


Selected ......... .. Forest Tree ne---i


Waterlogging
Tolerance
Rating GroupC

Most tolerant
Most tolerant
Most tolerant
Most tolerant
Most tolerant
Highly tolerant
Highly tolerant
Highly tolerant
Moderately
tolerant
Moderately
tolerant
Moderately
tolerant
Moderately
tolerant
Moderately
tolerant
Moderately
tolerant
Moderately
tolerant
Weakly tolerant
Weakly tolerant
Moderately
tolerant
Weakly tolerant
Moderately
tolerant
Least tolerant
Least tolerant
Least tolerant
Least tolerant








of waterlogged soils. All species except Betula nigra that have an FTI

number of 4 to 6 are facultative wet (FACW) or FAC and are rated as

moderately or weakly tolerant. B. nigra, an obligate species, occurs on

well-drained soils, often on natural berms. FTI numbers were computed

only for bottomland forests and do not reflect occurrence in other wet-

land types (e.g., pocosins and Carolina bays); thus, slight deviations

from the above pattern should be expected for some species. Also some

species may have genetic variants that possess varying degrees of flood

tolerance. Hook et al. (1988) reported that interspecific variation in

tolerant to waterlogging exists in loblolly pine (Pinus taeda). This

may also be true of other species.

Although the system using the National Wetlands Inventory (NWI)

indicator status (Reed 1988) does not allow comparison of hydrologic

definitions, the actual average duration of flooding in this study com-

pared with Hook's (1984) waterlogging tolerance rating definitions

yields strong agreement. For example, assuming an average 225-day grow-

ing season, the most tolerant rating can be defined as approximately

200 days (Appendix E). The boundary between Zones 2 and 3 (which theo-

retically would be slightly drier because the designation most tolerant

is best compared to Zone 2, not the boundary between Zones 2 and 3) has

a duration of inundation/soil saturation that ranges from 92 to

198 days. Highly tolerant ranges from 30 to 90 days versus 20 to

66 days for the Zone 3-4 boundary. Weakly tolerant ranges from 1 to

4 weeks, while the Zone 4-5 boundary ranges from 8 to 31 days. "Least

tolerant" is defined as waterlogging for a few days, but usually less

than 2 percent of the growing season. Using the 225-day growing season,








this could be assumed to be 3 or 4 days. The boundary between Zones 5

and 6 varied from less than 1 to 10 days.


Weighted Averaging


Weighted average estimates of species optima in this study were

calculated as FTI values as previously described, with importance values

used as the indicator of species abundance at each site.

The use of the weighted averaging approach requires that a number

of conditions be met, including: (1) species exhibit unimodal abundance

distributions, (2) species optima are equally spaced along the environ-

mental variable, (3) species have equal tolerances of the environmental

variable, and (4) species have equal maximum values for the environmen-

tal variable. Strict adherence to some of these conditions is not

always possible. Additional considerations should also be noted.

Species-rich samples should not occur at one end of the gradient.

Environmental tolerances of species should not vary substantially. The

standard deviation of FTI numbers is an estimate of tolerance in this

analysis. Species with narrow tolerances have low FTI numbers and

standard deviations and those with wide tolerance have high standard

deviations.

Some of the aforementioned conditions are not strictly met in this

study. Although species richness was fairly even across the hydrologic

gradient, few species had a peak abundance in Zone 5, perhaps because

Zone 5 was undersampled (n = 4) relative to the other zones. The condi-

tion of a unimodal abundance distribution is upheld for tree species

with peaks in hydrologic Zones 2, 3, and 6, but several species (e.g.,

Ilex opaca, Quercus nigra, Nyssa sylvatica, and Liquidambar styraciflua)








exhibit bimodal distributions with peaks in Zones 4 and 6. Again, this

pattern may have been influenced by fewer samples being taken in Zone 5.

The condition of equal tolerances is also violated somewhat. For

instance, those species with abundance peaks in Zones 2 and 6 have nar-

row tolerances; whereas, those that commonly occur in Zone 4 are also

fairly common in other hydrologic zones as well. However, these devia-

tions from the conditions for weighted averaging analysis do not neces-

sarily render the FTI method invalid. Additional analyses were applied

to help determine its validity.


Statistical Analysis of the Vegetation Data


To evaluate a method of identifying hydrologic zones based on

vegetative associations, the analysis must be based on common species

occurring within the region of interest. Species that occur infre-

quently may be excellent indicators of hydrologic conditions when pres-

ent, but their limited abundance makes evaluating their usefulness in

determining a hydrologic zone difficult. For this reason, species

occurring relatively infrequently in this study were not used in testing

the validity of the weighted averages (FTI numbers). More than half of

the 74 tree species (n = 44) recorded had 20 or fewer individuals

throughout the study region and, therefore, were not used in the analy-

ses. The remaining 30 tree species accounted for 90.4 percent of the

individual trees. One hundred and eighteen species of saplings and

shrubs were recorded. Only those 29 species which represented at least

one percent or more of the total and together accounted for 78 percent

of the total saplings and shrubs data set were used in the statistical

analysis. A total of 31 vine species were recorded; the 20 more common








species accounted for 96.5 percent of all individuals and were the spe-

cies analyzed statistically. The herbaceous ground-cover data set con-

tained 268 species. Most occurred only rarely in this study, and only

30 species whose abundances equaled or exceeded 1 percent of the total

herb individuals were included in the analysis. These species accounted

for 48.9 percent of all individuals in this vegetative category.


Cluster Analyses


A method of calibration suggested by Ter Braak and Prentice (1988)

is cluster analysis in which an environmental value hydrologicc zone) is

predicted through use of species abundance indicators (relative fre-

quency). In addition to relative frequency, cluster membership was used

as a predictor of hydrologic zone. Species were clustered according to

similar abundance distributions across the hydrologic gradient.

Cluster analysis was used to group the 30 tree species into five

clusters based on the five hydrologic zones. Results (Figure 5) show

that, with the exception of chinaberry (MEAZ) and deciduous holly

(ILDE), five distinct groups can be discerned. Table 7 gives the rela-

tive frequencies of occurrence of each species in each hydrologic zone

(species are grouped by cluster). Inspection of the data reveals why

chinaberry and deciduous holly did not group readily. Chinaberry is the

only species that occurs almost exclusively in Zone 5. Its occurrence

is also restricted to a single site. The distribution of deciduous

holly peaks in Zones 2 and 3, but not to the extent of other common

species in these zones. Because chinaberry and deciduous holly did not

group readily, they were eliminated from further analysis.






DISTANCES


MEAZ 0.000
OSVI
QUST
CATO
PITA
ILOP
QUNI
NYSY
LIST
CACA
QUPH
CELA
ULAM
FRPE
ACNE
ACRU
QULA
CAAQ
QULY
DIVI
NYBI
ACDR
FOAC
NYAQ
TADI
FRCA
NYOG
SANI
PLAQ
ILDE


1.000
0.843
0.000
0.000
0.003
0.112
0.018
0.108
0.019
0.124
0.027
0.025
0.050
0.131
0.067
0.047
0.152
0.007
0.003
0.003
0.025
0.008
0.001
0.427
0.000
0.004
0.006
0.000
0.021
0.121


Figure 5. Cluster diagram for trees. Distance metric is 1-Pearson
correlation coefficient, single linkage method (nearest neighbor)








Table 7

Relative Frequencies in Each Hydrologic Zone of Tree Species
Used in the Statistical Analyses: Groupings of Species


C ro EnnnA


tn Cl not-or Mo~mhareh4 n


Hydrologic Zone


Species

NYAQa
TADI
FRCA
NYOG
SANI
PLAQ

QULA
CAAQ
QULY
DIVI
NYBI
ACDR
FOAC

CACA
QUPH
CELA
ULAM
FRPE
ACNE
ACRU


2

85.0
82.7
79.1
93.9
96.5
67.7

0.9
0.8
6.5
0
15.3
7.1
3.6

0.6
0
2.0
1.9
0
0
6.8

0
0
0
0.3


Cluster


3

15.0
15.6
20.9
6.1
3.5
31.6

70.1
80.0
72.9
70.6
84.7
88.1
94.0

2.8
8.6
9.8
21.2
45.2
31.1
38.4

0
1.6
0
6.0

0
0
0
0


4

0
1.8
0
0
0
0.7

26.2
19.2
20.0
20.6
0
4.8
2.4

63.7
74.1
62.8
50.0
42.9
53.3
52.0

41.7
32.8
55.9
45.4

0
0
0
7.3


ILOP
QUNI
NYSY
LIST

OSVI
QUST
CATO
PITA


5

0
0
0
0
0
0

0
0
0.6
2.9
0
0
0

17.0
10.3
21.6
25.0
11.9
4.4
1.4

0
7.4
5.9
14.9

0
0
0
0


6

0
0
0
0
0
0

2.8
0
0
5.9
0
0
0

15.9
6.9
3.9
1.9
0
11.1
1.4

58.3
58.2
38.2
33.3

100
100
100
92.7


a Represents the first two letters of the plant genus and the first two
letters of the species name, i.e., NYAQ stands for Nyssa aquatica.
Species codes for all species are identified in Appendix E.


or espon ri t m i -se nms, 4








A closer inspection of Table 7 indicates that the five clusters

have modal peaks that correspond, with varying amplitude, to the five

hydrologic zones. Trees in cluster 1 are found almost exclusively in

Zones 2 and 3. Likewise, trees in cluster 2 occur most commonly in

Zone 3. Cluster 5 had the most restrictive distribution, with three of

the four species occurring exclusively in Zone 6. The remaining clus-

ters, 3 and 4, had less distinctive modes, but exhibited greater distri-

butions in Zones 4 and 6.

Cluster analysis on the entire sapling and shrub data set did not

produce distinct groupings (Figure 6). Silver bell (HADI) was not eval-

uated further because it occurred at a single site. The data set was

split into the two sapling and shrub (bush) components, and cluster

analysis was recalculated on each component. The saplings alone

(n = 17) grouped more distinctly (Figure 7) than the shrubs (n = 11)

(Figure 8). Shrubs, therefore, were not used in any further analyses.

Saplings alone grouped into five distinct groups but not as strongly as

trees. The relative frequency of occurrence and cluster membership of

the sapling species are given in Table 8.

The vine data produced five clusters, morning glory (IPWR) being a

single species cluster corresponding to Zone 2 (Figure 9). Two vine

species poison ivy (TORA) and trumpet creeper (CARA) were not included

in larger clusters, largely because of their apparent wide tolerances of

hydrologic conditions. Table 9 depicts the relative abundance distribu-

tions of each vine species across the hydrologic gradient within a clus-

ter and illustrates a pattern similar to but not as strong as that of

the tree species.







DISTANCES
0.000 0.500
ULAM
A 0.126
CELA
0.142
ILDE
0.152
ULAL
0.013
HADI
0.002
SYTI
0.000
VAAR
0.000
CATO
0.000
MYCE
0.016
QUNI
0.029
NYSY
0.027
LIST
0.115
ILOP
0.054
CACA
0.164
ACNE
VAL 0.126
VAEL -
0.016
OSVI
0.017
CODR
0.301
CRVI
0.309
QULA
E 0.145
FRPE
0.126
FOAC
S0.166
SANI
A 0.064
FRCA
0.074
PLAQ
0.083
CRAE
0.162
CEOC
0.183
ACRU
0.318
STAM


Figure 6. Cluster diagram for saplings and shrubs. Distance metric
is 1-Pearson correlation coefficient, single linkage method (nearest
neighbor)












DISTANCES


0.000
SANI
FRCA
OSVI
CODR
ULAM
CELA
ULAL
QUNI
CATO
NYSY
LIST
ILOP
CACA
ACNE
ACRU
QULA
FRPE


2.000
0.064
1.112
0.017
0.191
0.125
0.152
0.023
0.021
0.029
0.027
0.115
0.054
0.164
0.396
0.182
0.147


Figure 7. Cluster diagram for saplings alone. Distance metric is
1-Pearson correlation coefficient, single linkage method (nearest
neighbor)

















DISTANCES


0.000
ILDE
CRVI
VAEL
SYTI
VAAR
MYCE
STAM
PLAQ
FOAC
CREA
CEOC


Figure 8. Cluster diagram for shrubs alone. Distance metric is
1 Pearson correlation coefficient, single linkage method
(nearest neighbor)


1.000
0.300
0.613
0.736
0.000
0.000
0.747
0.368
0.169
0.167
0.161








Table 8

Relative Frequency of Occurrence of Each Sapling Species in the


Hvdrologic


2

99.4
61.0
1.4
0.0
4.7
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2


Zones along with Their s


3

0.6
21.2
35.8
50.0
53.3
2.0
13.1
0.0
0.0
0.0
0.4
0.0
0.0
5.4
0.0
1.4


4

0.0
2.7
35.1
21.9
36.7
42.1
47.5
0.0
1.5
0.8
14.2
0.0
24.7
24.2
55.5
40.6


5

0.0
0.0
4.6
21.9
3.6
32.2
19.5
71.2
80.0
17.7
11.1
0.0
22.7
14.6
5.0
15.7


6

0.0
15.1
23.0
6.3
1.8
23.7
19.9
28.8
18.5
81.5
74.2
100.0
52.6
55.8
39.5
42.0


Cluster

1



2


3


4





5


The herb data also did not cluster distinctly (Figure 10).

Inspection of the abundance distributions of the herbaceous species

reveals an absence of strong unimodal peaks that correspond to hydro-

logic zones; therefore, this vegetative type is less useful in classifi-

cation. The lack of significant clustering of the shrub and herbaceous

data sets was consistent with the results of ordination attempts of the

ground layer at Neches River sites by Mohler (1979). Mohler concluded

that the herbs and low shrubs were relatively unimportant components of

the forest when compared to the trees. Clearly, the low shrubs and

herbaceous species are more prone to bias due to successional processes

and vegetative change caused by local disturbances which may be

unrelated to the environmental gradient.


Species

SANI
FRCA
ACRU
FRPE
QULA
ULAM
CELA
OSVI
CODR
ULAL
QUNI
CATO
NYSY
LIST
ILOP
CACA


-- d ol vi Zones al n .. .. .. Th i ..... ... .. .. .... ........ ..













0.000
GESE
CLLI
ARTO
BICA
VIRO
TORA
CARA
VIPA
TRDI
BROV
VIRI
SMLA
COCA
SMWA
AMAR
PAQU
SMRO
BESC
SMBO
IPWR


DISTANCES


2.000
0.000
0.000
0.140
0.144
0.418
0.176
0.133
0.009
0.029

0.097
0.021
0.017
0.026
0.009
0.184
0.108
0.058
0.074
1.140


Figure 9. Cluster diagram for vines. Distance metric is 1-Pearson
correlation coefficient, single linkage method (nearest neighbor)








Table 9

Relative Frequencies in Each Hydrologic Zone of the Vine
Species Used in Statistical Analyses


Hydrologic Zone
Species 2 3 4 5 6 Cluster

IPWR 100 0 0 0 0 1

VIPA 0 64.1 30.5 5.5 0
TRDI 2.1 70.2 26.9 0 0.8 2
BROV 13.4 65.2 17.8 3.6 0

VIRI 1.0 33.3 43.8 7.6 14.3
SMLA 1.8 47.4 45.6 0 5.3
COCA 0 41.9 52.9 0 5.2 3
SMWA 9.1 36.4 53.8 0 0.7
AMAR 13.2 37.2 46.3 0 3.3

PAQU 1.5 55.9 2.2 40.4
SMRO 0 2.4 71.4 7.1 19.1 4
BESC 4.6 13.7 61.1 14.5 6.1
SMBO 6.7 13.3 43.3 26.7 10.0

GESE 0 0 0 0 100
CLLI 0 0 0 0 100
ARTO 0 0 0 0 100 5
BICA 0 19.8 3.1 25.0 52.1
VIRO 1.6 6.0 20.0 24.4 48.0



a Groupings of species correspond to cluster membership.



Results of cluster analyses suggest that cluster membership for

tree data provide better indicators of hydrologic zones than saplings or

vines. Shrubs and herbaceous species would be the least useful of all

strata types.


Discriminant Function Analysis


Discriminant function analysis (DFA) is used to determine func-

tions which allow one to classify an individual into one of the








0.000
SMRO
LIST
QUPH
MIRE
VIPA
LEVI
BROV
AMAR
QULA
JUOV
TRDI
COCA
CARA
ILDE
BESC
ULAM
ACRU
QULY
CAAQ
TORA
TADI
PLAQ
BOCY
LYJA
CELA
EURA
VIRO
QUNI
BICA
CARE


DISTANCES


0.200
0.004
0.031
0.035
0.046
0.033
0.017
0.016
0.013
0.002
0.019
0.042
0.010
0.019
0.011
0.032
0.064
0.077
0.009
0.106
0.106
0.086
0.043
0.152
0.118
0.170
0.018
0.003
0.014
0.095


Figure 10. Cluster diagram for herbs. Distance metric is 1-Pearson
correlation coefficient, single linkage method (nearest neighbor)








hydrologic zones using a number of independent variables. The indepen-

dent variables in this analysis are the average importance values for

each tree cluster and vine cluster in each zone. In cases where only

one species in a cluster was present in a zone, that importance value

alone was used. Table 10 lists the data set upon which this analysis

was based.

Three analyses were performed using DFA. The first analysis

developed a classification model using only the importance value from

tree clusters. This model correctly classified 47 of 55 sites for an

overall misclassification rate of 14.6 percent (Table 11). In all

cases, errors involved an assignment to a neighboring zone. For

instance, a single Zone 3 site was assigned to Zone 2, and three Zone 6

sites were assigned to Zone 5. The greatest error in classification

occurred for Zone 4 sites, which is expected because of the greater

tolerance of species for this zone. As previously shown (Figure 4),

Zone 4 is that portion of the hydrologic gradient that has species com-

mon to all three distribution patterns identified. Three Zone 6 sites

were assigned to Zone 5 primarily due to the lack of Zone 6 species

being identified as commonly occurring. A better agreement would be

expected if all species were used.

The second model examined the ability of vine cluster importance

values to predict zone. Although vines did not cluster as well as

trees, a DFA was examined because their cluster patterns were similar.

However, this model correctly classified only 26 of 55 zones for a

52 percent misclassification. This model was not considered adequate,

and the results of its classification are not presented.








Table 10

Mean Importance Values for Species in Each Cluster Used


in ihe DFA. Arran~ed by


7onn /Simn1 tl


Zone TCla TC2 TC3 TC4 TC5 VC2 VC3 VC4 VC5


78.77
70.26
65.12
91.7
62.38
83.49
73.45
68.41
58.62
58.12
74.99

29.57
30.3
0
0
24.37
22.84
99.86
30
0
32.72
7.39
39.94
10.06
44.79
11.06

0
0
0
5.94
16.64
0
0
0
10.3
0
0
0


0
0
0
8.26
0
21.16
6.19
12.58
32.76
4.19
33.22

39.47
47.5
50.08
71.81
66.92
84.85
25.52
47.29
85.22
54.74
48.01
56.21
99.1
43.95
48.2

15.04
6
72.7
40.19
31.88
8.35
0
17.68
12.76
55.18
7.69
68.09


0
0
0
8.32
0
9.48
0
0
0
0
3.34

0
0
26.13
0
7.92
22.59
6.41
11.72
13.67
10.42
16.18
14.8
47.87
17.52
36.72

65.41
52.66
41.12
66.29
155.17
52.08
40.72
21.24
18.17
33.31
19.56
27.28


0
0
0
0
0
5.66
0
0
0
0
0

8.4
19.12
0
0
0
0
0
0
11.78
0
37.94
15.66
13.23
0
0

31.36
40.82
20.82
48.27
34.4
72.63
126.62
81.27
187.41
35.37
67.09
27.85


0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
14.53


0
0
0
64.06
0
0
141.17
112.22
0
0
300

101.56
68.2
12.98
100
93.14
51.13
82.8
53.87
19.27
89.33
4.2
0
0
130.77
0

135.02
76.47
13.16
30.2
73.4
5.46
30.46
5.74
40.49
5.21
0
9.97


0
0
0
0
0
103.24
38.76
32.13
41.32
300
0

63.02
47.04
65.86
0
0
16.28
45.12
32.81
28.8
78.78
34.88
233.26
144.56
0
0

0
5.96
24.22
68.16
79.8
13.51
28.14
25.98
13.95
44.58
0
58.12


0
0
0
203.9
0
0
53.5
0
0
0
0

33.87
0
8.65
0
4.89
5.57
0
18.92
22.65
0
0
0
47.53
0
0

28.46
7.9
8.55
25.77
0
42.44
28.88
5.97
9.71
60.2
16.41
6.32


0
0
0
0
0
162.5
0
11.33
0
0
0

0
0
0
0
0
0
0
0
58.54
53.11
12.78
0
5.97
0
0

23.4
59.12
8.83
0
0
0
0
76.54
81.55
25.9
84.39
0


a TC1 indicates tree cluster 1; VC2 indicates vine cluster 2.


in thp nFA Arranged hy







Table 10--continued


Zone TC1 TC2 TC3 TC4 TC5 VC2 VC3 VC4 VC5


0 42.94
28.79 56.29


0
0
5.81
6.02

0
0
0
5.66
4.61
0
0
16.54
0
0
0


22.97
26.84
13.86
39.94

35.05
10.82
35.32
5.8
26.58
11.79
6.06
0
0
0
0


31
0

76.73
64.91
64.23
28.87

29.44
21.19
33.28
71.55
50.91
15.8
12.3
20.22
29.17
59.44
71.67


18.46 0 125.37 0 150
0 44.81 22.4 88.65 0


0
0
0
0

26.42
53.32
0
0
0
34.58
36.91
69.98
154.4
89.43
120.16


4.34
15.28
6.84
0

0
0
0
0
5.3
0
0
0
4.35
0
0


17.4
0
0
39.86

0
0
24.69
3.75
0
11.21
52.25
0
6.12
0
0


16.49
10.36
14.87
9.59

0
27.3
26.87
14.2
17.08
14.04
10.44
0
4.35
0
90.37


9.3
51.22
73.04
155.82

150
136.35
52.85
54.37
49.18
55.3
49.67
96.96
114.03
0
5.84


Table 11


Predicted Hydrologic Zones (Columns) and Actual Zones (Rows)
Based on DFA Results Using Only Tree Importance
Values


Actual
Hydrologic Predicted Hydrologic Zonesa
Zones 2 3 4 5 6 Total

2 11 0 0 0 0 11
3 1 14 0 0 0 15
4 0 2 10 2 0 14
5 0 0 0 4 0 4
6 0 0 0 3 8 11
12 16 10 10 8 55


a Misclassification rate = 14.6 percent.








A final model was generated to examine the effectiveness of using

the average site FTI numbers of all observed tree species at a site for

each of the 55 sites. This model correctly classified 45 of 55 sites for

an overall misclassification rate of 18.2 percent (Table 12). All

misclassifications were in an adjacent zone. Zones 2, 3, and 5 were mis-

classified twice; Zone 4 was misclassified 3 times; and Zone 6 was mis-

classified only once. Using FTI numbers for all observed tree species

improved the results obtained in the first DFA using the importance values

of only the commonly occurring species.

In examining the data (Table 13), it appears that three of the mis-

classifications occur at site 10 (Pearl River, observations 27 to 30).

The hydrology at this site was obtained from two gauges, and it is



Table 12

Predicted Hydrologic Zones (Columns) and Actual Zones (Rows)
Based on DFA Results Using Average FTI Values
for All Observed Tree Species at the Site

Actual
Hydrologic Predicted Hydrologic Zonesa
Zones 2 3 4 5 6 Total

2 9 2 0 0 0 11

3 0 13 2 0 0 15

4 0 2 11 1 0 14

5 0 0 2 2 0 4

6 0 0 0 1 10 11

9 17 15 4 10 55


a Misclassification rate = 18.2 percent.








Table 13

Cross-Validation Results of Zone Membership Using Linear
Discriminant Function Analysis


Posterior Probability of
Membership in Zone:


Observation


From
Zone


To
Zone

2

3

4

6

2

3

4

6

2

3

4

5

6

3

3a

3

4

3

4

3a

3


2
5

0.9876
0.0000
0.1759
0.0000
0.0000
0.1019
0.0000
0.0075
0.9876
0.0000
0.2436
0.0000
0.0000
0.0361
0.0000
0.0007
0.9801
0.0000
0.0137
0.0001
0.0000
0.0526
0.0000
0.6341
0.0000
0.0958
0.1244
0.0000
0.0012
0.0025
0.0597
0.0000
0.0003
0.0075
0.0806
0.0000
0.0000
0.1909
0.1909
0.0000
0.0748
0.0000


3
6

0.0124
0.0000
0.8212
0.0000
0.0062
0.0001
0.0000
0.9924
0.0124
0.0000
0.7548
0.0000
0.0376
0.0000
0.0000
0.9993
0.0199
0.0000
0.8865
0.0000
0.0202
0.0000
0.0000
0.0993
0.0000
0.9016
0.8703
0.0000
0.5429
0.0000
0.9245
0.0000
0.2747
0.0000
0.9092
0.0000
0.0017
0.0005
0.8079
0.0000
0.9138
0.0000


4

0.0000

0.0029

0.8919

0.0000

0.0000

0.0016

0.9263

0.0000

0.0000

0.0997

0.9273

0.2666

0.0026

0.0052

0.4534

0.0158

0.7175

0.0103

0.8069

0.0013

0.0114








Table 13--continued


Posterior Probability of
Membership in Zone:


Observation


From
Zone


To
Zone

4

3

4

3

6

3a

4a

4

5a

2

3

4

4a

3a

4a

6

2

3

4

6

2

5

6


2
5

0.0000
0.0594
0.2436
0.0000
0.0008
0.0036
0.0150
0.0001
0.0000
0.0002
0.0205
0.0000
0.0011
0.0027
0.0000
0.1286
0.0000
0.6307
0.8970
0.0000
0.3105
0.0000
0.0000
0.0526
0.0000
0.3664
0.0014
0.0021
0.0000
0.4502
0.0000
0.0785
0.9503
0.0000
0.2436
0.0000
0.0000
0.4057
0.0000
0.0013
0.9332
0.0000
0.0000
0.6216
0.0000
0.0013


3
6

0.0163
0.0000
0.7548
0.0000
0.4558
0.0000
0.8945
0.0000
0.0000
0.9998
0.9672
0.0000
0.4321
0.0000
0.0039
0.0002
0.0000
0.0044
0.1030
0.0000
0.6886
0.0000
0.0202
0.0000
0.0003
0.0060
0.5719
0.0000
0.0001
0.0119
0.0000
0.9196
0.0497
0.0000
0.7548
0.0000
0.0002
0.0047
0.0000
0.9987
0.0668
0.0000
0.0000
0.0699
0.0000
0.9987


4

0.9242

0.0016

0.5398

0.0904

0.0000

0.0124

0.5641

0.8673

0.3649

0.0000

0.0009

0.9273

0.6274

0.4246

0.5378

0.0019

0.0000

0.0016

0.5894

0.0000

0.0000

0.3085

0.0000








Table 13--continued


Posterior Probability of
Membership in Zone:


Observation


From
Zone


To
Zone

2

4a

6

2

3

4

6

2

3

5a

6


2
5

0.9503
0.0000
0.0001
0.0093
0.0000
0.0015
0.8729
0.0000
0.2144
0.0000
0.0000
0.1806
0.0000
0.0011
0.7969
0.0000
0.1004
0.0000
0.0000
0.7509
0.0000
0.0020


3
6

0.0497
0.0000
0.1163
0.0000
0.0000
0.9985
0.1271
0.0000
0.7835
0.0000
0.0020
0.0005
0.0000
0.9989
0.2031
0.0000
0.8923
0.0000
0.0000
0.0739
0.0000
0.9980


a Miclassified observation.


4

0.0000

0.8743

0.0000

0.0000

0.0020

0.8170

0.0000

0.0000

0.0074

0.1753

0.0000








possible that an error was made in combining the hydrologic data from

the two gauges. It seems ironic that the only zone correctly predicted

at this site was Zone 4. Table 13 also shows the percentage probability

of each of the 55 sites occurring in a zone.

Figure 11 shows the mean site FTI plotted against the observed and

predicted hydrologic zones. As expected, mean FTI numbers were greater

than the observed hydrologic zones at the low end (Zone 2) and less than

the zones at the upper end (Zone 6) of the hydrologic gradiant due to

the lack of outlying zones (e.g., Zones 1 and 7) to pull these averages

toward either extreme. Therefore, using DFA classification decision

points shows that average site FTI numbers as high as 3.45 would still

be in Zone 2, and average site FTI numbers as low as 5.33 would still be

in Zone 6. The lower end of the predicted Zone 4 (4.16) compares favor-

ably with the observed (4.0). Zone 5 predicted and observed zone values

are the same (5.0).


Regional Variation in Species FTI Numbers


Because the 17 sites in this study occur over a broad geographic

area, the possibility of regional differences in species FTI numbers was

a concern. To test for possible differences, the sites were grouped

into three regions: Gulf Coast (sites 1, 2, 10, 11, 12), Mississippi

Valley (sites 3 through 9), and Atlantic Coast (sites 13 through 17).

A two-factor analysis of variance (ANOVA) was used to test for

differences in importance values between regions and clusters for trees.

There was no significant interaction between region and cluster trees

(F = 0.71, p = 0.68); therefore, importance values of species within a

cluster do not differ among regions. There was a significant difference


















7



ZONE
6 6
w
I z
Z
o
N
5. 33
I ZONE 5 -
5 5.00 o
I-
I ZONE
< 4
4 | 4

U
4 4.16 S

I ZONE -
3 3
S3.45 a


ZONE
2



2 3 4 5 6
OBSERVED HYDROLOGIC ZONE












Figure 11. Mean tree FTI numbers plotted versus observed
and predicted hydrologic zones for all 55 sites








between regions (F = 4.02, p = 0.019), reflecting the fact that impor-

tance values of trees were greater in the Mississippi Valley, averaging

49.1 + 47.6, than the Gulf Coast (37.5 + 37.2) and Atlantic Coast

(39.3 34.4) regions. A number of factors may contribute to this phe-

nomenon, including stand maturity and localized disturbances.

Another two-factor ANOVA also was used to determine whether the

predicted zone values generated were more accurate in one region or

another. The absolute value of the difference between the predicted and

actual zones was used as the dependent variable. There was neither an

interaction (F = 1.44, p = 0.18), nor a main effect (F = 1.54, p = 0.22)

involving region, which indicates that the hydrologic zones can be pre-

dicted with equal accuracy among the specified regions.















SUMMARY AND CONCLUSIONS


Bottomland hardwood forests are dynamic and complex systems.

Frequent flooding from adjacent streams provides the forcing function

that characterizes the affected plant communities. Frequency and dura-

tion of floodwater determine the extent of anaerobic soil conditions

that directly affect plant populations. Plant species adapted for life

in anerobic soil conditions are located in the topographically lowest

areas subject to long duration flooding. Species composition changes as

the elevational and associated moisture gradient changes from wettest to

driest, and reflects species adaptations to the prevailing hydrologic

regimes.

Determination of a hydrologic gradient often requires extensive

data gathering over a long period. However, many studies have shown

that a definite relationship exists between plant species and the tim-

ing, frequency, and duration of inundation and soil saturation (Larson

et al. 1981). This study was undertaken to express quantitatively the

optimum position of various plant species along a hydrologic gradient.

Previous studies have estimated the location of plant species and

communities along a hydrologic gradient. Various systems have been

proposed that use vegetation to predict the duration and/or frequency of

flooding. However, previous studies were limited to a small geographic

area, the developed systems are qualitative, and vegetation data used to








predict the degree of flooding for the entire southeastern United States

previously have been literature-based involving many studies with vary-

ing research designs.

Vegetation data resulting from this study related four vegetation

strata and three life forms occurring in 55 stands at 17 sites through-

out the southeastern United States. Hydrologic regimes were calculated

for a 10- to 20-year period of record for each stand. A flood tolerance

index (FTI) system of weighted averages based on importance values was

developed, and FTI numbers were calculated for various life stages of

each species identified in the study.

Three hundred and twelve species were identified for each of

4 strata in the study including 74 tree species, 188 species of saplings

and shrubs, 31 species of woody vines, and 268 species of herbs and

woody seedlings. Comparison of the FTI numbers with two other systems

(Hook 1984; Reed 1988) using vegetation to estimate wetness showed gen-

eral agreement among the systems, especially for mature trees.

Cluster analysis and discriminant function analysis were used to

evaluate the weighted averaging technique and explore the best method

for using the FTI numbers in predicting hydrologic regimes.

Tree, sapling, and vine data clustered into distinct groups. Her-

baceous and shrub data did not group distinctively. Tree and vine

importance values for each cluster in a zone/sample (data taken in a

single zone at a site) and FTI numbers for tree data were used as inde-

pendent variables for the discriminant function analyses. Tree species

were found to be more useful than saplings, shrubs, vines, or herbaceous

species in predicting hydrologic zones. The tree data alone using

importance values provided 85 percent accuracy. Tree data alone using







FTI numbers was only slightly less accurate at 82 percent. All misclas-

sifications assigned membership to a neighboring zone. Misclassifica-

tions are understandable for two important reasons. Zone 4 contains the

more facultative species because as wetness decreases, other environmen-

tal conditions begin to exert greater influence. Also, since Zone 5 is

so narrow compared to other zones and most species occurring in Zone 5

occur in greater abundance in either Zone 4 or 6, very little difference

in the community structure exists between the top of Zone 4 and the

bottom of Zone 6.

The accuracy of these predictions may be somewhat inflated,

because hydrologic zone was a parameter used to derive species FTI

numbers.

There were no regional (Gulf Coast, Lower Mississippi Valley, and

Atlantic Coast) differences in the accuracy of the weighted averaging

and predicted values. Therefore, a single FTI number calculated for

each species can be used to predict zones for the entire study area.

The implication of this study is that the calculated FTI numbers

can be used to estimate hydrologic regimes in bottomland forest systems

of the southeastern United States. Trees were determined to be the most

reliable vegetative growth form for determining hydrologic zones. How-

ever, this study was conducted in relatively undisturbed areas. Because

trees can remain for decades following hydrologic disturbance, a modifi-

cation of the method using saplings and seedlings may prove to be more

reliable.

Techniques used in this study to develop FTI numbers in the south-

eastern United States may be applicable to regions of the country that

have similar types of riverine forest systems.








































APPENDIX A

SITE DESCRIPTIONS AND MAP LOCATIONS















Neches River (Sites 1 and 2)


Location (Neches River Basin) These sites are located in the National
Big Thicket Preserve, Jack Gore Baygall unit, 6.4 km north of Evadale in
Jasper County, Texas. Reference U.S. Geological Survey (USGS) map,
Silsbee, Texas, N3015-W9400/15, 1955.


Hydrology data
staff gauge on
gauge datum to
profile.


- Twenty years of hydrology data were obtained for a
U.S. Highway 96 bridge at Evadale. Slope correction from
study site was determined by using a water surface


General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatica in deep sloughs to Quercus alba-Pinus taeda and Fagus
grandifolia on the nearby ridges. Intermediate communities consist of
Quercus lyrata, Carya aquatica, Quercus michauxii, Liquidambar styraci-
flua, Ulmus americana, and Carpinus caroliniana.

Soils and climate Soils vary from the very poorly drained Angelina
series in sloughs to the moderately well drained Spruger series on
ridges. Other soil series encountered were Bleakwood, Urbo, and
Attoyac. Average annual rainfall in the area is 170 cm, and the average
growing season is 234 days.


Delineated zones Zones 2, 3, 4,
Zone 5 was too narrow to reliably
slope of the floodplain terrace.


and 6 were delineated for both sites.
delineate due to its position on the























































Liu
r2 Z
z a








Steele Bayou (Site 3)

Location (Yazoo River Basin) This site is located in the Yazoo
National Wildlife Refuge, 6.4 km northeast of Glen Allen in Washington
County, Mississippi. Reference USGS map, Percy, Mississippi, N3300-
W9052.5/7.5, 1967.

Hydrology data Twenty years of hydrology data were obtained for a
gauge on the bridge over Steele Bayou 6.4 km south of Grace, Missis-
sippi. Slope correction from gauge datum to study site was determined
by a water surface profile.

General vegetation Plant communities ranged from a Salix nigra-
Taxodium distichum community at lower elevations to a Sassafrass
albidum-Liquidambar styraciflua-Quercus nigra community at the highest
elevation. Intermediate communities are dominated by Planera aquatica,
Forestiera acuminata, Quercus lyrata, Carya aquatica, Fraxinus pennsyl-
vanica, Celtis laevigata, Acer negundo, and Ulmus americana.

Soils and climate Soil series range from Sharkey at the lowest eleva-
tions to Dundee at the higher elevations. Average annual rainfall in
the area is 132 cm and the average growing season is 213 days.

Delineated zones Zones 2 through 6 were delineated for study at this
site.




66










S\ \
SI \ \
S> l \ \ o

\ \ /\
o I I

V < //, / z a

oI o
Ii




a a\ \ /
,// \
10t1
U < U- 3
N 0 a a* 0

< U.4




I- i


-\\\Js
InE
/1UlN
UJ ____ ( fc^








Ouachita River (Sites 4 and 5)

Location (Ouachita River Basin) These sites are located in the Felsen-
thal National Wildlife Refuge, 8 km west of Crossett and .8 km east of
Felsenthal, respectively, in Union County, Arkansas. Reference USGS
map, Felsenthal, Arkansas-Louisiana, N3300-W9200/15, 1981.

Hydrology data Nineteen years of hydrology data were obtained for a
gauge on a U.S. Highway 81 bridge, 8 km west of Crossett, Arkansas. A
slope correction for site 5 was determined using a water surface pro-
file. Site 4 was adjacent to the gauge, so no correction was necessary.


General vegetation Plant communities range from a Taxodium distichum-
Cephalanthus occidentalis dominated community in lower areas to a nearly
monotypic stand of Pinus taeda in higher areas. Intermediate communi-
ties are dominated Carya aquatica, Quercus lyrata, Diospyros virginiana,
Quercus phellos, Quercus nuttallii, and Liquidambar styraciflua.

Soils and climate All encountered soils are in the Una series. Aver-
age annual rainfall is 140 cm and the average growing season is
211 days.

Delineated zones Only Zones 3 and 4 were delineated for both sites 4
and 5. Zone 2 was not used because the hydrology was not reflected by
the gauge data. Zones 5 and 6 were not used due to major disturbances
from recent silvicultural and agricultural practices.




68





l SITE 4


A. LINE
ZONE 4
ZONE3 PLOT4
PLOT 3 't
SWAMP,, --. \ U.S. HWY 82










I I

I



I I



















SITE 5 S' -
I'






I I
I I

















ZONE 3
II






















ZONE4
PLOT 4 I
DAM UNDER CONSTRUCTION
SNOT TO SCALE


Figure A-3. Ouachita River (sites 4 and 5)








Yazoo River (Site 6)

Location (Yazoo River Basin) This site is located on the north side of
the Yazoo River, 8 km west of the U.S. Highway 61 bridge and 12.0 miles
north of Vicksburg, Mississippi in Issaquena County. Reference USGS
map, Long Lake, Mississippi-Louisiana, N3222.5-W9052.5/7.5, 1962.

Hydrology data Twenty years of hydrologic zone elevations were com-
puted by analyzing flow data from gauges on the Mississippi River at
Vicksburg, Mississippi, on the Yazoo River 2.4 km east of the site, and
at the Steele Bayou control structure immediately adjacent to the study
site.

General vegetation Plant communities range from Quercus lyrata-Carya
aquatica at lowest elevations to a Liquidambar styraciflua-Ulmus ameri-
cana-Celtis laevigata association at the highest elevations. Other
commonly occurring species include Ilex decidua, Carya illinoensis, and
Cercis canadensis.

Soils and climate Soils were determined to be in the Sharkey series.
Average annual rainfall is 127 cm. The average growing season is
221 days.

Delineated zones for study Only Zones 3 and 4 were delineated for this
study. Zones 2, 5, and 6 were either too narrow or too disturbed to
provide reliable data.





70









































co





o>
"r4
0

'4-





0
N


0s





<~








Big Black River (Site 7)

Location (Big Black River Basin) The site is located on the south bank
of the Big Black River adjacent to the Fisher Ferry bridge on Fisher
Ferry Road, 24 km south-southeast of Vicksburg, Mississippi. The site
is in Claiborne County, Mississippi. References USGS map, name N3207.5-
W9045/7.5, 1963.

Hydrology data Twenty years of hydrologic data were analyzed for a
flow gauge on the U.S. Highway 80 bridge, 3.7 km east of Bovina, Missis-
sippi. A slope correction from gauge location to site was determined by
a water surface profile.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatica at lower elevations to Liquidambar styraciflua-Celtis
laevigata-Ulmus americana at higher elevations. Intermediate communi-
ties are dominated by Planera aquatica, Carya aquatic, Quercus lyrata,
and Fraxinus pennsylvanica.

Soils and climate Soils range from the Waverly series (depressional
phase) in lowest elevations to the Faylaya series at higher elevations.
Average annual rainfall for this area is 132 cm and the average growing
season is 226 days.

Delineated zones Zones 2, 3, and 4 were delineated for study. Zones 5
and 6 were not delineated due to major vegetation disturbance induced by
silvicultural and agricultural practices.































































Figure A-5. Big Black River (site 7)








L'Anguille River (Sites 8 and 9)

Location (L'Anguille River Basin) Sites 8 and 9 are located on the
west bank of L'Anguille River, 0.8 km east and 7.2 km southeast, respec-
tively, of Palestine in St. Francis County, Arkansas. Reference USGS
map, Marianna, Arkansas, N3445-W9045/15, 1957.

Hydrology data Twenty years of hydrologic data were analyzed for a
gauge located on the U.S. Highway 70 bridge, 0.8 km east of Palestine,
Arkansas. A slope correction was computed for site 9 using a water
surface profile. No slope correction was necessary for site 8 because
it was adjacent to the gauge.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatic dominated communities at the lowest elevations to a Carya
tomentosa-Quercus alba-Liquidambar styraciflua dominated association on
adjacent ridges. Intermediate communities are dominated by Quercus
lyrata, Carya aquatica, Diospyros virginiana, Fraxinus pennsylvanica,
and Ulmus americana.

Soils and climate Soil series range from Zachary at lower elevations
to Loring on adjacent ridges. Average annual rainfall is 132 cm, and
the average growing season is 219 days.

Delineated zones Zones 3 and 4 were delineated for sites 8. Zones 3
and 6 were delineated for site 9. All other zones were unacceptable.













HWY
U ELECTRIC
SUBSTATION


II

II
(1


WOODEN
BRIDGE \ RD

II
| HOUSES



II
ELCANNON & II SOYBEAN
AME CHURCH I FIELD
II

HII
FIELDS II HOUSES


NOT TO SCALE


L'Anguille River (sites 8 and 9)


Figure A-6.








Pearl River (Site 10)

Location (Pearl River Basin) This site is located in the Pearl River
State Wildlife Management Area, 8 km north of Slidell in St. Tammany
Parish, Louisiana. Reference USGS map Nicholson, Mississippi-Louisiana,
N3022.5-W8937.5/7.5, 1955.

Hydrology data Twenty years of hydrology data were analyzed for two
gauges. First 10 years data was extrapolated to present gauge on the
Southern Railway bridge at Pearl River, Louisiana. No slope correction
was necessary because the site is adjacent to the gauge.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatic dominated communities in sloughs to Liquidambar
styraciflua-Quercus nigra dominated communities on low ridges. Interme-
diate communities are dominated by Quercus laurifolia, Acer drummondii,
Fraxinus pennsylvanica, Carpinus caroliniana, and Ilex opaca.

Soils and climate Soil series range from Rosebloom (depressional
phase) in sloughs to Prentiss on ridges with Arkabutla at intermediate
elevations. Average annual rainfall is 152 cm, and the average growing
season is 237 days.

Delineated zones Zones 2, 3, 4, and 6 were delineated. Zone 5 could
not be reliably delineated due to topography.

































































Figure A-7.


Pearl River (site 10)







Apalachicola River (Sites 11 and 12)

Location Site 11 is located on the west bank of the Apalachicola
River, immediately south of the Florida Highway 20 bridge, 1.6 km east
of Blountstown in Calhoun County, Florida. Site 12 is located on the
east bank of the river, 4.8 km north of Bristol in Liberty County,
Florida. Reference USGS maps Blountstown, Florida, N3022.5-W8500/7.5,
1945, and Bristol, Florida, N3022.5-W8452.5/7.5, 1945, respectively.

Hydrology data Twenty years of hydrology data for Site 11 were ana-
lyzed for a gauge located 0.8 km south of Highway 20 bridge at the Neal
Lumber Company Landing. Hydrology data for Site 12 were analyzed from
data from a previous study (Leitman et al. 1984).

General vegetation Vegetation for both sites ranges from Nyssa
aquatica--dominated swamps at lower elevations to Nyssa sylvatica-
Juglans nigra-Sassafras albidum dominated associations at higher eleva-
tions. Intermediate plant communities consist of Fraxinus caroliniana,
Gleditsia aquatica, Quercus lyrata, Carya aquatica, Ulmus americana,
Melia azederach, Celtis laevigata, and Quercus nigra.

Soils and climate Soil series range from Bibb at lower elevation to
Ochlochonee at higher elevations. Soils series occurring at intermedi-
ate elevations were Chastain, Enoree, Jena, and Chewacla. Average
annual rainfall is 137 cm, and the average growing season is 267 days.

Delineated zones Zones delineated for site 11 were 2, 3, 4, and 5.
Essentially all of Zone 6 has been developed for agriculture. Zones
delineated for site 12 were 4, 5, and 6. The hydrology of Zones 2 and 3
had been altered by an extensive network of beaver dams and was not
reliable.













APALACHIC


OLA RIVER


BERM


PLOT 5
CmC


ZONE 5- -"


STAFF
STAFF


pLOT4

ZONE
Z014f-4


GAUGE __.--SLOUGH----


SITE 11


ZONE3


NOT TO SCALE


Figure A-8. Apalachicola River (site 11)


.7






79


































r-4

414

"4




-1.

Cd
Ci
r-4
O
0
ce
*-I


C
-4



r)
*r












*-I
(fl








Ocmulgee River (Site 13)

Location (Ocmulgee River Basin) This site is located across the river
from Lumber City, and adjacent to the east side of Southern Railway and
U.S. Highway 23 and 341 in Jeff Davis County, Georgia. Reference USGS
map, Lumber City, Georgia, N3152.5-W8237.5/7.5, 1971.

Hydrology data Twenty years of hydrology data were analyzed for a
gauge on the U.S. Highway 23 and 341 bridge adjacent to the site. No
slope correction was necessary.

General vegetation The plant communities range from Taxodium
distichum-Nyssa aquatica communities at the lowest elevations to a Carya
tomentosa-Quercus alba-Pinus glabra dominated association at the higher
elevations. Intermediate communities consist of Planera aquatica, Quer-
cus lyrata, Carya aquatica, Ulmus americana, Liquidambar styraciflua,
Quercus phellos, Carpinus caroliniana, and Quercus nigra.

Soils and climate The soil series range from Bibb in the lowest areas
to Riverview at the highest elevations. The Chastain series occurs at
intermediate elevations. Average annual rainfall in this area is
117 cm, and the average growing season is 232 days.

Designated zones Zones 2, 3, 4, and 6 were delineated for study.
Zone 5 was too narrow to provide reliable data.






























































Figure A-10. Ocmulgee River (site 13)








Altamaha River (Site 14)

Location (Altamaha River basin) This site is located in the northeast
quadrant at the intersection of U.S. Highway 1 and the Altamaha River,
50 km north of Baxley in Toombs County, Georgia. Reference USGS map,
Baxley NE, Georgia, N3152.5-W8215/7.5, 1970.

Hydrology data Twelve years of hydrology data were analyzed for a
gauge on the U.S. Highway 1 bridge adjacent to the site. No slope cor-
rection was necessary.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatica at lowest elevations to a Juniperus virginiana-Quercus
stellata-Carya tomentosa-Pinus taeda community at highest elevations.
Intermediate communities are dominated by Fraxinus pennsylvanica, Quer-
cus michauxii, Quercus phellos, and Carpinus caroliniana.

Soils and climate Soils range from the Osier series in lowest eleva-
tions to the Riverview series at highest elevations. The Chewacla
series occurs at intermediate elevations. Average annual rainfall is
117 cm, and the average growing season is 232 days.

Designated zones Zones delineated for this study site were 2, 5, and
6. Zones 3 and 4 had ridge and swale topography which prevented sepa-
rating them reliably.


















SITE 14


I


//


U I
01





I /
I' ,


RIDGE & SALE






RIDGE & SWALE


^ ^.~-----''
ZONE 5
4 ( B PLOT 5

I-----
.----- --r --- SLOUGH --- ---
ZONE 6

A1 PUCPLOT 6
SLOUGH-- ___ I ZONE 2
-- ------R-- PLOT 2- ---- SLOUGH-
RIVER
GAUGE TBM, BERM

FoL O ALTAMAHA RIVER

NUCLEAR is
I _PLANT
.-------------------------~~I----------------


NOT TO SCALE


Figure A-11. Altamaha River (site 14)


v-F


I

a


USGS
BM
201-Z



I I

I

I
1a1


IO l
It
I r1

I I
I I
I I



BT--

BOAT RAMP


1

I








Edisto River (Site 15)

Location (Edisto River Basin) This site is in Givhans Ferry State
Park, north and west of South Carolina Highway 61 bridge, 4.8 km west of
Givhans in Colleton County, South Carolina. Reference USGS map, Maple
Cane Swamp, South Carolina, N3300-W8022.5/7.5, 1979.

Hydrology data Twenty years of hydrology data were analyzed for a
gauge on the South Carolina Highway 61 bridge adjacent to the site. No
slope correction was necessary at this site.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatica-Fraxinus caroliniana at lowest elevations to a Pinus
taeda-Quercus virginiana-Quercus nigra dominated community at highest
elevation. Species in communities at intermediate elevations include
Quercus lyrata, Quercus laurifolia, Planera aquatica, Liquidambar styra-
ciflua, Quercus nigra, and Carpinus caroliniana.

Soils and climate Soil series range from Osier at lowest elevations to
Chipley at highest elevations. The Torhunta soil series occurs at
intermediate elevations. Average annual rainfall is 132 cm, and the
average growing season is 213 days.

Delineated zones Zones 2, 3, 4, and 6 were delineated, but Zone 5 was
too narrow due to its topographic position.







85












--
ua
0
I-
z




-C -







03 CL0
A-I3dO Id 3.LVS / -o \





) 4. )
NL -. "I



/ ,/ y / 1.zo D
I





I' I
I I




\
II I




\ I

I-,
i" N^ /
v) '\/








Lynches River (Site 16)

Location (Lynches River Basin) This site is located in Lynches River
State Park, 1.6 km south of Effingham in Florence County, South Caro-
lina. Reference USGS map, Florence West, South Carolina, N3400-
W7945/15, 1940.

Hydrology data Twenty years of hydrology data were analyzed for a
gauge on the bridge on U.S. Highway 52, 1.6 km south of Effingham. A
slope correction from gauge to site was determined by a water surface
profile.

General vegetation Plant communities range from Taxodium distichum-
Nyssa aquatic at the lowest elevations to Quercus falcata-Quercus
stellata-Carya tomentosa at higher elevations. Species occurring in
communities at intermediate elevations include Quercus lyrata, Quercus
laurifolia, Liquidambar styraciflua, and Quercus phellos.

Soils and climate Soil series range from Chastain at lowest elevations
to Chipley at highest elevations, with the Wehadkee series at intermedi-
ate elevations. Average annual rainfall is 107 cm, and the average
growing season is 237 days.

Delineated zones Zones 2, 3, and 6 were delineated for study. Zones 4
and 5 were too narrow to delineate due to their topographic positions.






























































Figure A-13. Lynches River (site 16)