Vegetative regeneration in wetland forests of Florida

MISSING IMAGE

Material Information

Title:
Vegetative regeneration in wetland forests of Florida
Physical Description:
ix, 195 leaves : ill. ; 29 cm.
Language:
English
Creator:
Davis, Mary M
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 184-194).
Statement of Responsibility:
by Mary M. Davis.
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 - 001613899
notis - AHN8317
oclc - 23536018
System ID:
AA00003742:00001

Full Text










VEGETATIVE REGENERATION IN WETLAND
FORESTS OF FLORIDA


















by

MARY M. DAVIS
17


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


UNIVERSITY OF FLORIDA LIBRARIES


1990
Wap -b












ACKNOWLEDGMENTS

I would like to thank Jack Putz and Ronnie Best for

their support and guidance; their energy and dedication to

science serve as career standards to which I aspire. Kathy

Ewel and Jon Johnson were particularly helpful in the design

of the experimental portion of this research; their comments

were welcome throughout this research project. Walter Judd

very kindly agreed to join my supervisory committee after my

research was well underway; his time and comments were

appreciated.

The Lake Oklawaha study was greatly facilitated by

Stephen Tennenbaum, without whose energy and expertise with

calipers the project would not have been done as easily.

Jeff Bielling, Charlotte Pezeshki, Jack Pierce, and Karen

Merritt were very helpful in the field and with data entry.

Special thanks go to Jon van der Venter and the U.S. Army

Corps of Engineers for their time, assistance, and airboat.

The survey of natural wetlands in north-central Florida

was an enjoyable experience because of the many wonderful

people from the Center for Wetlands. In particular, I would

like to thank Mark Brown, Pete Straub, Jim Feiertag, Scott

Swank, Jack Pierce, Shane Best, Bob Tighe, John Richardson,

Linda Crowder, and Pete Wallace for their help with the many

aspects of BIGPHOS.







Bea Pace, Bill Davis, and Maggie Davis provided wel-

comed extra hands and energy for the apical dominance ex-

periments. I enjoyed working with them very much, they made

the long and tedious tasks easier. Steve Linda provided

statistical counseling for all aspects of this research.

I have already acknowledged the help of my husband,

Bill, and my daughter, Maggie, for their help with my ex-

periments, but I especially want to thank them for their

understanding support through all my highs and lows. Of all

the people I have been involved with during this program, I

would not have wanted to go through it without Bill and

Maggie.

My parents, Howard and Margaret Butler, and my sisters,

Pat, Lois, and Fran, have been wondering for a long time

when I was going to get out of school and get a job. In

spite of this, my family has been very patient and suppor-

tive all these years, and I have a great deal of respect for

them. Each of my family members has chosen a different path

to follow in life, and although each one of us had the

capability, I was the one that chose to pursue a higher

education. It is for this reason that as I celebrate my

graduation, I celebrate it with and for my family.


iii














TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS. .. . ... ii


ABSTRACT . . vii


CHAPTERS

1 PATTERNS, PROCESSES, AND MECHANISMS OF VEGETATIVE
REGENERATION . ........ .. 1

Introduction . .... 1
Definition of Vegetative Regeneration 2
Consequences of Vegetative Regeneration 4
Phenotypic Plasticity. . .. 5
Distribution of Vegetative Regeneration .. 7
Sprouting in Stressful Environments 8
Stress Induced by Inundation . 9
Mechanisms of Apical Dominance .. ... .12
Research Objectives . 15


2 RECOVERY BY SPROUT PRODUCTION IN THE IMPOUNDED
FLOODPLAIN FOREST OF LAKE OKLAWAHA, FLORIDA 20

Introduction . .20
Site Description. . ... 21
Methods . ... 26
Results . . 27
Species Distributions of the Natural
Floodplain Forest . .. 27
Stem Mortality in Lake Oklawaha 30
Species Distribution in Lake Oklawaha 30
Recovery Following Impoundment 35
Discussion . . .. 39


3 VEGETATIVE REGENERATION IN NORTH-CENTRAL
FLORIDA WETLANDS . . 44

Introduction .. ... . 44
Site Descriptions. . .. .48








Methods . . 58
Transect Establishment . 58
Ground Surface Profile .. 59
Ground Water Fluctuation .. 59
Soil Chemistry . .. 60
Tree Species Distributions ... .61
Data Analysis . . 62
Results . . .62
Wetland Tree Species Distributions 62
Sprout Production by Dicot Tree Species 67
Distribution of Stems Producing Sprouts 72
Distribution of Stems of Sprout Origin 73
Sprouting Conditions of Common Wetland
Tree Species. . .. .74
Discussion . 83


4 APICAL DOMINANCE: MECHANISMS AND ECOLOGICAL
SIGNIFICANCE . . .. 88

Introduction . . 86
Methods . . 93
Red Maple Field Experiment .. .93
Shadehouse Experiment. ... 95
Humidity Chamber Experiment .. 101
Results . . .. 102
Red Maple Field Experiment. .. 102
Shadehouse Experiment .. .105
Humidity Chamber Experiment 139
Discussion. .. .... .... 145
Mechanisms of Apical Dominance .. 145
Plant Water Relations of Flooded
Seedlings. .. . 147
Transpiration Rates and Humidity 149
Apical Dominance of Flooded Plants 150
Sprout Production in Wetlands 151


5 SUMMARY AND CONCLUSIONS. .... 152

Sprout Production and Growth in Wetlands 152
Effects of Inundation on Vegetative
Regeneration . 153
Effects of Inundation on Mechanisms of
Apical Dominance . 154
Sprouting and Recovery of Chronically
Stressed Trees. . .155
Intraspecific Variation in Strength of
Apical Dominance. . .. .156
Vegetative Regeneration in Wetlands ... 157









Consequences of Sprouts in Wetland
Populations ........... 158
Vegetative Regeneration in Mature Seral Stage
Communities .. .... 159
Effects of Stress on Reproductive Tactics .. 159
Vegetative Regeneration in Disturbance-
Maintained Communities .163
Maintaining Wetland Processes in a Changing
Landscape . 164
Additional Research of Inundation
and Sprouting . .. 165


APPENDIX . . 166


REFERENCES . . 184


BIOGRAPHICAL SKETCH ... . .195












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

VEGETATIVE REGENERATION IN WETLAND
FORESTS OF FLORIDA

by

Mary M. Davis

August 1990


Chairman: Francis E. Putz
Cochairman: G. Ronnie Best
Major Department: Botany

Environmental stress on plants limits reproductive

success, often with differential effects on vegetative and

sexual reproduction. Effects of stress on reproduction

differ among species depending on degree of stress tolerance

and capacity to recover following stressful events. Conse-

quently, relative contributions of seedlings and sprouts in

natural communities vary with disturbance regime and

community composition.

This study investigated the relationship between

hydrologic regime and basal sprouting of trees in north-

central Florida wetlands. Basal sprouting patterns in

natural wetlands were compared with sprouting responses in

those parts of the Oklawaha River floodplain that had been

artificially impounded for 18 years. Interpretations of


vii







sprouting responses in relation to hydrologic regimes were

based on experimental investigations of apical dominance

mechanisms.

Trees sprouted throughout the range of natural hydro-

logic settings. The relative abundance of sprouts in na-

tural communities depended on tree species composition and

growing conditions. Sprouts were generally more common

among dicots rooted on hummocks in wetlands with brief but

frequent periods of deep inundation.

Stems that originated as basal sprouts were common in

natural, unimpounded portions of the Oklawaha River flood-

plain forest but became more abundant with increasing water

depth in Lake Oklawaha. Trees have persisted in the reser-

voir either by simply tolerating altered hydrologic regimes

or by increasing sprout production, thereby rejuvenating

individuals. Contrary to predictions from earlier studies

of this system, species richness has remained high in

standing water depths up to 0.6 m due to this sprouting

response.

Experimental investigations of apical dominance

mechanisms showed that inundation does not affect sprout

production either by direct anoxia-induced physiological

responses or indirect responses to altered root-to-shoot

ratios following inundation. Flooding inhibited growth of

sprouts, as well as of parent stems, due to increased

stomatal resistances.


viii







Successful regeneration by wetland trees depends on

avoidance of inundation; continual production of basal

sprouts and rapid sprout growth are primary factors

contributing to the success of vegetative regeneration in

wetlands. Successful vegetative regeneration after

disturbances such as inundation reduces species turnover

rates, stabilizes community structure, and provides a

survival mechanism that allows individuals to survive

despite long-term stress.












CHAPTER 1
PATTERNS, PROCESSES AND MECHANISMS OF
VEGETATIVE REGENERATION



Introduction

Regeneration of wetland plants is limited by the

presence of excess water. Seeds of most species do not

germinate under water, and seedlings die if submerged.

Establishment is, therefore, restricted to extended periods

of soil exposure during which seedlings can grow above the

deleterious affects of subsequent inundation. Annual inun-

dation of natural wetlands is, however, fairly predictable

(sensu Colwell 1974), and suitable periods for seedling

establishment in wetlands are infrequent. Alternatively,

angiosperms and certain gymnosperms are capable of sprout-

ing, a form of vegetative regeneration, which may not depend

upon exposed soils.

Population regeneration and survival depend on inter-

actions between genetically controlled characteristics, such

as sprouting, and the environment (Bazzaz 1984). In wet-

lands, where seedling establishment is relatively uncertain

(Huenneke and Sharitz 1986), relatively greater investment

by populations in vegetative regeneration may be advanta-

geous.









Patterns, processes, and mechanisms of vegetative

regeneration are therefore integral to the understanding of

plant demography and ecology in wetlands and in other natur-

al systems in general. In order to understand patterns of

vegetative regeneration and the role of sprouts in wetland

community dynamics, it is necessary to determine what

species-specific and environmental conditions affect sprout

initiation, growth, and survival.



Definition of Vegetative Regeneration

Vegetative regeneration is the production of genetical-

ly identical offspring that often remain physically and

physiologically connected to the parent stem (Silvertown

1984, Pitelka and Ashmun 1985). Through vegetative regener-

ation, genetic individuals are able to increase stem num-

bers, disperse into under-exploited areas (e.g., Holbrook

and Putz 1982), expand areas of dominance (e.g., gap

closure), recover from injury (Powell and Tryon 1979,

Malanson and Westman 1985, Terwilliger and Ewel 1986), and

survive in spite of severe or chronic stress (Held 1983,

Paillet 1984).

Plants use a variety of mechanisms to reproduce vegeta-

tively. Regeneration and dispersal of ramets (genetically

identical individuals) are often accomplished with rhizomes

and stolons. Vegetative reproduction of individuals capable

of independence from the parent stem is also possible by









division of underground bulbs, tillering, and air-layering.

Vegetative expansion, recovery, and maintenance of individu-

al stems, however, are the results of sprouting: the produc-

tion and expansion of buds.

Sprouts arise from either axillary or adventitious buds

(Esau 1960). Axillary buds differentiate from leaf tissue

as meristematic cells in axils of leaf primordia. When

axillary buds remain dormant, secondary growth carries the

bud away from its point of origin; vascular connections with

the main axis are developed and maintained. In contrast,

adventitious buds develop from meristematic cells differen-

tiated from callus tissue and may arise during any point of

development, either on the stem surface or deeply seated in

stem or root tissues.

Sprout type can be determined by the bud origin, loca-

tion on the stem, and conditions that stimulated bud growth

(Kramer and Kozlowski 1960). For example, basal or stump

sprouts often arise from dormant buds on the lower tree

trunk or root crown. Epicormic branching, dormant buds

sprouting on main stems and branches, is often associated

with increased light levels (Blum 1963, Smith 1966, Trimble

and Smith 1970, but see Wilson 1979, Bryan and Lanner 1981).

Root sprouts are adventitious.









Consequences of Vegetative Regeneration

The degree to which plants are able to produce sprouts

and the effect of sprouts on plant survival have important

implications for population and community dynamics. The

principal effect of sprouting is the reduction of mortality

rates of established genetic individuals. Increased plant

longevity stabilizes populations, while genetic variability

is maintained (Hamrick 1979). Species turnover rates are

slower and possibly arrested in communities with a high

proportion of species that can sprout. This is apparently

the case in some mature plant communities that are stable in

species composition (Odum 1969) and have relatively more

successful vegetative than sexual reproductive efforts

(Abrahamson 1980).

Interpretation of stand age structure and history

becomes more complicated when age distributions of stems do

not match root age distributions. For example, 58% of

seedlings sampled in hardwood forests of central Pennsyl-

vania had younger shoots than roots; young stems had roots

ranging up to 25 years old (Ward 1966). Under these condi-

tions, standard population ecological methods using stem

size distributions to deduce age relationships and patterns

of regeneration and mortality of genetic individuals are









inadequate. Application of models describing density-

dependent relationships and competition for resources be-

comes very difficult in communities with a high rate of

sprouting.

Sprouting has further consequences for natural selec-

tion and population distributions. For example, competition

among genetic individuals over time results in locally

adapted clones (Abrahamson 1980). Furthermore, spatial

distributions of ramets are generally more aggregated than

are genetic individuals (Huenneke 1985).



Phenotypic Plasticity

Phenotypic plasticity is variation in expression of a

character that can be altered by environmental conditions

(Bradshaw 1965). Degree of plasticity is variable both

within and among species. Consequently, the degree of

character plasticity is species-specific, varies with en-

vironmental conditions, and is alterable by selection

(Abbott 1976a).

Environmental conditions influence plant phenotypes by

inducing changes in physiological and morphological charac-

ters and by favoring existing phenotypes, particularly if

environmental changes are rapid (Bradshaw 1965). Phenotypic

plasticity has been noted in response to many environmental

conditions, including inundation, degree of exposure, and

fertilization. For example, finely dissected leaves are









found on submerged portions of aquatic plants, such as

Utricularia inflata Walt. and Myriophyllum heterophyllum

Michx., while emergent leaves of the same plant are entire

(Godfrey and Wooten 1981). Genetic variability often paral-

lels variation in exposure; the relative dwarf habit of

Senecio vulgaris L. progeny from plants exposed to strong

winds on a cliff site in Wales was retained throughout

development, but there was greater variation in heights of

progeny from sites with variable levels of exposure (Abbott

1976b).

Phenotypes tolerant of rapid environmental changes are

favored when plants are unable to avoid damage (Bradshaw

1965). For example, Held (1983) reported a partial shift

from seedling establishment of American beech (Faqus gran-

difolia Ehrh.) toward regeneration by root sprouts in areas

where the climate is severe, presumably because root sprouts

are more tolerant of freezing than young seedlings. Similar

trends were reported from tropical deciduous and sub-

tropical forests of India (Khan et al. 1986).

Increased longevity of sprouting individuals increases

the chances that plants will experience changes in environ-

mental conditions. Long-lived perennial plants exhibit a

large degree of phenotypic plasticity by sprouting under

different environmental conditions (Jefferies 1984). For

example, allometric diversity of sprouting plants allows

adjustment to changes in availability of water, major







7

inorganic nutrients, light, and carbon dioxide, and to shoot

and root pruning (Wilson 1988).



Distribution of Vegetative Regeneration

Patterns of sprout production are the result of inter-

actions between a species' propensity to sprout (i.e.,

species-specific apical dominance strength) and environmen-

tal conditions. Under environmental conditions favorable

for sprout production, species will differ in sprout produc-

tion rates depending on their specific strengths of apical

dominance (Kramer and Kozlowski 1979). Within communities,

proportions of stems of sprout origin are therefore partial-

ly a function of species composition.

Sprout production is more common in lower plants and

angiosperms than gymnosperms due to the typically strong

apical dominance exhibited by the latter (Kramer and Koz-

lowski 1979). Abrahamson (1980) reports variable occurren-

ces of vegetative reproduction in angiosperm-dominated

communities. Plants commonly reproducing vegetatively have

been reported from all latitudes and for both terrestrial

and aquatic species. The relative importance of vegetative

to sexual reproduction of trees, however, decreases in the

tropics.

Sprouting is thought to be favored in late successional

and disturbance-maintained communities (Bradshaw 1965,

Abrahamson 1980), as well as under severe environmental









conditions (Held 1983, Khan et al. 1986). Tree seedlings of

northeastern deciduous forests survive long periods of sup-

pression in unfavorable understory conditions partially by

replacing dead stem apices with sprouts (Powell and Tryon

1979). Above-ground portions of shrubs and trees can re-

cover following fire damage by the production of stump

sprouts and epicormic branching (e.g., Ewel and Mitsch 1978,

Ohmann and Grigal 1981, Malanson and Westman 1985). Final-

ly, American chestnut maintains individuals following devas-

tation by the chestnut blight by the persistent production

of sprouts (Paillet 1984).



Sprouting in Stressful Environments

Sprouting frequency is often high in plant communities

subjected to natural stress-inducing disturbances of various

intensities, frequencies, and forms (Bazzaz 1983). Physio-

logical stress is imposed on plants subjected to wind (Putz

and Brokaw 1989), extreme temperatures (Held 1983), grazing,

fungal or insect infestations (Paillet 1984, Maschinski and

Whitham 1989), logging (Terwilliger and Ewel 1986), unstable

soils (Mooney and Billings 1961), increased exposure from

community fragmentation (Trimble and Smith 1970), drought,

and excess water (Johnson 1987). These environmental condi-

tions can cause irreversible alterations in physiological

processes (Johnson 1987); in such cases, sprout production

can be considered a recovery mechanism.







9

Many references to vegetative responses to disturbances

pertain to fire-maintained communities. For example, Keeley

(1981) described the relative importance of sprout produc-

tion versus seedling establishment for chaparral shrub

populations. Sprout production is favored by frequent fires

but seedling establishment becomes more important under fire

regimes with longer fire return intervals.

Over 90% of the groundcover species in longleaf pine-

wiregrass savannas of the southeastern coastal plain are

capable of vegetative regeneration (Davis unpublished data).

These communities are maintained by frequent low-intensity

fires. Stem production and mortality rates of fire-tolerant

species are determined primarily by fire frequency; seedling

establishment is rare. Consequently, spatial patterns of

fires carried through the groundcover determine the extent

of even-aged patches of mixed-species stems.

Occurrences of sprout regeneration in wetlands have not

been well documented, although note has been made that

species capable of regenerating vegetatively are favored in

continuously flooded sites (Malecki et al. 1983). Stress

imposed on plants by long periods of inundation should

affect vegetative regeneration.



Stress Induced by Inundation

Inundation is stressful to many plants because of its

effects on aerobic biochemical processes (Ponnamperuma







10

1972). Oxygen diffuses through air 10,000 times faster than

through water; as water displaces air from soil pores,

molecular oxygen availability for respiration is greatly

reduced. If aerobic organisms and organic matter are pre-

sent, molecular oxygen is consumed within hours of inunda-

tion; the changing oxidation state consequently alters

respiration pathways and mineral balance.

Excess water becomes physiologically stressful to

plants for three reasons: anoxic conditions develop, toxins

accumulate, and nutrient availability is altered (Kozlowski

1984b). Many flood-tolerant species are able to tolerate

only short-term anoxia because energy production is limited

by anaerobic respiration. Plant tolerance to inundation

often depends on an ability to adapt metabolically and to

aerate the rhizosphere (Hook 1984).

As diffusion rate of oxygen in water is slow, diffusion

of toxic metabolic products, such as carbon dioxide and

ethylene, is also limited. In addition, increased con-

centrations of iron and manganese with reduced oxidation

state result in phytotoxic concentrations. Again, the

capacity of the plant to aerate the root system, thus oxidi-

zing many toxins, is an important flood-tolerance mechanism

(Hook 1984).

Stomatal closure is often the initial response to root

system inundation; consequently, transpiration and photosyn-

thesis rates are reduced (Kozlowski 1984a). If inundation






11

continues additional stress symptoms, such as leaf epinasty,

chlorosis, and root mortality, develop in relatively flood-

intolerant species; ultimately, flood-intolerant species die

(Gill 1970).

Flood-tolerant species, particularly woody angiosperms,

develop physiological, morphological, and anatomical adap-

tations to increasing levels of stress due to anoxic soils

(Gill 1970, Hook 1984). Short-term flooding induces revers-

ible alterations of metabolic processes, such as glycolosis

and growth rates, whereas long-term flooding is more likely

to induce irreversible morphological and anatomical changes.

Anatomical and morphological adaptations of flooded

plants primarily increase aeration of flooded roots. Devel-

opment of aerenchyma, lenticel formation and hypertrophy,

and pneumatophore production form an aeration system

throughout the flooded portion of the plant. Although roots

are often lost following inundation, flood-tolerant species

regenerate new roots on the original root system, on the

submerged part of the stem, or on both (Kozlowski 1984c).

Intensity of anoxia-induced stress is related to dura-

tion, depth, and frequency of inundation. Root mortality

rates of artificially impounded floodplain trees increase

with depth of inundation due to corresponding increases in

anoxia of the inundated soils (Harms et al. 1980). Due to a

mixing effect that carries oxygen throughout the water






12

column, however, flowing water is less detrimental to plants

than standing water (Kozlowski 1984c).

Depth, duration, and frequency of inundation differ

among wetlands. Variations in water sources (i.e., runoff,

groundwater, and throughfall) and water loss (i.e., outflow,

evapotranspiration, and percolation) determine the site-

specific hydrologic regime (LaBaugh 1986), and consequently

the intensity of anoxia-induced stress. For example, cy-

press domes hydrologicallyy isolated wetlands common in the

southeastern coastal plain) receive little runoff from

surrounding drainage basins and do not drain to communities

at lower elevations; their hydrologic regimes are charac-

terized by slow seasonal fluctuations (Heimburg 1984).

Rivers, in contrast, receive a large proportion of water by

runoff from upstream sources in addition to base water flow,

which flows rapidly through and out of the system; as a

consequence, floodplain forests are typically inundated

frequently for relatively short durations (Wharton et al.

1982). Greater degrees of anoxia are expected in cypress

dome substrates than in floodplains.



Mechanisms of Apical Dominance

Initiation of growth of lateral buds follows release

from apical dominance (Rubinstein and Nagao 1976, McIntyre

1987, see Zimmerman and Brown 1980). Removal of apical buds

or girdling stems below apical buds often results in lateral









bud growth. The most commonly noted condition that stimu-

lates sprout production is direct injury or removal of some

portion of the plant. Coppice growth occurs in response to

cutting (Hook et al. 1970, McGee 1975), mechanical damage

(e.g., Putz et al. 1983, Gartner 1990), and fire (Malanson

and Westman 1985), as well as insect and fungal infestation

(Paillet 1984, van der Meijden et al. 1988). Root sprout

production is thought to be stimulated by abrasion or freez-

ing and thawing (Jones and Raynal 1987).

The mechanism of release from apical dominance is often

ascribed to removal of the source of auxin, a class of plant

growth-promoting hormones, and a reduction in the auxin-to-

cytokinin concentration ratio at the lateral bud; the initi-

ation of lateral bud growth can be stopped by external

applications of auxin (Rubinstein and Nagao 1976, Powell

1987). This mechanism explains the phenomenon of sprout

production under a variety of environmental conditions such

as fire or insect infestation, but not necessarily in plants

stressed by inundation. Whereas mechanically injured trees

in wetlands produce sprouts, there is no visual evidence,

such as canopy stress or suppression, to explain the initia-

tion of sprout growth by the majority of trees with normal

water level fluctuations. Furthermore, stem concentrations

of auxins may increase following inundation while cytokinin

production is reduced (Reid and Bradford 1984), conditions

increasing the auxin-to-cytokinin ratio that reinforce









rather than reduce apical dominance (Salisbury and Ross

1985).

McIntyre (1987) suggested an alternative hypothesis to

explain the mechanism of apical dominance: growth of lateral

buds is initiated when sink strength of lateral buds for

stem water increases relative to apical bud strength.

Lateral buds of pea and sunflower seedlings were induced to

grow under high light and high nutrient conditions when the

relative humidity was increased to near 100%. Sink strength

of the apical stem for stem water was reduced relative to

the sink strength of the lateral buds as transpiration rates

of the seedlings were reduced, but xylem water potentials

remained high. This mechanism does not depend upon plant

hormone concentrations for initiation of lateral bud growth.

(Auxin concentrations, however, are presumably important for

continued growth of the sprout through their effect on cell

differentiation and strengthening of the vascular connection

between the developing sprout and parent stem.)

Wetland tree species differ in their tolerance to

inundation (Hosner 1960, Gill 1970). Survival following

inundation depends upon how quickly the tree can resume

normal function. Stomatal closure, a common initial re-

sponse to flooding, interrupts normal rates of both photo-

synthesis and transpiration (Kozlowski 1984a). For example,

cherrybark oak (Quercus falcata var. pagodaefolia Ell.) is

flood-intolerant as evidenced by closed stomates for weeks






15

following inundation (Pezeshki and Chambers 1985b). Pereira

and Kozlowski (1977) showed, however, that stem water poten-

tials of inundated wetland trees remain high during periods

of stomatal closure.

Reduced transpiration demand by apical portions of

stems and available stem water below the apex are suitable

conditions for release of lateral buds from apical domin-

ance. Under these conditions, duration of the period of

stomatal closure following inundation should be related to

the propensity of that species to produce sprouts in natural

wetlands. Lateral buds of those species with rapidly open-

ing stomata following inundation are least likely to have

time for lateral bud release (i.e., initiation of sprout

growth) before normal transpiration rates resume. Converse-

ly, species with slowly opening stomata may have adequate

time for lateral buds to initiate growth and increase their

sink strength for stem water before the parent, stem resumes

normal functions.


Research Objectives

Reproduction is limited in disturbance-maintained plant

communities by periodic stressful or destructive events,

such as fire, grazing, or drought. Seedlings are par-

ticularly susceptible to stress and damage, especially

during early establishment phases when energy reserves are

inadequate for recovery. Sprouting is common in these







16

communities; sprouts are formed rapidly by parent plants and

benefit from parent resources for rapid growth.

Relative proportions of stems established as sprouts or

seedlings have been related to disturbance regimes (Keeley

1981). Frequent low-intensity disturbance events often

favor sprout production; seedlings are unable to become

established, yet recovery time for sprouting individuals is

sufficient to replenish energy reserves before a subsequent

stressful event.

Inundation of wetland communities is a periodically

occurring stressful phenomenon analogous to other distur-

bance regimes. Whereas, for example, fire regimes are

described in terms of intensity and frequency, degrees of

anoxia-induced stress depend on duration, depth, and fre-

quency of inundation. Effects of fire on sprouting in-

dividuals are clearly definable as the removal of stems

followed by a pulse of high nutrient availability; interac-

tions of inundation and mechanisms of sprout production,

however, are not so clear.

The research addressed in this report falls into three

phases. The first two are descriptions of sprout distribu-

tions in the chronically stressed wetland forest of Lake

Oklawaha (Chapter 2) and in natural forested wetlands with

different hydrologic regimes (Chapter 3). Sprout production

patterns in these wetlands were related to several environ-

mental parameters estimating degree of anoxia-induced









stress. Interpretation of environmental correlates with

sprout production, however, required a third research phase:

experimental investigations of interactions of environmental

conditions with mechanisms of apical dominance (Chapter 4).

Sprout production by floodplain tree species apparently

increased following the creation of Lake Oklawaha in 1968 by

artificially impounding a portion of the Oklawaha River.

Parts of the floodplain forest have been inundated continu-

ously for over 18 years; depth of inundation varies with

location in the lake. Stress levels experienced by this

forest exceed levels prior to impoundment due to the dura-

tion and, in most areas, depth of inundation. Analyses were

made of sprouting production and proportions of stems of

sprout origin with increasing depth of water.

Twelve natural wetlands were selected to represent the

range of hydrologic conditions in swamps of north-central

Florida. Several tree species commonly occur in three or

more of these wetlands; these populations experience dif-

ferent hydrologic regimes. Sprout patterns resulting from

species-specific characteristics and site-specific environ-

mental conditions were identified at both species and com-

munity levels.

Comparisons of sprout distributions in Lake Oklawaha

with those in natural wetlands isolated effects of site-

specific selection pressures and stress-tolerance threshold

levels. Wetland populations undergo site-specific selection









for flood tolerance, as well as other characteristics.

Consequently, patterns of sprout production in natural

wetlands may not be directly related to flood tolerance but

only correlated with those characteristics that increase the

flood tolerance of an individual. Trees of the Oklawaha

River floodplain forest developed under the same hydrologic

conditions; correlations of sprouting with altered hydrolog-

ic conditions are therefore independent of site-specific

selection pressures and serve as controls for correlations

found elsewhere.

Relationships between sprouts and hydrologic parameters

are not expected to be linear across all stress levels; for

example, there is likely to be a stress threshold beyond

which neither sexual nor vegetative reproduction is success-

ful. Trees in Lake Oklawaha were flooded to depths greater

than are found where trees grow in natural wetlands. Upper

tolerance limits of vegetative reproduction to anoxia-

induced stress are clearer in Lake Oklawaha than in natural

wetlands.

Correlations of sprout distributions with environmental

conditions indicate processes by which the environment

interacts with mechanisms of sprout production; correla-

tions, however, can only suggest mechanisms by which a

process leads to a pattern. For example, increased per-

centages of stems of sprout origin with inundation depth may

be due to lower rates of seedling establishment in deep







19

water, a stimulating effect on sprout production, or both.

The final experimental phase is designed to elucidate

interactions of inundation with mechanisms of apical domin-

ance and sprout initiation.













CHAPTER 2
RECOVERY BY SPROUT PRODUCTION OF THE IMPOUNDED
FLOODPLAIN FOREST IN LAKE OKLAWAHA, FLORIDA



Introduction

Impoundment of natural waterways creates chronic levels

of stress (sensu Levitt 1972) in floodplain ecosystems.

Reservoirs reduce natural water level fluctuations while

increasing the depth and area of flooding in floodplain

forests. The fate of the flooded forest depends on the

perseverance of flood tolerant individuals (see Gill 1970).

Bald cypress (Taxodium distichum (L.) Rich.), swamp tupelo

(Nyssa sylvatica var. biflora (Walt.) Sarg.) and water

tupelo (Nyssa aquatica L.) are examples of species generally

able to tolerate continuous inundation following impoundment

(Eggler and Moore 1961, Harms 1973). Many other species,

which may be dominants in pre-impoundment forests, often die

within a few years, changing the species composition and

physiognomy of the forest (Conner et al. 1981).

Furthermore, standing water limits regeneration of many

species; seedling establishment is usually precluded by

constant deep flooding (Green 1947, Eggler and Moore 1961).

In areas where the forest floor is rarely or never exposed






21

above water, sprouting is the only means of regeneration for

trees (Eggler and Moore 1961, Malecki et al. 1983).

In 1968, the U.S. Army Corps of Engineers dammed a

portion of the Oklawaha River in central Florida. A Federal

Interagency Task Force assessed the environmental impacts of

the impoundment in 1972 and 1975, 3 and 6 yr following

impoundment (Gardner et al. 1972, U.S.D.A. Forest Service

1972, and Harms et al. 1980). I participated in the forest

census that was repeated in 1987, 18 growing seasons follow-

ing impoundment of the Oklawaha River. Objectives of the

study were to describe the status of the forest after 18

years of inundation, to compare the forest structure with

predictions from the earlier studies, and to characterize

the strategies or patterns by which tree populations have

persisted under continuously stressful conditions.


Site Description

The Oklawaha River (290 30' N, 810 50' W) flows north

along the western border of the Ocala National Forest in

central Florida before emptying into the St. Johns River

(Figure 2.1). The climate is characterized by long, warm

growing seasons (about 300 d, mean annual temperature of

21.40C and fewer than 10 d/yr at or below freezing; Lugo and

Brown 1984). Total annual rainfall is about 1300 mm (Na-

tional Oceanographic and Atmospheric Administration 1986),

typically with wet summers and winters. The lowest water

















ORANGE SPRINGS
LANDING Ja


>200


Ocala


National


STUDY
AREA


Forest


3 SWAMP FOREST


0 1 2 3 miles
0 4 kms.
0 1 2 3 4 kms.


26
(control)


Figure 2.1. Detail of the Lake Oklawaha, Florida,
study site. Lines across floodplain delineate sample
populations. Numbers for each sample population indicate
average water depth (cm). Note: direction of river flow is
to the north (redrawn from Harms et al. 1980).







23

levels of the river are in May; peak water levels in Septem-

ber often inundate floodplain forests (Lugo 1972).

Two dams, the Eureka Dam and the Rodman Dam, were

constructed on the Oklawaha River (Figure 2.1). The Eureka

Dam is approximately 26 km upstream from the Rodman Dam and

has never been closed; water flows north around it in the

natural channel. Floodplain forests immediately adjacent to

and on the upstream side of the Rodman Dam were removed

before the dam was closed in 1968. The impoundment re-

sulted in a 5265-ha reservoir, flooding most of the remain-

ing 1620 ha of floodplain forest between the two dams.

Depths of inundation depend primarily on lake levels

maintained at Rodman Dam and decrease with distance upstream

(Figure 2.1). Lake level was initially held near 6.2 m

above sea level for 1 year (Figure 2.2). This level cor-

responds to 2 m above average water level near the remaining

floodplain forest in the deepest water near Orange Springs

(Figure 2.1). At this time, all of the floodplain forests

downstream from Eureka Dam were flooded. Subsequently, the

lake level was lowered to 5.5 m above sea level, where it

has been maintained except for occasional short-term draw-

downs (Figure 2.2). Water depth averages 0.2 m in portions

of the forest farthest from the Rodman Dam below the Eureka

Dam. This is essentially natural river level. Depth of

standing water in impounded forests increases to 1.2 m in









00

0 0
(a 4



0V
0
.9-I $4
M440
4.)

02
ro as
LI'






$4c
0
00


$40
w 00

00
.rq









O4
'do











0 r.
to -rOl
rcl in


-d cZ








4JQ
O










0 -0

4J (d
0O4
000~








'u-I -r4

a)0 0
IJd4

-0

WN 00



$4
*H00
$40
iai



rZx $4


00
















0O













-I


:3







0I
































N 14 -- 0


Cn 0oo -- CO tu O 3

(73A31 V3S 3AOaV I) NOIIVA3-13









the most deeply flooded portion of the study area (Figure

2.1, Harms et al. 1980).



Methods

The Federal Interagency Task Force censused the flood-

plain forest in the spring and fall of both 1972 and 1975

(Harms et al. 1980). In May and June, 1987, the forest was

recensused using the stratified random sampling design of

the previous censuses. Seven contiguous populations differ-

ing in average water depth were delineated along the length

of the impoundment (Figure 2.1). All the populations were

censused except the open water reservoir adjacent to Rodman

Dam, where no live trees remained. Two control populations

were delineated in the natural floodplain forest, one up-

stream from Eureka Dam and one downstream from Rodman Dam.

Populations were sampled with randomly located 0.04 ha

circular plots, the locations of which were noted on aerial

photographs (Gardner et al. 1972). Ten plots were sampled

per population (except 8 plots at 83 cm; 18 plots at 103 cm;

4 at 120 cm; see Figure 2.1) for a total of 80 plots. Stems

> 2 cm diameter at breast height (DBH) were measured for

DBH, identified to species, and classified as live or dead.

In 1987, the previous locations of plots were approx-

imated from the annotated aerial photographs. In addition

to the above mentioned data, we noted the origin of each

stem (seedling or sprout) and if the stem was producing









sprouts. A sprout was defined as the result of active

growth from a bud below breast height. A parent stem can

release more than one sprout; several sprouts on the same

individual often survive and grow to large diameters. Stems

were considered to be sprouts if they were attached at their

bases with no evidence of stems of the same tree having

grown together (e.g., discontinuities or seams in the bark).

Densities of live trees for the different census dates

are contrasted in a repeated measures multivariate analysis

of variance (GLM: SAS 1985). The lack of replication in the

research design does not meet the basic assumptions of the

statistical analysis; therefore, significance levels are

presented only as indications of differences in tree den-

sities between years for Lake Oklawaha.


Results

Species Distributions of the Natural Floodplain Forest

The Oklawaha floodplain forests, under natural condi-

tions, are dominated by red maple (Acer rubrum L.), bald

cypress, swamp tupelo, and two ash species (Fraxinus caro-

liniana Mill. and F. profunda (Bush) Bush; Table 2.1), which

form a closed canopy 20-30 m high. Dahoon holly (Ilex

cassine L.) and stiffcornel dogwood (Cornus foemina Mill.)

are frequent sub-dominants. Shrubs (e.g., Itea virginica

L., Cephalanthus occidentalis L.) are generally clumped on

raised microsites. Herbaceous groundcover is sparse.








28

Table 2.1. Importance values of tree species in natural
floodplain forests of the Oklawaha River (control) and in
Lake Oklawaha forests differing in depth of water. An
indication of the frequency of stems of sprout origin is
indicated for each species. Censused in May, 1987.


SPECIES COMMON NAME SPROUTS


Taxodium distichum

Nyssa sylvatica
var. biflora

Sabal palmetto

Fraxinus spp.

Acer rubrum

Ulmus americana
var. floridana

Ilex cassine

Cornus foemina


Ouercus laurifolia

Persea palustrus

Magnolia virginiana

Crataecus spp.


cypress

swamp tupelo


cabbage palmetto

ash

maple

American elm


dahoon holly

stillcornel
dogwood

laurel oak

swampbay

sweetbay

hawthorn


rare

rare


none

many

many

few


many

few


few

few

few

few


*(relative dominance + relative frequency + relative
density) / 3


















WATER LEVEL (cm)


26 20 52 66 83 103 120
(control)


IMPORTANCE VALUE*


13.5

7.3


3.9

44.7

14.2

5.8


5.4

0.8


1.6

0.3

2.5

0.3


18.8

3.7


7.4

40.5

12.3

6.6


7.6

1.7


1.0


15.6

7.0


4.3

47.1

13.2

2.7


7.6

2.5


23.2

6.3


4.9

44.8

12.7

2.2


32.6

10.6


0.0

49.4

6.1

1.3


62.0

24.2


1.9

8.4

3.7


67.5

17.5


15.0


3.2

2.7









Stem Mortality in Lake Oklawaha

Following impoundment of the floodplain forest, stem

mortality rates increased with depth of inundation in a size

(DBH) dependent fashion. Stem size-class distributions

gradually changed with increasing water depth from large

percentages in small size classes to more evenly distributed

numbers of stems in all size classes (Figure 2.3). The fact

that similar total numbers of trees were standing in water

greater than 60 cm deep from 1972 to 1975 suggests that most

tree mortality due to inundation had occurred between 1968

and 1972 (Table 2.2).

Species Distributions in Lake Oklawaha

Tree species richness declined with increasing water

depth in the impoundment (Table 2.1). Most species with low

importance values in the natural forest became increasingly

scarce in deeper water (e.g., sweetbay, dogwood, and elm);

swamp tupelo and sabal palm, however, are important excep-

tions (see below).

Mortality rates of all species increased with depth of

inundation; however, tree species differed in tolerance to

standing water (Harms et al. 1980). Cypress, swamp tupelo

and sabal palm were the only species that tolerated the

deepest water (120 cm). Contrary to evidence from 1972 and

1975 that ash and maple would eventually die out in inundat-

ed portions of the impoundment, both species have maintained

individuals in standing water 100 cm deep. Moreover, ash




























Figure 2.3. Stem diameter size class distributions of
trees with increasing depth of inundation and in control
populations of Lake Oklawaha at three census dates. Years
are given for census dates 3, 6 and 18 years following
impoundment, respectively. Stem origin as a sprout (cross-
hatch) or seedling (diagonal) was determined only in the
last census.







Water
depth
(cm)

26
(control)




20



C-
0
0
52 o

x

e--
/U)
C
U,
66

E

v)


83






103


1972













2-
1 -I













1-

0
2

1

o
2


1






0r-

2- 14- 27- 40+
13 26 40


1975









































2- 14- 27- 40+
13 26 40


32
1987









































2- 14- 27- 40+
13 26 40


Size class (cm)








Table 2.2. Mean (SD) stem densities (stems/ha) of dominant
tree species in natural floodplain forests of the Oklawaha
River (control) and in Lake Oklawaha forests differing in
depth of water. Significant differences between adjacent
years of censuses are indicated by asterisk between columns
(P < .05).


WATER CYPRESS ASH
DEPTH
(cm)

1972 1975 1987 1972 1975 1987


26 120 106 114 1115 1108 1108
control (177) (185) (64) (396) (384) (697)

20 108 193 408 920 988 1298
(70) (159) (295) (469) (478) (324)

52 63 100 190 763 1110 1550
(56) (115) (226) (270) (462) (397)

66 78 83 185 428 530 1145
(52) (60) (137) (212) (165) (364)

83 206 219 159 331 494 778
(174) (174) (137) (108) (214) (569)

103 242 210 223 46 20 30
(152) (153) (254) (72) (31) (60)
















MAPLE TUPELO TOTAL




1972 1975 1987 1972 1975 1987 1972 1975 1987


256 270 279 80 75 49 1902 1935 1771
(203) (214) (250) (104) (96) (45) (646) (646) (764)

180 295 425 73 65 13 1565 1775 2508
(115) (334) (317) (95) (82) (13) (466) (440) (858)

150 20 320 25 15 30 1203 1750 2345
(87) (125) (273) (33) (34) (39) (248) (424) (570)

240 188 243 53 65 43 878 990 1723
(177) (162) (306) (104) (88) (58) (234) (195) (385)

91 63 34 69 25 31 709 860 1006
(104) (75) (50) (48) (27) (29) (177) (136) (503)

50 3 10 38 110 58 421 343 325
(57) (8) (32) (44) (173) (68) (265) (214) (30)






35

has maintained relative importance values in water depths up

to about 80 cm similar to importance values in natural

forests through the production of sprouts (Table 2.1, Figure

2.4). Several subdominant species (American elm, dahoon

holly, and dogwood) not usually found growing in standing

water are also able to maintain importance values in inun-

dated areas similar to natural areas by sprouting (Table

2.1).

Recovery Following Impoundment

Although some trees continued to succumb under con-

tinuous deep inundation, recovery of stem density in moder-

ately deep portions of the flooded forest was apparent by

1975. For example, stem densities between 2 and 13 cm DBH

in water 83 cm deep increased from 272 stems/ha (SD=125;

n=8) in 1972 to 431 stems/ha (SD=133; n=8) in 1975.

Overall, total live stem densities increased signifi-

cantly from 1975 to 1987 in water depths up to 66 cm (Table

2.2). Numerous cypress and red maple seedlings were found

in the shallowest areas of the impoundment adjacent to the

Eureka Dam, which have had essentially natural river condi-

tions since the second year of impoundment (Figure 2.2; see

Johnson 1972 for details). The increases in stem densities

by 1987 and the recovery of natural stem densities in stand-

ing water greater than 20 cm deep, however, were due to the

sprouts produced by surviving individuals (Figure 2.3).

Furthermore, the proportions of 2-13 cm DBH sprouts





























Figure 2.4. Stem diameter size class distributions of
ash and cypress trees with increasing depth of inundation
and in control populations of Lake Oklawaha after 18 years
of artificial impoundment. Stem origin indicated for sprouts
(cross-hatch) and seedlings (diagonal).





Water 37
depth Ash Cypress
10
(cm)

26 5 -
(control)
0 1 I I I-1 1
10

20 5



a 10
S 5
52 o

I 0
-e 10

66 5 <

(U 0 i 0 i
10

83 5-


10
10 -




0
103 5 -

0 I I I -T 7 t
2- 14- 27- 40+ 2- 14- 27- 40+
13 26 40 13 26 40

Size class (cm)









increased with depth of inundation from 29% in natural

populations up to 67% in 83 cm of water (Figure 2.3).

No trends were observed between increasing water depths

and the proportions of stems that originated as sprouts in

larger stem size classes. This may be due to short life

spans of sprouts, or more likely there was inadequate time

for sprouts to reach diameters greater than 13 cm following

inundation.

The recovery of natural stem densities in the more

deeply flooded areas of Lake Oklawaha was primarily a result

of sprouts produced by ash (Table 2.2). In fact, ash stem

densities increased in water depths between 66 and 83 cm,

primarily due to sprouts, while the other dominant tree

species at best only maintained densities similar to those

measured in 1975. Although red maple stem densities did not

increase significantly, high stem densities have been main-

tained by sprouts for 18 yr in greater water depths than

where red maple is naturally found.

Ash and red maple tree stems that originated as sprouts

are common in natural portions of this floodplain forest.

One-third of stems from 2 to 13 cm DBH in control popula-

tions are sprouts from bases of both live and dead parent

stems. Proportions of stems of sprout origin in larger size

classes tend to be underestimated, because diameter growth

of a stem can mask evidence of sprouting from a parent stem

especially if only the sprout remains. Conservative esti-








mates of the proportions of stems of sprout origin in the

larger size classes in natural populations are 2.3% for 14-

26 cm DBH; 0.4% for 27-40 cm DBH; and less than 0.1% for

greater than 40 cm DBH. Although we did not find a large

percentage of sprouts in large size classes, these data

indicate that stems of sprout origin can be long-lived and

contribute to the maintenance of tree populations under

natural conditions in these forests.



Discussion

Based on censuses 3 and 7 years following impoundment

of the Oklawaha River floodplain forest, Harms et al. (1980)

predicted that bald cypress and swamp tupelo would "adapt to

and be little affected by water depths of 0.6 m or less but

the less tolerant ash and maple [would] eventually die out."

In fact, after 18 years of impoundment, species richness has

remained high in Lake Oklawaha. In addition to dominant

tree species, many subdominant species have maintained

importance values very similar to those in the adjacent

natural floodplain forests in water depths up to 0.6 m.

The capacity for the tree species of the Oklawaha River

floodplain to survive continuous inundation for 18 years was

not evident from their distributions in natural systems.

Bald cypress and swamp tupelo are commonly found along

natural lake fringes, standing in water for long periods of







40

time. Red maple, ash, and the subdominant species, however,

are usually restricted to seasonally inundated sites.

Capacities of wetland species to adapt physiologically

and morphologically to inundation generally reflect patterns

of flood tolerance (Gill 1970, Kozlowski 1984c). After 3

years of impoundment, Harms et al. (1980) found increased

root mortality with increasing water depth. Cypress and

swamp tupelo, however, had begun to develop new secondary

root systems, whereas maple and ash had not.

In spite of high mortality rates of ash and maple

following impoundment and other evidence that these species

were relatively intolerant of long periods of inundation, a

substantial number of individuals of these species, as well

as other relatively flood-intolerant species, have persisted

in Lake Oklawaha for 18 years. The mechanisms by which

these individuals adapted to and tolerated continuous inun-

dation are not clear. Intraspecific variation in physio-

logical adjustments to the stress of inundation were likely

to be involved (Keeley 1979, Sena Gomes and Kozlowski 1988,

Chapter 4).

The trees in Lake Oklawaha had two maintenance patterns

that allowed them to persist under the conditions of con-

tinuous and chronic stress imposed on the floodplain forest

by the impoundment. Ash, red maple and many sub-dominant

species effectively responded to artificial impoundment by

regenerating vegetatively. Cypress and swamp tupelo toler-








ated continuously inundated conditions in this system, as

well as other impounded areas, as long-lived stems, with no

sexual or asexual regeneration (Eggler and Moore 1961,

Conner et al. 1981).

The contrast between these two patterns is evident in

the size class distributions of ash and cypress with in-

creasing water depths (Figure 2.4). The longevities of

cypress and swamp tupelo under continuous flooding determine

the ability of these populations to recover by seeding in

the event that water levels are lowered. The futures of ash

and maple populations rely on the production of new sprouts

to rejuvenate existing individuals. Both maintenance stra-

tegies are effective in sustaining populations capable of

returning to natural regeneration patterns if water levels

are lowered in the future.

Species with the ability to sprout are favored in

stressful environments, particularly in many fire maintained

communities (Abrahamson 1980, Malanson and Westman 1985),

because sprouting is a regenerative mechanism in recovery

from injury. Tree crowns snapped off by wind often resprout

(Putz and Brokaw 1989). "Sprout forests" occur in many

areas of the United States; these are primarily hardwood

forests recovered from clearcutting in the early part of

this century (Braun 1974). In fact, coppice management is

an ancient hardwood forest management technique in Europe,







42

which is still in common use for the production of fuelwood

(Auclair 1986).

Stress on the impounded floodplain forest in Lake

Oklawaha, however, is much greater in both degree and dura-

tion than stress imposed on individuals by wind, fire, or

logging. Even if continuously flooded plants have the

capacity to adjust metabolically to anoxic conditions,

growth rates never match those plants that are inundated

periodically or by flowing water (Brown 1981). This is

likely to be due to the adverse effects of flooding on

photosynthesis and translocation of carbohydrates (Kozlowski

1984a).

It is because of this difference between degrees of

stress to the individual that sprout production of the

continuously inundated trees in Lake Oklawaha differs in

function from the sprouting responses to a gap opening in a

forest or loss of a stem to fire. Under these latter condi-

tions, sprouting individuals can extend into unoccupied

areas to exploit available resources (e.g., a gap) or simply

restore balance to the root-to-shoot ratio (Kramer and

Kozlowski 1979). In contrast, the release of sprouts under

continuously flooded conditions serves as a survival mechan-

ism that maintains the tree under long-term stress.

Novel and chronic perturbations to communities, such as

infestation by viruses or insects, acid rain, climate

change, and long-term flooding, dramatically shift communi-






43

ties to domination by species able to tolerate the situa-

tion. Vegetative reproduction serves as one mechanism by

which plants can maintain themselves for long periods of

time under chronically stressed conditions, perhaps until

future conditions improve.













CHAPTER 3
VEGETATIVE REGENERATION IN NORTH-CENTRAL
FLORIDA WETLANDS



Introduction

Increases in vegetative regeneration in association

with changing environmental conditions are common in plants

(Bradshaw 1965, Abrahamson 1980). Two mechanisms exist by

which such morphological variability can be expressed by a

genotype (Bradshaw 1965): sprouting can be environmentally

induced, or it can be an expression of a preadaptive trait

that increases genet survival rates under new environmental

conditions. Environmental cues for sprouting usually in-

volve loss of aerial structures, as with, for example,

sprouting responses to fire (Ohmann and Grigal 1981, Malan-

son and Westman 1985), grazing, wind damage (Putz et al.

1983), and logging (Braun 1974, Terwilliger and Ewel 1986).

In contrast to sprouting in response to environmental cues,

American beech (Fagus grandifolia Ehrh.) continuously pro-

duces root suckers throughout its geographical range.

However, stems of root sucker origin were favored in the

most severe climates because they were more tolerant of

freezing early in the growing season than seedlings (Held

1983).








Many types of plant communities experience recurring

disturbances differing in return frequency, intensity of

stress induced, and predictability; these disturbances are

usually integral to regeneration and competition processes.

Longleaf pine forests of the southeastern Coastal Plain, for

example, are considered to be disturbance-maintained commu-

nities with predictably frequent, low-intensity fires. When

fire is excluded, species composition shifts from fire-

tolerant species typical of these savannas toward fire-

intolerant species of mixed-hardwood forests (Monk 1968).

Black grama (Bouteloua eriopoda) dominated grasslands of

southern New Mexico shifted to shrub-dominated communities

when grazing pressure was intensified by domestic livestock

(Schlesinger et al. 1990).

Sprout production, an expression of phenotypic plas-

ticity, is adaptive in fluctuating environmental conditions;

the rapidity of vegetative recovery following a destructive

event gives sprouting plants an initial advantage over

seedling establishment in reoccupying an area (e.g., Keeley

1981). Relative contributions of vegetative and sexual

reproduction to perpetuation of populations often vary with

the frequency, intensity, and seasonality of recurring

stressful environmental conditions. For example, frequent

low-intensity fires favor sprout production of chaparral

shrubs (Keeley 1981), whereas intense summer fires in Pinus

banksiana Lamb. forest communities destroy vegetative repro-







46

ductive structures, such as dormant buds on corms, favoring

seedling establishment (Ohmann and Grigal 1981). Under less

intense disturbances such as logging, sprout production is

generally greater following disturbances that occur during

the dormant season than during the early growing season

(Belinger 1979, Harrington 1984). Seasonal fluctuations in

carbohydrate and nitrogen reserve levels in roots are corre-

lated with shoot growth vigor and may explain these patterns

(Tromp 1983).

Disturbances such as fire, grazing, and logging are

sources of physiological stress for plants. Partial removal

of stems disrupts the root-to-shoot ratio, and consequently,

the physiological state of the plant. If a vegetative

response to stress is adaptive in communities subjected to

such disruptive events, sprouting may also be adaptive under

other less destructive stressful conditions. The observa-

tion that sprouts are more common in wetland than upland

communities (personal observation) suggests that sprout

production by wetland plants may be an adaptation to stress

imposed by water level fluctuations.

North-central Florida wetland hydrologic regimes are

analogous to fire regimes in that both have characteristic

return frequencies and intensities that induce stress in

plants. Depending on the relative position of wetlands in a

drainage basin, depth, duration, and frequency of inundation

will vary (Chapter 1). As fire intensity is determined by








temperature and speed of travel, the degree of anoxia and

intensity of stress imposed on plants in wetlands is deter-

mined by the depth, duration, and frequency of inundation

(Reddy et al. 1980).

Effects of hydrologic regime on relative contributions

of vegetative and sexual reproduction to perpetuation of

wetland tree populations should be evident from comparisons

of population sprouting patterns among different wetlands.

These patterns are likely to be significant for populations,

communities, or both levels. Species-specific interactions

with hydrologic regimes may differ. If, for instance,

anoxia-induced stress affects mechanisms of apical dominance

of red maple in a different manner than sweetbay (Magnolia

virginiana L.), different sprout production patterns of

these species would be expected where they co-occur. If, on

the other hand, interactions of hydrologic regimes with

mechanisms of apical dominance are not species-specific,

consistent sprout production patterns within wetland com-

munities would be expected, but a species' sprout production

pattern might differ from wetland to wetland.

Sprouting patterns of trees in relation to inundation-

related stress were investigated by describing the impor-

tance of sprout production and establishment (stems of

sprout origin reaching at least 2 cm DBH) for common tree

species occurring in wetlands of north-central Florida.

These community and species-specific sprout patterns were








compared with hydrologic parameters to determine if

increased stress induced by inundation increases rate of

sprout production and establishment in natural wetlands.



Site Descriptions

Twelve forested wetlands differing in both hydrologic

regimes and species composition were selected throughout

north-central Florida (Figure 3.1). Study sites were part

of a larger study of common Florida plant communities con-

ducted in cooperation with the Center for Wetlands, Univer-

sity of Florida, funded by the Florida Institute of Phos-

phate Research. All sites were at least partially logged

during the first half of this century. Site selection was

based on the presence of natural hydrologic regime (i.e., no

indication of artificial drainage or inflow), the mature

condition of the forest, and the long-term accessibility

needed to monitor water level fluctuations over several

years.

Seven of the study wetlands were hydrologically iso-

lated from other wetlands (i.e., not connected by overland

flow): two cypress domes (Figures 3.1A and 3.1B), three

bayheads (Figures 3.1C and 3.1D), and two mixed-hardwood

forests (Figures 3.1C and 3.1D). Water levels in these

wetlands depend primarily on rain input and groundwater

fluctuations. Water does not flow perceptibly in these








u Q) rla

4 ---00
G0 X 4 O.P V
rw Ngw 3 0

4k -0 W 4 <-H44


Q -. .) 1 to -4
V > .--o 0

a ~ t ua
WMn 3 I (A 4-4
04- 0 i : -4)
Q,^Qi% 0- 0O
0 > .-- 0 !) M .-,
10 4 4J >4 < >






o # O J "-'( 0
** ) c U 4 ) k

m w--4 0 4
01 O 044 0.,


0 0 -r 0 -
- Q0 0M r
a, ) 0') ( 0 a a
(0 M OM >-H V ()


0U 0) ( 0 (0
a, 01 p >1 0 0r1
0 4k N0 k **H






4-- M 0 0 .4) P
w '-) m C g0H 00
g 4C Ul 044
Wa as M toa 0C


-4 0 -4 -t M



Sr-- O. ( M U



(UP O ** O
d 4) 0 Cm >4 O)
) C( r d 4
M 0 (dE A V *









P 0 0 V W 0U V
*C0 L C M Mr O
* C- 4 0 r- 9C
(0 0 ri tO C 0





0 M V (0 0 -H
4J -0 r 0 -t >
SU :3 fr-I V 4
Z 0 -. V49 0 W





V(U E-4 P4 4 0 0
* (d0 En :a o0


0 ( U M 0 0(


r -0 0 -0 -VU >C
W 0 VP 0 0 r4l 41 t
.; tA t r-4 P4, (n >1(1)
0 4 M $4 :3-A
>40 l0 l000 04J> 0l
a w==X a a wo: E











50









Cu

0


ot
U)


















04

D..
m0
b*r













clit























ClQ)
II








IKI



a0 1 I 0








































Figure 3.1--continued.


























SOST1


*OST2


) SMT2


) SMT1


0 2 4 6
kilometers


Figure 3.1--continued.













Park



LotelBa


Lake
Letta


0 2 4

kilometers


Dinner
Lake


bring


HHP4


Figure 3.1--continued.


6
g


g










54








0
'aa)
E
U






ca
1N14
0


















c Pi
o 0


























too




w





csju








00


0 I





rz

























































kilometers


Figure 3.1--continued.


















0 2 4 6
kilometers


0


Figure 3.1--continued.























































0 2 4 6
kilometers


(H)


Figure 3.1--continued.







58

wetlands; water loss is through percolation and evapotrans-

piration. Consequently, water level fluctuations are rela-

tively slow and there is often standing water in the summer

and winter.

The remaining five wetlands were parts of overland

drainage systems: a cypress strand (Figure 3.1E), a lake

fringe forest (Figure 3.1F), a slough (Figure 3.1C), a

floodplain forest (Figure 3.1G), and a forest along a

groundwater fed stream (Figure 3.1H). These wetlands re-

ceive rain, groundwater, and overland flow in varying pro-

portions. Water level fluctuations are usually seasonal, as

in the isolated wetlands, but the rise and fall are more

rapid and dramatic. Water flows through these systems, and,

therefore, conditions associated with standing water (e.g.,

low nutrient availability, low reduction-oxidation poten-

tial) are less intense then in the hydrologically isolated

wetlands.



Methods

Transect Establishment

Line transects were established across each wetland.

Transects were surveyed from a random point in an upland

association and were oriented perpendicular to the wetland

boundary. Transects were divided into the component plant

communities, generally uplands, wetland, and ecotones (areas

of transition between uplands and wetland). This study used








information only from wetland portions of each transect as

determined by ground surface profile and wetland plant

distribution.

Delineation of forest community type followed Monk

(1968) with the exception that mixed hardwood swamps in this

study were further divided by hydrologic setting: hydro-

logically isolated mixed hardwoods, cypress strand, lake

fringe, slough, floodplain, and groundwater-fed stream

forests. Whereas no species-area relationships were deter-

mined for this study, it was determined that wetlands were

adequately sampled because species composition of the forest

types agreed with Monk (1968).

Ground Surface Profile

The physical profile of each transect was described by

surveying relative elevation changes along each transect

line. Elevation readings were taken every 2 m along the

line except when microtopographic features (i.e., hummocks

and stream channels) required that readings be taken at 0.1

m intervals. Organic soil depth was recorded at 4 m inter-

vals by probing to mineral soil with a 1 cm diameter iron

rod.

Ground Water Fluctuation

Piezometer wells were established at the midpoint or

lowest elevation in each wetland. Wells consisted of 3.2-cm

diameter schedule 40 PVC pipe, tipped with a PVC well

screen. Each well was installed in an augered hole and soil









was backfilled around the pipe. Well depths varied from 1

to 4 m depending on location and proximity of groundwater to

the soil surface. Wells were installed deeply enough to

maintain contact with groundwater during dry periods and

they extended above the ground high enough to remain visible

during inundation. Groundwater levels were measured monthly

from a reference mark of known relative elevation on the

pipe.

Daily water levels were estimated from monthly measure-

ments by non-linear interpolation using a spline technique

(Michael Miller, unpublished BASIC computer program).

Hydrologic regimes were calculated for each wetland using

measurement records from January 1986 to December 1987. For

each elevation in the wetland (0.01 m increments) the fol-

lowing parameters were calculated to define the hydrologic

regimes: average depth of inundation, percentage of time

inundated during the two year period (duration), and number

of times water rose above the elevation during the two year

period (frequency). Hydroperiod calculations for the deep-

est point in the wetland were used for the community des-

cription.

Soil Chemistry

Soil samples were collected from each well location.

Three 5-cm diameter soil cores were taken from the top 10 cm

of the soil profile within 1 m of the well. Samples from

each well were mixed and subsampled for analysis. Analyses









were performed by the Institute of Food and Agricultural

Science (IFAS) Extension Soil Testing Laboratory, University

of Florida, Gainesville, Florida. Soils were analyzed for

pH, organic matter content, and nutrients (NH4, NO3, P, Ca,

Mg, K, Fe, Mn, Zn, Cu, Cl, Na, and Al). Double acid extrac-

tion was used for P, K, Ca, Mg, Mn, Cu, and Zn, and samples

analyzed by inductively coupled argon plasma (ICAP) spec-

troscopy.

Tree Species Distributions

Species and diameter at breast height (DBH) were

recorded for all woody stems > 2 cm DBH within 5 m of the

transect. Euclidean coordinates of each stem were recorded.

For stems situated on hummocks, hummock heights were also

recorded.

Two sprouting conditions were noted for each stem:

stem origin (parent stem or sprout) and the occurrence of a

stem producing sprouts. Stem characterization was made in

the same manner as for the Lake Oklawaha study (see Chapter

2).

Individual tree positions were used to describe ranges

of environmental tolerance of tree species to hydrologic and

soil conditions. Relative elevations and organic soil

depths were interpolated for each tree from the physical

profile. Frequency, depth, and duration of inundation were

then determined for each tree based on measured water levels

at each site.









Data Analysis

Multiple regression analyses of environmental variables

against arcsin square root transformed percentages of stems

of sprout origin and of stems producing sprouts were per-

formed using general linear models (PROC GLM, stepwise

procedure, SAS 1985). Regression equations are reported in

the text with regression coefficients and one standard error

in parentheses. Categorical maximum likelihood analyses

(PROC CATMOD, SAS 1985) regress ln(P / l-P) against indepen-

dent parameters, where P is the probability of an event at

different levels of the independent parameter. Maximum

likelihood analyses were used to investigate the relation-

ships between sprouting conditions of stems (stem origin and

sprout production) to environmental parameters. Analysis

summaries for results presented in tables are in the appen-

dix.



Results

Wetland Tree Species Distributions

Twenty-seven tree species were recorded in the 12

wetlands (Table 3.1). Most species were found in only one

or two wetlands and had low relative importance values (less

than 10). Eight species, however, were dominant or co-

dominant species in several wetlands, and were used in both

community and species level analyses of sprouting patterns.










Table 3.1. Diversity and importance values of trees in
wetlands differing in hydrologic regime. Common species
used in further analyses are denoted with asterisk.


HYDROLOGICALLY
ISOLATED


Cypress
Dome Bayhead

#1 #2 #1 #2 #3

TRANSECT AREA (m2) 1420 1130 1700 2200 1200

SPECIES RICHNESS: 5 8 7 7 4

SHANNON-WEINER
DIVERSITY (H): 0.06 0.18 0.55 0.62 0.30

EVENNESS (H / Hmax): 0.08 0.20 0.65 0.73 0.50

SPECIES IMPORTANCE VALUES
((relative dominance + relative abundance) / 2):

Taxodium spp. 98.4 94.8 23.4 9.7

Ilex cassine 1.0 2.1 14.8 11.9

Nvssa svlvatica 0.1 1.0 4.0 2.1 14.3
var. biflora *

Persea palustris 0.4 0.1 2.1 1.2

Quercus virqiniana 0.1 0.7

Acer rubrum 0.3

Ilex opaca 0.4

Quercus laurifolia 0.5

Quercus nigra 0.3

Magnolia virginiana 3.6 17.1 74.0

Gordonia lasianthus 53.0 55.7 9.9
















HYDROLOGI-
CALLY
ISOLATED


Mixed
Hardwoods


#1 #2


1050 950

10 7


0.71 0.45

0.71 0.53






6.1


27.0


1.7


2.5


41.1



0.3



13.0

4.4


6.4



2.8


HYDROLOGICALLY
CONNECTED


Cypress Lake Flood- Ground-
Strand Fringe Slough plain water


2250


0.54

0.57





53.5

2.8

19.3


0.6



0.2


1650


0.58

0.68





41.0



2.8


3.0



14.3


2000

9


0.78

0.82





12.2

46.7

17.7


2.7



5.4


2600

11


0.76

0.73





9.5

0.7


12.3


7.1


1.7

7.6


12.6

1.7


2300

17


0.87

0.71






4.0




7.4



3.9



2.9

1.9

6.9


1.2









Table 3.1--continued.


HYDROLOGICALLY
ISOLATED



Cypress
Dome Bayhead


#1 #2 #1 #2 #3


Pinus serotina 1.3

Pinus elliottii 1.1

Fraxinus caroliniana *

Liquidambar styraciflua

Sabal palmetto

Ulmus americana
var. floridana

Pinus palustris

Cornus foemina

Carpinus caroliniana

Celtis laevigata

Carva acuatica

Chamaecyparis thyoides

Morus rubra

Pinus taeda

Tilia caroliniana


CUMULATIVE IMPORTANCE VALUES:

dicots 1.6 5.2 75.5 89.0 100.0

conifers and
monocots 98.4 94.8 24.5 11.0 0 0


--~- --~-


.... ---- ---












HYDROLOGI-
CALLY
ISOLATED


Mixed
Hardwoods


#1 #2


2.9


2.4


58.1

0.3 1.5

26.8


2.7


HYDROLOGICALLY
CONNECTED


Cypress Lake Flood- Ground-
Strand Fringe Slough plain water


13.9


32.7


0.6

0.4


42.5


1.0


16.6


5.5


0.2


2.1

2.0

49.7

0.3


4.0


1.7

0.2

2.4

6.0


1.6


14.6

0.2

0.4

0.8


94.7 73.2


5.3 26.8


28.6


59.0


86.8


90.5


35.1


71.4 41.0 13.2


9.5 64.9








Species richness and evenness of trees in hydrologi-

cally connected swamps generally were greater than in hydro-

logically isolated swamps (Table 3.1). Cypress domes and

bayheads had the lowest species richness, with pond cypress

dominant in cypress domes and broad-leaved, evergreen

dicots, such as loblolly bay (Gordonia lasianthus (L.)

Ellis.) and sweetbay, dominant in bayheads. Mixed-hardwood

swamp forests generally had the greatest species richness

and were dominated by one or more deciduous dicot tree

species. Greater tree diversity in wetlands with water flow

may be related to hydrologic conditions more conducive for

seed dispersal; in addition, more variation in hydrologic

conditions of connected wetlands than in isolated wetlands

provides a wider range of establishment conditions.

Sprout Production by Dicot Tree Species

Stems producing basal sprouts (Table 3.2) and stems of

sprout origin (Table 3.3) were found in all wetlands.

Multiple regression analyses indicate that sprouting condi-

tions of trees within wetland communities were not related

to the importance values of the common tree species. The

importance values of dicots versus other species of trees

(i.e., conifers and monocots), however, were positively

related to percentage of stems producing sprouts (Tables 3.1

and 3.2):









Table 3.2. Environmental condition significantly related to
percentages of stems actively producing sprouts in natural
forested wetlands.


HYDROLOGICALLY
ISOLATED


Cypress


Cypress
Dome

#1 #2


Bayhead

#1 #2 #3


Percentage of stems
producing sprouts:


Iron (mg Fe / kg soil):


5.6


7.9


13.8 14.8


28.3 13.6 22.9


58.3 34.9 33.3














HYDROLOGI-
CALLY
ISOLATED


Mixed
Hardwoods


#1 #2





29.1 40.2



51.8 64.3


HYDROLOGICALLY
CONNECTED


Cypress Lake Flood- Ground-
Strand Fringe Slough plain water


3.7


28.1


21.8


23.4 43.8 23.1


28.5


5.5


81.1 1.6









Table 3.3. Environmental conditions significantly related
to percentages of stems of sprout origin in natural forested
wetlands.


HYDROLOGICALLY
ISOLATED


Cypress


Cypress
Dome


#1 #2


Bayhead


#1 #2 #3


SPROUTING CONDITION:

Percentage of stems
of sprout origin


7.2


HYDROLOGIC REGIME:

Average duration of 65.5
inundation
(% of 2 year record)

Average frequency of 2.3
inundation
(# of times in 2 years)

Average depth of 14.2
inundation
(cm)


STEM DISTRIBUTION:

Percentage of stems
on hummocks


0.9


7.1



41.6


3.1


10.1





6.6


40.4 30.7


7.8


13.0 13.0 13.6


1.4


1.3


1.4


0.7


65.7 29.8


3.1


0.7





0.7


SOIL CHARACTERISTICS (mg nutrient / kg

pH 4.1 4.3

Organic matter 3.7 6.4

Zinc (Zn) 1.1 1.5

Copper (Cu) 0.0 0.0

Chlorine (Cl) 4.6 8.8


soil):

3.8

18.6

2.9

5.9

31.6


3.6

26.7

3.0

2.5

15.2


3.7

14.0

1.8

0.0

19.7














HYDROLOGI-
CALLY
ISOLATED


Mixed
Hardwoods


#1 #2





22.4 10.9





41.6 83.2


3.3


2.2


5.9 18.9


40.6





4.5

18.1

4.1

3.2

45.5


1.7





4.1

21.3

3.5

0.0

19.4


HYDROLOGICALLY
CONNECTED


Cypress Lake Flood- Ground-
Strand Fringe Slough plain water


4.4





61.5



3.5



8.0





5.1





3.7

13.7

0.8

0.0

9.5


27.1





47.8



3.4



16.0





0.0





5.6

4.4

1.2

0.0

0.0


22.6





48.1



3.6



8.1





41.7





4.4

26.0

3.1

2.0

26.7


34.5





8.9



7.3



3.1





0.2





6.0

1.3

6.0

0.0

4.4


5.2





5.4



0.1



1.7





0.6





5.6

50.0

0.3

0.0

83.9









percent producing sprouts = 0.222 ( 0.068) + 0.004

( 0.001) dicot importance value


(R2 = 0.57, df = 1, 10, P < .01). Importance values of

dicots in these wetlands were also positively related to

percentage of stems of sprout origin (Tables 3.1 and 3.3):


percent sprout origin = 0.234 ( 0.084) + 0.003

( 0.001) dicot importance value


(R2 = 0.39, df = 1, 10, P = .03).

Distribution of Stems Producing Sprouts

Percentages of stems in wetlands that were producing

sprouts were not related to any combination of environmental

parameters describing the hydrologic regime, topography, or

stem distribution (e.g., stem density, basal area). This

indicates that sprouts are equally likely to be produced

under a wide variety of conditions in wetlands. Depth,

duration, and frequency of inundation apparently did not

stimulate or suppress sprout production. Light penetration

and competition (indexed by stem densities and basal area)

also were not correlated with sprout production.

Iron concentration was the only soil nutrient associat-

ed with the likelihood of basal lateral buds on stems break-

ing dormancy (Table 3.2):








percent producing sprouts = 0.222 ( 0.053) + 0.006

( 0.001) Fe



(R2 = 0.69, df = 1, 10, P < .01). Soil iron concentrations

are associated with buffering of the reduction-oxidation

potential of flooded soils. It accounted for the greatest

amount of variance in proportions of stems producing sprouts

in these wetlands.

Distribution of Stems of Sprout Origin

In contrast to basal sprout production, percentages of

stems of sprout origin in wetland communities are related to

hydrologic regime parameters and percentage of stems situa-

ted on hummocks (Table 3.3):



percent sprout origin = 0.263 ( 0.075) 0.008

( 0.002) duration + 0.041 ( 0.016) frequency +

0.031 ( 0.011) depth + 0.007 ( 0.001) hummock



(R2 = 0.83, df = 4, 7, P < .01). Largest percentages of

stems of sprout origin were in wetlands with relatively high

frequencies of water level fluctuations and deep average

water depths but where sprouts could escape long periods of

inundation on elevated microsites, the hummocks.

In addition to hydrologic regime and hummocks, several

soil characteristics were significantly related to the







74

percentage of stems of sprout origin in wetland communities

(Table 3.3):



percent sprout origin = -0.281 ( 0.099) + 0.132

( 0.022) pH + 0.023 ( 0.010) Zn + 0.084

( 0.010) Cu 0.007 ( 0.001) Cl + 0.007

( 0.002) organic matter


(R2 = 0.96, df = 5,6, P < 0.01). The pH is negatively cor-

related with reduction-oxidation potential, and, thus, high

pH is associated with less reduced soils and better growing

conditions. High organic matter is closely associated with

hummock formation.

Sprouting Conditions of Common Wetland Tree Species

Stems producing sprouts and stems of sprout origin

comprised a substantial portion of the 5154 stems censused

in these 12 study wetlands (Table 3.4). Of the 27 species

recorded, stems of 20 species were actively producing

sprouts, whereas 15 species had stems at least 2 cm DBH that

had originated as sprouts. All of the 8 most common wetland

tree species both produced sprouts and had basal sprouts

grow to at least 2 cm DBH.

The common tree species exhibited wide ranges of sprout

production capacity and percentages of stems of sprout

origin (Table 3.4). For example, production and establish-

ment of sprouts by cypress were relatively infrequent when












Ut
i-4



*4







0
OU
-I V



H*$4
*--1
U











0
0
0)





wM











40
OM
0 0
U0

C4


0 0
0
t0
-l4 $4

.04
O U

O(U

U (



me
o


o ia
r-l




M5
0(U
o


c

e-40



E-i


Q C






MOH
0







0 P
EO







0fl
0 N










SU)




























H

I
0










U)


0V 10 r-4 co
4 T-4l ,-4








- ti n In %D
-i i-i .-


o N IA aO N o
'N 1i (V (* -I 1-








N 14 -In -4 I. L
m< N4 N N











w cn in w w0 %0


0 0
> >0 0


0 ( 0 0 4) 4)
(- O l ) ~
A rl A


0 0 0 ( 3 3> 0
k a r-i w i 0


(t
c a a)
r*H to f
H 4) 0 *4

0- (U -rI 0 1 4
4 (-4 4 *4 4- r-4 u 4
S> > r-i 0t
S. a fa a. 3(
S 0 -4 to r-4

x 5 O | > 0 4)
( p 0 0 ( X t
rL, -r f


75


in co






(N

0 %0
I il


Nr o> T-
M cO N
N 10 %D









compared with ash. In spite of this range of sprouting

behavior, species importance values were not good predictors

of sprouting in wetland communities.

Maximum likelihood analyses relating stem origin (Table

3.5) and sprout production (Table 3.6) of individual tree

species to hydrologic parameters and hummock height revealed

no consistent patterns. Red maple stems of sprout origin,

for instance, were more common in situations with relatively

shallow, more frequent and shorter inundation periods than

stems that were not sprouts. In contrast, species such as

swamp tupelo were more likely to have stems of sprout origin

if the parent stems were on hummocks.

In spite of the suggestion of the maximum likelihood

analyses that each species responded independently to en-

vironmental cues, if one species in a community had a rela-

tively large percentage of stems of sprout origin, other

species were also likely to have relatively more stems of

sprout origin (Table 3.7). For example, in Bayhead #1, most

species had relatively large percentages of stems of sprout

origin, and, consequently, percentages that were high in

comparison with sprouting of the same species in other

communities. Percentages of stems of sprout origin of

individual species can be ranked for every wetland where

they occur. For instance, sweetbay occurred at seven sites

(Table 3.7). Mixed hardwoods #1 had the lowest percentage

of sweetbay stems of sprout origin, and the slough had the









Table 3.5. Mean (SD) hummock height and hydrologic condi-
tions of stems by origin (P parent; S sprout) for common
tree species in study wetlands. Significant maximum likeli-
hood coefficients are indicated with asterisk between means
(P < .05). Note: Summaries of maximum likelihood analyses
are in Appendix A.1.

NUMBER HEIGHT
OF STEMS OF HUMMOCK
(cm)

SPECIES P S P S


Acer rubrum 187 47 4.0 9.4
(14.4) (9.4)

Fraxinus caroliana 233 256 0.0 0.0
(0.0) (0.0)

Gordonia lasianthus 406 215 8.3 8.3
(11.6) (10.5)

Ilex cassine 204 193 7.9 11.0
(10.4) (11.7)

Magnolia virginiana 479 63 2.1 5.2
(6.7) (9.3)

Nyssa sylvatica 380 42 2.0 8.8
var. biflora (6.2) (8.8)

Persea palustris 91 13 3.6 11.9
(7.5) (12.8)

Taxodium spp 1613 59 0.7 0.6
(3.9) (3.2)

















DURATION FREQUENCY DEPTH
OF INUNDATION OF INUNDATION OF INUNDATION
(%) (per 2 yr) (cm)

P S P S P S


31.6 29.9 5.2 3.7 6.5 4.5
(32.6) (31.8) (4.0) (2.7) (7.1) (5.5)

55.6 32.3 3.4 6.5 14.5 10.4
(33.6) (30.4) (3.6) (4.8) (9.4) (10.3)

11.3 14.3 1.5 1.7 0.9 1.2
(18.5) (21.3) (1.9) (1.9) (2.1) (2.1)

32.2 24.6 2.4 2.6 5.0 3.2
(33.0) (28.2) (2.0) (2.3) (7.6) (5.9)

22.8 30.8 2.5 3.3 2.8 4.2
(27.5) (28.4) (2.0) (2.0) (5.3) (6.5)

48.4 23.0 3.6 2.1 6.3 2.8
(29.1) (28.9) (1.8) (2.2) (5.7) (5.6)

18.7 21.9 1.6 2.8 2.8 1.9
(29.3) (23.1) (2.0) (2.3) (6.0) (2.2)

60.3 54.0 3.1 3.1 11.2 10.1
(23.3) (16.8) (1.8) (1.4) (7.8) (4.4)









Table 3.6. Mean (SD) hummock height and hydrologic condi-
tions of sprout production (NP not producing sprouts; P -
producing sprouts) for common tree species in study wet-
lands. Significant maximum likelihood coefficients are
indicated with asterisk between means (P < .05). Note:
Summaries of maximum likelihood analyses are in Appendix
A.2.


NUMBER HEIGHT
OF STEMS OF HUMMOCK
(cm)

SPECIES NP P NP P


Acer rubrum 135 99 7.3 2.0
(17.1) (5.7)

Fraxinus caroliana 336 153 0.0 0.0
(0.0) (0.0)

Gordonia lasianthus 465 156 7.8 9.8
(11.2) (11.2)

Ilex cassine 314 83 9.0 11.2
(11.1) (11.4)

Magnolia virginiana 406 136 2.3 2.9
(7.1) (7.1)

Nyssa sylvatica 403 19 2.6 6.1
var. biflora (6.7) (7.9)

Persea palustris 84 20 4.5 5.3
(8.9) (8.5)

Taxodium spp 1594 78 0.7 0.7
(3.8) (4.4)


















DURATION FREQUENCY DEPTH
OF INUNDATION OF INUNDATION OF INUNDATION
(%) (per 2 yr) (cm)

NP P NP P NP P


35.2 25.8 4.4 5.5 6.5 5.5
(33.9) (29.5) (3.6) (4.1) (7.4) (6.1)

38.6 53.8 5.7 4.4 11.4 14.3
(33.4) (33.1) (4.7) (3.9) (10.2) (9.3)

12.8 11.1 1.6 1.5 1.0 0.9
(19.9) (18.5) (2.0) (1.7) (2.1) (2.1)

28.4 28.7 2.6 2.2 3.8 5.1
(30.6) (32.4) (2.2) (1.9) (6.3) (8.7)

20.1 34.3 2.4 3.2 2.3 4.8
(25.5) (31.0) (2.1) (1.8) (5.0) (6.4)

46.2 37.0 3.5 3.2 6.0 5.1
(29.9) (32.3) (1.9) (1.9) (5.8) (6.8)

18.7 20.7 1.6 2.6 2.8 2.2
(29.4) (25.0) (2.0) (1.9) (6.0) (3.6)

60.7 48.1 3.0 3.8 11.2 9.9
(22.8) (26.1) (1.6) (3.3) (7.8) (6.6)









Table 3.7. Percentages of stems on hummocks and stems of
sprout origin of common dicotyledonous wetland trees species
in study areas. Stem percentages are ranked (numbers in
parentheses) across sites. See text for explanation of
analysis.


HYDROLOGICALLY
ISOLATED


Cypress
Dome


Bayhead


#1 #2


#1 #2


PERCENTAGE OF STEMS
ON HUMMOCKS:


0.0 6.6
(1.5) (8)


65.7 29.8
(12) (9)


COMMON DICOT TREE SPECIES:
Acer rubrum

Fraxinus caroliniana


Gordonia lasianthus


Ilex cassine


14.3 37.5
(1) (4)


38.5 35.4
(7) (5)
64.2 58.9
(8) (7)


Magnolia virginiana

Nyssa sylvatica
var. biflora

Persea palustris


AVERAGE RANK FOR
COMMUNITY


21.7
(6)


0.0
(3)


0.0
(3)


25.0
(7)


3.7
(6)


3.0
(4)


9.6
(4)


66.7 53.3
(10) (9)


0.0 11.8 0.0
(1.5) (4) (1.5)


6.5
(12)


5.8
(9)


0.7
(5)


0.0
(2)


3.2
(1)


9.2
(3)
0.0
(3)


2.1
(1)
















HYDROLOGI-
CALLY
ISOLATED

Mixed
Hardwoods

#1 #2

40.6 1.7
(10) (6)




29.9 10.0
(7) (4)
14.8
(1)


22.2
(2)
41.2
(5)


3.0
(1)


17.9
(8)
28.6
(8)


0.0
(3)


HYDROLOGICALLY
CONNECTED


Cypress Lake Flood- Ground-
Strand Fringe Slough plain water


5.1
(7)




0.0
(2)


0.0
(1.5)




0.0
(2)
50.9
(3)


36.0
(6)
21.7
(3)
14.3
(5)


1.7
(6)


0.0
(3)


20.0
(5)


5.2 2.7
(8) (2.5)


4.5
(7)


2.7
(2.5)


41.7
(11)




35.1
(8)


25.0
(3)
43.8
(6)
23.9
(7)
11.6
(7)
21.1
(6)

6.2
(11)


0.2
(3)




11.8
(5)
75.6
(4)


100.0
(9)


5.4
(3)

6.0
(10)


0.6
(4)




16.7
(6)
25.0
(2)
33.3
(4)
17.4
(2)
3.6
(2)


3.2
(5)









highest percentage; sweetbay had rank 1 at Mixed hardwoods

#1 and rank 7 at the slough. Community averages of these

ranks of relative sprouting success were positively

correlated with ranked percentages of stems that occurred on

hummocks (Kendall Coefficient of Rank Correlation = 0.51, P

< .05, Table 3.7).



Discussion

Basal sprouts of trees in north-central Florida wet-

lands are widely distributed across both species and hydro-

logic regimes. As described for other dicot-dominated

vegetation (see Abrahamson 1980, Davis unpublished data), a

large percentage of these wetland tree species (87%) was

capable of reproducing vegetatively. Forty-four percent of

the wetland tree species, principally the common dicots, had

individuals of sprout origin > 2 cm DBH.

Sprouts are produced throughout a wide hydrologic range

(i.e., depth, duration, frequency) in north-central Florida

wetlands. Whereas stems of sprout origin were generally

more common on hummocks in wetlands with brief but frequent

periods of deep inundation, trees producing sprouts were

found in all hydrologic settings.

These results indicate that, unlike sprouting in re-

sponse to disturbances such as fire, sprout production is

neither in response to nor adversely affected by water level

fluctuations. Establishment of sprouts as large diameter







84

stems depends upon surviving periods of inundation; sprouts

on parent stems situated on hummocks are best able to avoid

the adverse effects of inundation.

Two lines of evidence support the hypothesis that

importance of sprouts in north-central Florida wetlands is

primarily a community rather than a species-specific level

phenomenon. First, significant relationships between cer-

tain soil and hydrologic parameters and the percentage of

stems of sprout origin suggest better sprout success where

establishment conditions are most conducive for growth

(i.e., highly aerated substrate). Moreover, consistent

sprout success of many species within a community indicates

a common response to environmental conditions, particularly

the presence of hummocks. The lack of consistent patterns

of sprout production and growth in relation to environmental

conditions for the individual species reflects the incon-

gruency of their environmental tolerance ranges.

Most wetland trees are capable of producing sprouts

throughout their natural range of hydrologic conditions, as

well as under altered regimes (see Chapter 2). This indi-

cates that the primary mechanism by which sprout production

is expressed, and becomes important, is as an endogenously

produced trait that is adaptive under certain environmental

conditions.

Plants cannot respond to rapidly changing environmental

conditions, such as fluctuating water levels, by directly






85

adaptive genetic changes (Bradshaw 1965). Continual sprout

production in a fluctuating environment would be advanta-

geous for vegetative reproduction because of the rapidity of

environmental changes; sprouts produced in response to

inundation would not be likely to survive the event and

therefore would not be adaptive. Continuous sprout produc-

tion allows the plant to be in the appropriate regenerative

state before the critical environmental changes occur (Brad-

shaw 1965); sprout survival is increased by growth prior to

inundation, before physiological stress and the consequent

suppression of growth.

Sprout-producing species have the potential to regener-

ate regardless of the hydrologic state of the wetland.

Whereas sprouts are susceptible to the effects of flooding,

tree seedling establishment in wetlands is rare (Huenneke

and Sharitz 1986, Brandt and Ewel 1989) primarily due to

reduced metabolic rates following submersion (Pereira and

Kozlowski 1977). Under these circumstances, increased

longevity of a sprouting individual potentially increases

fitness; the likelihood of being present to produce seeds

increases when environmental conditions allow seed regen-

eration (Abrahamson 1980), for example, when water levels

remain low following a good period of seed production.

Thirty percent of the 2900 stems of the 7 common dicot

species recorded in this study of wetlands originated as

sprouts; given the methods used to determine stem origin, I









believe this estimate is conservative. Such high percen-

tages of sprout-originated stems among the most common

species should have important repercussions for the

stability of communities and for rates of species succession

(see Chapter 5). Indeed, based on community diversity and

structural complexity, Monk (1968) ranked dicot-dominated

bayheads and mixed hardwood swamps in north-central Florida

as seral stages replacing cypress domes. Although cypress

domes are not obligatory for the formation of mixed hardwood

swamps, in the absence of fire, particularly following

disturbance (i.e., logging, drainage), cypress is not likely

to replace itself and cypress dominated systems are then

encroached upon by dicots (Rochow 1985, Brandt and Ewel

1989).

Abrahamson (1980) describes a general pattern of shift-

ing from seed reproduction to vegetative reproduction with

succession. The propensity of dicots to continually produce

advance regeneration in the form of sprouts, even when there

is no mechanical injury to the tree, that are capable of

surviving a range of environmental conditions is an obvious

advantage over conifers: the possibility of species turn-

over is much reduced. Increased vegetative reproduction in

late successional forests is likely to be a consequence of

this advantage and contributes to the unlikelihood of coni-

fers naturally replacing a hardwood forest.









In conclusion, establishment of stems by sprouting is

an important mechanism of regeneration in north-central

Florida wetlands, especially for dicots capable of con-

tinually producing basal sprouts. Sprout survival of all

species appears to be determined by the ability to avoid

inundation, and is facilitated by the presence of hummocks

on which parent stems are situated. Successful vegetative

regeneration is likely to be a contributing factor in

reduced species turnover rates and stabilization of mature

community structure.













CHAPTER 4
APICAL DOMINANCE: MECHANISMS AND ECOLOGICAL SIGNIFICANCE


Introduction

Interactions between specific environmental conditions

and mechanisms of apical dominance need to be understood in

order to predict sprout production patterns. Whereas the

temptation exists to infer process and mechanism from pat-'

terns, the complexity of ecological systems often undermines

such simplistic approaches. Such is the case for sprouting

patterns in wetlands of north-central Florida; inference of

mechanisms of apical dominance from sprout distributions in

Lake Oklawaha would lead to different conclusions than

inferences from distributions in natural wetlands.

Sprout distributions in natural wetlands suggest that

there is little interaction between inundation and sprout

initiation, insofar as the capacity to produce sprouts is a

species-specific trait and is independent of hydrologic

conditions. Avoidance of submersion, however, is the key to

sprout survival (Chapter 3). Observed increases in propor-

tions of stems of sprout origin with increase in water depth

in Lake Oklawaha suggest that increased anoxia-induced

stress stimulates sprout production. Inundation of plant

roots leads to a sequence of responses that could affect






89

apical dominance of sprout initiation; results of studies of

the artificially impounded floodplain forest in Lake Okla-

waha and natural forested wetlands do not clearly indicate

the nature of that relationship.

Sprout production as a consequence of relaxed apical

dominance could be induced by a number of environmental

stimuli resulting both directly and indirectly from flood-

ing, such as altered root-to-shoot ratios, light on stems,

hormone balance, or plant water status. An initial result

of flooding is generally high root mortality (Harms et al.

1980). A root system with many damaged roots cannot support

a large amount of foliage, and thus flooding is often fol-

lowed by crown dieback. There is evidence for this in the

sparse foliage of the floodplain forest canopies 18 years

after impoundment in Lake Oklawaha. Loss of leaf area would

result in the reduction of auxin production sites. The

theory of apical dominance by hormonal control (e.g., Ruben-

stein and Nagao 1976) predicts that the consequent reduction

in auxin concentrations would reduce apical dominance and

initiation of sprout growth could occur. Whereas this is

feasible, there is some evidence to suggest that flooded

plant auxin concentrations increase due to reduced basipetal

movement and reduced auxin breakdown rates; in addition,

cytokinin production is reduced following root death (Reid

and Bradford 1984). The altered auxin and cytokinin con-

centration balance presumably enforces apical dominance







90

(Salisbury and Ross 1985). Indirect effects of foliage loss

are increased light penetration and temperature of the

stems. Epicormic branching is thought to be stimulated by

the increased light reaching stems when canopies are opened

(Trimble and Smith 1970) or the associated heating of stems.

Although crown loss and thus light penetration into the

forest often increase with depth of inundation, it is un-

clear whether these indirect effects of flooding initiated

sprout production in Lake Oklawaha. Foliage stress and

dieback do not occur in swamps with natural water level

fluctuations. Proportions of stems of sprout origin in

natural swamps are not related to stem density, which sug-

gests that there is no relationship with canopy closure.

Furthermore, there is apparently no direct effect of flood-

ing on sprout production in natural swamps; sprout produc-

tion is not related to any of the measured hydrological

parameters (Chapter 3).

In contrast with the hormonal explanation of apical

dominance, the theory of apical dominance by lateral bud

sink strength for plant water predicts initiation of lateral

bud growth when their relative sink strength for available

xylem water is greater than apical bud sink strength

(McIntyre 1987). These conditions may exist following

inundation when stomata close, a common initial response to

flooding (Gill 1970). With stomatal closure, transpiration

rates are reduced and stem xylem water potential remains









high (Pereira and Kozlowski 1977, Reid and Bradford 1984).

These conditions are necessary to initiate sprout growth by

increasing relative sink strength of lateral buds for water.

To the extent that the apical dominance theory of bud

sink strength applies, species that differ in the duration

of time required to resume normal plant water relations

following inundation are expected to have different periods

of time during which apical dominance is reduced by flood-

ing. Red maple (Acer rubrum L.) and swamp tupelo (Nvssa

sylvatica var. biflora (Walt.) Sarg.) are common in many

types of forested wetlands in north-central Florida. Swamp

tupelo is very flood tolerant (Eggler and Moore 1961, Harms

1973) and is commonly found in standing water under natural

conditions. Red maple grows more often in areas that drain

frequently; only scattered individuals can be found growing

naturally in standing water. In Lake Oklawaha and in natur-

al wetlands, red maples are prolific sprout producers where-

as swamp tupelo sprouts are rare. For these two species

under these conditions, the degree of flood tolerance is

negatively associated with sprout production. I hypothesize

that flood tolerant species like swamp tupelo recover normal

transpiration rates more rapidly following inundation than

less flood tolerant species, like red maple. A longer

period of inhibited transpiration of red maple trees would

allow lateral buds to initiate growth, thus increasing

sprout production rates.