Ecology of maritime forests of the southern Atlantic Coast: a community profile

              Biological Report 30
May 1995
National Biological Service
The National Biological Service publishes five technical report series. Manuscripts are accepted from Service
employees or contractors, students and faculty associated with cooperative research units, and other persons whose
work is sponsored by the Service. Manuscripts are received with the understanding that they are unpublished.
Manuscripts receive anonymous peer review. The final decision to publish lies with the editor.
Editorial Staff
WAGING EDITOR
Paul A. Opler
ASSISTANT BRANCH LEADER
Paul A. Vohs
SCIENTIFIC EDITORS
Elizabeth D. Rockwell
James R. Zuboy
TECHNICAL EDITORS
Jerry D. Cox
Deborah K. Harris
VISUAL INFORMATION SPECIALIST
Constance M. Lemos
EDITORIAL ASSISTANT
Martha W. Nichols
EDITORIAL CLERK
Donna D. Tait
Series Descrhtions
Biological Report ISSN 0895-1926
-Technical papers about applied research of limited scope.
Subjects include new information arising from comprehensive
studies, surveys and inventories, effects of land use on fish
and wildlife, diseases of fish and wildlife, and developments
in technology. Proceedings of technical conferences and
symposia may be published in this series.
Fish and Wildlife Leaflet ISSN 0899461X
Summaries of technical information for readers of non-technical
or semitechnical material. Subjects include topics of
current interest, results of inventories and surveys,
management techniques, and descriptions of imported fish
and wildlife and their diseases.
Fish and Wildlife Research ISSN 1040-2411
Papers on experimental research, theoretical presentations,
and interpretive literature reviews.
North American Fauna ISSN 0078-1304
Monographs of long-term or basic research on faunal and
floral life histories, distributions, population dynamics, and
taxonomy and on community ecology.
Resource Publication ISSN 0163-4801
Semitechnical and nonexperimental technical topics
including surveys; data, status, and historical reports;
handbooks; checklists; manuals; annotated bibliographies;
and workshop papers.
Copies of this publication may be obtained from the Publications Unit, U.S. Fish and Wildlife Service,
I849 C Street, N.W., Mail Stop 130, Webb Building, Washington, D.C. 20240 (call 703-358-17111, or
may be purchased from the National Technical Information Service (NTIS), 5285 Port Royal Road,
Springfield, Virginia 22161-0002 (call toll free l-800-553-6847).
Biological Report30
May 1995
BY
Vincent J. Bellis
Janet R. Keough, Project Officer
National Biological Service
Southern Science Center
700 Cajundome BZud.
Lafayette, Louisiana 70506
U.S. Department of the Interior
National Biological Service
Washington, D.C. 20240
Page
Preface ..................................................
Abstract .................................................1
CHAPTER 1. General Introduction .................................. 3
Definitions ..............................................4
Maritime Shrub Community ................................... 4
Maritime Evergreen Forest .................................... 4
Maritime Deciduous Forest .................................... 5
Coastal Fringe Evergreen Forest ................................. 5
Coastal Fringe Sandhill ...................................... 6
Maritime Swamp Forest ..................................... 6
Maritime Shrub Swamp ..................................... 7
InterdunePond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Geographical Distribution ...................................... 7
Barrier Island Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
CHAPTER 2. The Maritime Environment ...............................17
Introduction .............................................18
Climate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8
Oceanic Salts ............................................ .23
Soil Formation and Mineral Cycling ................................27
Hydrology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
Wetlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..3 1
Fire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..3 3
CHAPTER 3. Flora of Maritime Forests ................................35
Introduction .............................................36
Latitudinal Gradient in Floristic Composition ...........................36
Donation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..3 6
Succession ..............................................41
Origin of the Maritime Forest ....................................45
Fungi and Lichens ..........................................45
CHAPTER 4. Fauna of Maritime Forests ...............................47
Introduction .............................................48
Invertebrate Fauna .........................................48
Snails and Slugs (Pulmonate Castropods) ............................48
Spiders...............................................4 8
Insects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...4 9
Insects of the forest floor .....................................49
Wasps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..4 9
Blood-feeding arthropods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Nuisance insects ..........................................50
Vertebrate Fauna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50
Introduced Fauna ..........................................55
Introduction ............................................55
Domestic Animals .........................................56
Exotic Birds and Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Fauna1 Diversity .........................................60
ii
c ER 5. Management of Maritime Forests ...........................61
Introduction .............................................62
Management of Native Vegetation .................................63
IvIaritime Forest Fragmentation ...................................64
Effects of Highway Construction ..................................64
Recreational Impact 0x1 the Biota ..................................65
Effects of Subdivision Development on the Herp&ofawm . . . . . . . . . . . . . . . . . . . . . 69
Fire Management ..........................................70
Rare Plants and Animals ......................................70
Current Status of Maritime Forests .................................72
Regulation of Development: A Case History from North Carolina ................72
CHAPTER 6. Research Needs .................. . . . . . . . . . . . ..- . . . . 75
Introduction .......................... . . . . . . . . . . . . . . . . . . 76
Ecological Questions ...................... , . . . . . . . . . . . . . . . . . 76
Management Needs ...................... . . . . . . . . . . . . . . . . . . 76
Acknowledgments ....................... ..................77
References ........................... . . . . . . . . . . . . . . . . . . 77
Appendix A. Draft Use Standards for Maritime Forest Areas of Environmental
Concern................................................8 9
Appendix B. Checklist of Vertebrates Inhabiting the Barrier Islands of Georgia . . . . . . . . . 90
Figures
Fig. 1.1. Composite location map of barrier islands of the Atlantic coast of United States ....... 8
Fig. 1.2. Cross-sections of Gulf coast barrier islands ......................... 10
Fig. 1.3. Formation of barrier islands by submergence ........................ 10
Fig. 1.4. Development of barrier islands through breaching of complex spits ............ 11
Fig. 1.5. Types of barrier islands forming the Outer Banks of North Carolina ........... 12
Fig. 1.6. Geologic age of the barrier islands of Georgia ........................ 13
Fig. 1.7. Areas of natural vegetation on barrier islands of the Atlantic coast of Florida
from Duval County south to Volusia County ............................ 14
Fig. 1.8. Areas of natural vegetation on barrier islands of the Atlantic coast of Florida
from Brevard County south to Dade County ............................ 15
Fig. 2.1. Zones and subzones of broad-leaved forest in Florida ....................18
Fig. 2.2. Isoth-ms of mean daily minimum temperature of the coldest month in
Florida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..19
Fig. 2.3. Limits of the d&tfibutiOn of trees typical of three different forest types in
Florida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..2 0
. . .
111
Fig. 2.4. Hurricane probability at numbered stations along the Atlantic and gulf
coasts of the United States .....................................21
Fig. 2.5. Live oak defoliated by Hurricane Hugo at Monck’s Corner, South Carolina ........ 22
Fig. 2.6, Loblolly pine damaged by Hurricane Hugo at Francis Marion National
Forest, South Carolina ........................................23
Fig. 2.7. Depths and ages of sea level indicators from the Atlantic Continental Shelf
of the United States ........................................ .23
Fig. 2.8. Canopy height and atmospheric contribution of selected salts of the
maritime forest canopy at various distances from the ocean, Bogue Banks,
North Carolina ............................................25
Fig. 2.9. Chronological pattern of chloride deposition into the maritime forest canopy
at two locations on Bogue Banks, North Carolina .........................26
Fig. 2.10. Reference points for calculations of mineral inputs from salt aerosols at
Bogue Banks, North Carolina ....................................26
Fig. 2.11. Percent of salt spray collected at various points across a barrier island ......... 27
Fig. 2.12. Hydrologic cycle of a typical Holocene barrier island ...................30
Fig. 2.13. Idealized diagrammatic cross section of a barrier island, showing
water-flow pattern in the freshwater lens .............................30
Fig. 3.1. Latitudinal range limits of 50 species of trees and shrubs reported in
literature as forest constituents at 32 barrier island forest locations between
Miami, Florida, and Cape Cod, Massachusetts ...........................37
Fig. 3.2. Number of taxa occurring at l-degree latitude intervals from 25”N to 45”N ....... 38
Fig. 3.3. Taxonomic similarity (Jaccard’s Index) for assemblages of maritime forest
trees and shrubs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Fig. 3.4. Transect diagrams showing generalized physiography of forested and
unforested portions of Shackleford Banks, North Carolina ....................39
Fig. 3.5. The basic physiographic and ecological zones of a typical barrier island .......... 40
Fig. 3.6. Typical barrier island profiles found along the east coast of the United States ...... 40
Fig. 3.7. Generalized zonation of maritime vegetation: a comparison between a
northeastern and a southeastern barrier island ..........................41
Fig. 3.8. Successional relationships between plant communities on Cumberland
Island, Georgia ............................................42
Fig. 3.9. Schematic representation of successional stages in vegetative cover on
Shackleford Banks, a North Carolina barrier island ........................ 42
Fig. 3.10. Successional stages in the development of coastal dunes in Great Britain ........ 43
Fig. 3.11. Trends in plant community succession within the Fire Island, New York,
SunkenForest............................................4 3
iv
Fig. 3.12. Generalized transect across a system of parallel dune ridges at
Portsmouth, North Carolina, showing vegetation succession . . . . . . . . . . . . . . . . . . . 44
Fig. 3.13. Hypothetical profile development and succession of vegetation zones on an
accreting barrier island . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44
Fig. 4.1. Animal population drift in response to closure of inlets by longshore currents
and subsequent succession of old flood-tide deltas to hammocks in North
Carolina’s inter-capes zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Fig. 4.2. Relationship between amount of woodland habitat and number of reptile or
amphibian species inhabiting nine Atlantic coastal barrier islands . . . . . . . . . . . . . . . . 56
Fig. 5.1. Vegetation cover on a barrier island stabilized by artificial barrier dunes,
compared with a natural barrier pattern . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . 63
Fig. 5.2. Generalized patterns of onshore winds across undisturbed and disturbed
barrier island forest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65
Fig. 5.3. Effect of salt spray on red cedar canopy and sprout regrowth exposed by
cutting highway right-of-way . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Fig. 5.4. Effect of salt spray on live oak canopy and sprout regrowth exposed by
cutting highway right-of-way . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Fig. 5.5. Effect of salt spray on yaupon and red bay canopy and sprout regrowth
exposed by cutting highway right-of-way . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Fig. 5.6. Hypothetical effects of salt spray on maritime vegetation . . . . . . . . . . . . . . . . . 67
Fig. 5.7. Distribution of plant communities on a hypothetical barrier island, showing
design of two road systems . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . 69
Tables
Table 2.1. Climatological data for selected Atlantic Coast locations of the
southeastern United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 2.2. Mean values of physical-chemical parameters for five freshwater ponds in
the Nags Head Woods, North Carolina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 3.1. Characteristic plant communities of the barrier islands of the southeastern
UnitedStates....................................._.......38
Table 3.2. Bolete fungi in the Nags Head Woods, North Carolina . . . . . . . . . . . . . . . . . . 46
Table 4.1. Frequency of orb weaving spiders on several South Carolina barrier islands . . . . . . 49
Table 4.2. Insects collected in pitfall traps in maritime forest habitat on Cumberland
Island, Georgia, summer 1973 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Table 4.3. Numerical comparison of tetrapod vertebrate fauna of Shackleford Banks,
North Carolina, and the immediately adjacent mainland . . . . . . . . . . . . . . . . . . . . . 51
Table 4.4. Vertebrates of maritime forests on the Cape Hatteras National Seashore
and vicinity, North Carolina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
V
Table 4.5. Ubiquitous reptiles and mammals inhabiting barrier islands on the
Atlantic and Gulf of Mexico Coasts . . . . . . . . , . . . . . . . . . . . . . . , . . . . . . . . . 53
Table 4.6. Occurrence of the nonmarine species of coastal plain reptiles and
amphibians on Atlantic Coast barrier islands . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Table 4.7. Comparative colonization trends of herpetofauna on nine Atlantic Coast
barrier islands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
Table 4.8. Comparison of occurrence of mammals on 16 islands off the coast of
Georgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...57
Table 4.9. Numbers of feral ungulates on the Shackleford Banks, North Carolina, in
late summer 1978-1980 . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . 58
Table 5.1. Densities of the most common breeding forest birds along transects
through sites with differing disturbance levels on three barrier islands in South
Carolina................................................68
Table 5.2. Summary of reptile and amphibian transect data in maritime forest
habitats for three disturbance levels on two barrier islands in South Carolina . . . . . . . . . 70
Table 5.3. Status of maritime forest on the southeastern coast of the United States . . . . . . . . 72
vi
This is a synthesis of scientific information and literature concerning the maritime forests of the
southern Atlantic Coast of the United States. Information was gathered from many sources, including
published scientific literature, dissertations and theses, government agency reports and newsletters,
and unpublished reports.
Maritime forests are among the rarest and least studied coastal biological communities. Even the
term “maritime forest” remains ill-defined. Maritime forest8 are largely confined to barrier islands and
ocean-fringing sand dune systems. Published studies pertaining specifically to maritime forest are rare;
however, much information about maritime forest origin, development, and ecological function is
contained in the literature dealing with barrier islands. Most information about maritime forests is
descriptive in nature. Basic concepts about the causes of community zonation, the pattern of ecological
succession, the origin of wildlife populations, the degree of genetic isolation among animal populations,
the ecological significance of feral animal populations, and the possible barrier-island stabilizing of
maritime forests remain unresolved and controversial.
On the Atlantic Coast of the United States, the maritime forest resources have been neither identified
nor inventoried. Thus, there is a real danger that most maritime forest habitat not currently protected
by design or by accident (inclusion within areas protected for other reasons) will be destroyed or at least
functionally impaired by urban development by the end of this century.
This reports provides an understanding of the geological processes and environmental conditions
needed to evaluate controversies related to maritime forest ecology and management. The information
should be most useful to persons who desire, in a single source, a synopsis of the existing literature and
will provide a useful source of information for persons whose duty is to interpret maritime forests to
visitors.
Since some of the important literature is obscure, a reference section rather than the usual list of
literature cited has been provided. The final chapter enumerates some of the information gaps and
suggests some specific research needs. We hope this publication will stimulate additional support for
critically needed long-term and experimental research on understanding the ecological structure and
ecosystem functions of maritime forests.
This community profile was originally intended to be a part of one in a series coordinated by the U.S.
Fish and Wildlife Service’s National Wetlands Research Center, now the National Biological Service’s
Southern Science Center. Questions or comments about this community profile or others in the series
should be directed to:
Director
National Biological Service
Southern Science Center
700 Cajundome Boulevard
Lafayette, Louisiana 70506
vii
Vincent J. Bellis
I~epartmenr of Biology
East Carolirla University
Greenville, North Carolina 278.58-4353
Janet R. Keough
Project Officer
National Biologicai Service
Southern Science Center
700 Cajundome Boulevard
lkfayette, Louisiana 70506
Abstract. Maritime forests dominated by broadleaved evergreen trees and shrubs occur in a
discontinuous narrow band along the barrier islands and on the adjacent mainland from North Carolina
to Florida. The flora and fauna of maritime forests typically consist of a distinctive subset of theregional
biota that is particularly well adapted to survive the elevated salt content, limited availability of fresh
water, soil erosion and dune migration, periodic seawater inundation, and wind damage associated
with oceanic storms. Maritime forests cover the more stable portions of barrier islands and coastal
dune ridges. They function as refugia for wildlife, provide storage capacity for groundwater, and help
stabilize the soil. Recent recognition of the relatively greater physical stability of maritime forests
compared to the beachfront has resulted in intensified urban development within them. Maritime
forests across the range have been increasingly impaired by clearing for roads and parking lots and
fragmented by subdivision development. Further development within maritime forests should
minimize impairment of their critical biological and ecological functions. Maritime forest management
should be directed toward reducing forest fragmentation and toward protecting their ecological
integrity.
ECOLOGY OF Mmrim~ FORESTS OF THE SOUTHERN ATLANTIC COAST 3
General Introduction
4 BIOL~GICALREP~RT 30
Definitions as a basis for distinguishing among maritime forests of the
southern Atlantic coast of the United States (modified
“Maritime forest” is a broadly inclusive term that can from Schafale and Weakley 1990).
be used to distinguish woody vegetation growing near any
of the world’s oceans. These forests often exhibit canopy
height limitations resulting from salt-aerosol impact and
have been distinguished from other types of coastal forest
on the ba.Qs of differences in growth form and the relative
abundance of particular woody plant species. The concept
of which forests are “maritime forests” can vary widely,
depending on the relative weighting of growth form and
species composition. Wells (1939) described a “salt spray
climax” community along the southeastern coast and
noted that the geographic limits of this community corre-sponded
closely with those of southern live oak (Quercus
virginiana), often a conspicuous component of the com-munity.
Several later authorities also emphasized the im-portance
of evergreen oaks in this forest type: evergreen
oak forest (Braun 1950) and maritime live oak forest
(Bourdeau and Oosting 1959; Burk 1962a). Otherauthori-ties
defined the type without mention of oaks: arborous
zone of the salt spray community (Boyce 1954), maritime
closed dunes (Raynor and Batson 1976), and upland mari-time
strand forest (Wharton 1978). An early description of
the coastal forests of North Carolina (Pinchot and Ashe
1897) used the term “maritime” in its general, meaning “of
the sea.” Pinchot and Ashe apparently accepted more than
one canopy type in their concept of maritime forest be-cause
they referred to the “maritime forests” of North
Carolina.
Until recently, the question of defining maritime forest
only inspired arcane debates among academicians. Cur-rently,
the issue has achieved practical significance as
land-use planners and managers cope with the tasks of
identifying and managing the remaining maritime forests.
The North Carolina Coastal Resources Commission
(CRC) recently defined maritime forests (Appendix A) as
“those woodlands that have developed under the influence
of salt spray on barrier islands and estuarine shorelines.”
The CRC further differentiated maritime forests from in-land
forests by their adaptations to high wind velocities,
salt-aerosol impact, and sandy soils characteristic of the
coastal environment. Concomitantly, the North Carolina
Natural Heritage Program developed a classification sys-tem
(Schafale and Weakley 1990) for biological corrunu-nities
of the coastal zone that recognizes several related,
yet distinguishable, communities within the limits of for-ests
with maritime forest characteristics.
The following descriptive outlines are presented as an
overview of thephysiographic locations and general vege-tation
of maritime forest communities. The community
types were defined on the basis of their physical and
fioristic expression along the North Carolina coast; the
de=iptions should serve, with appropriate modification,
LocatiOn
Stabilized sand dunes, dune swales, and sand flats
protected from saltwater flooding and most extreme salt
spray.
Hydrology
Poorly to excessively drained sands. May have a high
water table. Subject to heavy salt spray.
Vegetation
Dense growths of shrubs, most frequently Myrica cerif-era,
Ilex vomitoria, Baccharis halimifolia, Juniperus vir-giniana,
Zanthoxylum clava-her&is, and stunted Quer-
US virginiana. Other species include Toxicodendron
(Rhus) radicans, Smilax spp., Parthenocissus quinquefo-lia,
Vitis spp., and Callicarpa americana.
Associatiolls
May grade into maritime evergreen forest. May contain
interdune ponds. Grades into or sharply borders on, dune
grass on less protected or more actively moving dunes.
Grades into or borders on dry or wet maritime grassland
in areas that receive overwash. May grade into salt shrub
in lower places subject to brackish or saltwater intrusions.
Distin@hing Features
Distinguished from maritime wet and dry grassland and
dune grass by the natural dominance of shrub-sized woody
vegetation and from maritime evergreen forest by its more
exposed environment and lower stature. Boundary defined
(by Schafale and Weakley 1990) at full canopy height of
5 m. Distinguished from salt shrub by its occurrence on
upland sites only rarely and catastrophically subject to
saltwater intrusion and by vegetation composition.
synonym
Maritime thicket.
Maritime Evergreen Forest
Location
Old, stabilized dunes and flats protected from saltwater
flooding and the most extreme salt spray.
Hydrology
Terrestrial, xeric to mesic, well to excessively drained,
subject to moderate to light salt spray.
Ve tion
Low to moderately high tree canopy, often stllrltcd or
pruned into streamlined shapes by salt spray, Dominated by
combinations of Quercus virginian~~, Pinus t(le& and
Q. hemisphnerica, with a few other spwies. Typical under-story
species Persea borbonia (sensu stricti,), carpi’zl4s
caroliniunri, Juniperus virginiana, Ci2~7u~~j~orid~2, Osnmrl-thus
americanus, Ilex opnca, Prunus caro/iniancl, and ZMI-tho.*
ylum clava-herculis. Shrubs include J/~_T \lorrljtoricl,
Myrica cerifera, Sabal minor, and Ca/iicnrpa americano.
Vines such as Toxicodendron (Rhus) rcrdictlns. Vitis rotundi-folia,
Smiler spp., Parthenocissus guirrquefolia, Bignonio
(Anisostichus) capreolata, Berchenrin scandens, Ampelop-sis
nrborea, and Gelsemium sempervirens arc often impor-tant.
The herb layer is sparse and low in diversity, with
species such as Mitchella repens, Asplenium platyneuron,
Chasmunthium (Uniola) laxurn, Piptochaetium ( S t i p a )
avenacea, Galium pilosum, Dicanthelium (Panicum) com-mutatum,
Elephuntopus nudatus, and Pnss$oru htea
Assaciations
Frequently grades into maritime shrub at more exposed
edges. May border on dune grass or maritime grassland at
the edge of actively moving sand dunes or overwash
deposits. May grade into maritime swamp forest, maritime
shrub swamp, or interdune pond in wet swalcs.
Distinguishing Features
Distinguished from maritime deciduous forest by the
occurrence of Quercus virginiana a& Q. hemisphuerica u
the dominant and often only canopy hardwoods. Pinus tueda
may occur in both types; its abundance is determined by
natural and artificial disturbance. A southcm variant of this
forest type occurs in the Smith Island complex on the
southern coast of North Carolina. This southern Variant
includes Sabul palmetto as an important canopy dominant
and becomes conspicuous further south in South Carolina
and Georgia. Maritime evergreen forest is distinguished
from maritime shrub by a tree canopy higher than 5 m. It is
separated from maritime swamp forest and maritime shrub
swamp by the dominance of the same suite of canopy species
that are found in ma&me evergreen forest. It is distin-guished
from coastal fringe evergreen forest by its occur-rence
on barrier islands or the ocean side of Peninsulas.
Synonym
Maritime forest.
Maritime DeciduOUs Forest
Locations
Most protected parts of old, stabilized dunes and beach
ridges on widest barrier islands.
Terrestrial, dry to mesic, with little salt spray.
Vegetation
~~~~~~~ dominated by mixtures of Pinus tuedu and vari-o
us hardwoods, particularly Quercus fakata, Fugus
grandifi,fi@, Liquidambar styraciflua, Q. nigra, Carya
glabrr;r, and C. pallida. Understory trees include Carpinus
c~~rolinic~rta, Ilex opaca, Cornus jlorida, Vaccinium ar-boreum,
Ostt-ya virginiana, Juniperus virginiana, Sassa-fr<
ls albilium, and Hamamelis virginiana. Shrubs and vines
inclu&G~lylussaciafrondosu,Arundinariagiganteu, Cal-licrrrpa
americana, Myrica cerifera, Rhus copallinu, Vac-cinium
stumineum, Vitis rotundifolia, Toxicodendron
(Thus) rrtdicans, Parthenocissus quinquefolia, Smilax
bona-no,r, and Gelsemium sempervirens. The herb layer
includes Mitchella repens, Pteridium aquilinum, Prenan-thes
serpentaria, Asterpatens, Solidago spp., Panicum sp.,
Schizmchyrium (Andropogon) scoparium, Desmodium
spp,, Cnidoscolus stimulosus, and Galium hispidulum.
Grades into maritime swamp forest, maritime shrub
swamp, and interdune ponds in wet swales. May grade into
maritime evergreen forest seaward.
Distinguishing Features
Sometimes regarded as similar to mesic forests inland
and sometimes regarded as only one extreme of the mari time
forest category. While both statements are true to some
extent, this community includes many species not normally
associated with the maritime environment, in a topographic
and climatic environment not found inland. In general,
differentiation of species along a topographic moisture gra-dient
seems to be poorly expressed. Species occur here in
associations not generally found inland. This may be a result
of the more frequent disturbance, the continuous input of
nutrients by salt spray, or the more moderate temperature.
Synonym
Madime mesophytic forest.
Coastal Fringe Evergreen Forest
Locations
Flats and low hills near the mainland coast.
HYbIogy
Terrestrial, mesic.
Vegetation
Forest dominated by various mixtures of Quercus hemi-sphaerica,
e. virginiana, and Pinus taeda. Other canopy
6 BIOLOGICALRGORT 30
trees include Quercusfalcata, Carya glabra, Q. nip-a, and
Pinus palustris. The understory may include Osmanthus
americana, Persea borbonia (sensu stricto), Magnolia
virginiana, Ilex opaca, Juniperus virginiana, and Sassa-fras
albidum. The most typical shrub is Ibex vomitoria.
Other shrubs include Myrica cerifera, Hamamelis virgini-ana,
Sabal minor, and species of the understory. Vines
such as Vitis rotundifolia, Smilax bona-nox, Gelsemium
sempervirens, and Campsis radicaas are sometimes nu-merous.
The herb layer is generally sparse and low in
diversity; Mitchella repens and Aspleniumplatyneuron are
most typical.
dominant species. Other common herbs include Rhyn-chospora
sp., Schizachyrium (Andropogon) scoparium,
Stipulicida setacea, Euphorbia ipecacuanhae, Stylisma
(Bonamia) patens, and Cnidoscolus stimulosus.
Macrolichens such as Cladonia evansii and Cladonia spp.,
and sandhill mosses such as Dicranum condensatum are
prominent and often dominate.
Associations
Grades into xeric sandhill scrub on the deepest, driest
sands. Grades into maritime forest, pond pine woodland,
or streamside pocosin in wetter places.
Associations Distinguishing Features
Frequently grades to coastal fringe sandhill on higher,
drier sites. Usually grades into salt marsh or brackish
marsh.
Distingutshing Features
Most easily distinguished from maritime evergreen
forests by the mainland location. Floristically, somewhat
to much more diverse than maritime evergreen forests.
Distinguished from coastal fringe sandhills by a closed
forest canopy structure and predominance of the canopy
species listed under vegetation over the sandhill species.
Distinguished from other mainland forest communities by
the significant occurrence of species typically confined to
maritime areas, such as Quercus virginiana, Osmanthus
americanus, and Ilex vomitoria.
Distinguished from pine-scrub oak sandhills and xeric
sandhill scrub by the occurrence of maritime-associated
species such as Quercus geminata, Q. hemisphaerica,
Q. virginiana, Ilex vomitoria, and Cladonia evansii; ap-pear
to be confined to locations near the coast. Distin-guished
from wet pine flatwoods and mesic pine flatwoods
by their structure, which includes a significant scrub oak
component and less shrub and herb layer. They often have
abundant lichens and bare sand.
synonyms
Sandhill, coastal scrub forest, pine-scrub oak sandhill.
Maritime Swamp Forest
Synonym Locations
Maritime forest.
Coastal Fringe Sandhill
Wet areas in well-protected swales, edges of relict
dunes, and edges of freshwater embayments.
Locations Hydrology
Sandy areas such as relict beach-ridge systems, gener- Palustrine, seasonally or intermittently flooded or satu-ally
within a few kilometers of the coast. Less commonly rated, to intermittently exposed.
on dry, sandy fluvial deposits, as in river floodplains.
Vegetation
Hydrology Forest dominated by various wetland trees such as
Terrestrial, xeric because of excessive drainage. Nyssa biflOra, Acer rubrum, Liquidambar styraciflua,
Fra.xinus americana, Taxodium distichum, Pinus taeda,
Vegetation Quercus nigra, and Q. michauxii. Understory trees and
Open to sparse canopy of Pinus palustris, sometimes shrubs may include Carpinus caroliniana, Persea bor-with
P. taeda. Quercus virginiana may form occasional to bon&, My&a cerifera, Cornus foemina, Magnolia vir-frequent
clumps. Open to sparse understory dominated by giniana, Vacciniumfuscatum (atrococcum), and V. corym-
Quercus geminata, Q. laevis, and Q. hemisphaerica. Other bosum. Arundinarib gigantea may be common. Common
understory species may include Sassafras albidum, Nyssa vines in&k Berchemia scandens, Toxicodendron (Rhus)
sylvatica, Q. incanu, Q. margarettae, and Vaccinium ar- radicans, and Vitis rotundifolia. The usually sparse herb
boreurn. Shrubs such as Gaylussacia dumosa, llex glabra, layer may con&in Woodwardia virginica, W. areolata,
Mm’ca cerifera, Ibex vomitoria, and Osrnanthus ameri- Osmunda cinnamomea, 0. regalis var. spectabilis,
canus may occur in sparse to dense patches. The herb layer &-&me& cylindrica, Saururus cernuus, Mitchella re-vaks
with woody cover, with A&j& stricta usually the pens, and Carex VP.
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 7
Grades into maritime forest or maritime mesophytic
forest, occurring as inclusions within them or between
them and marsh.
istinguis
Distinguished by occurrence in nontidal maritime wet-lands
and its dominance by wetland trees other than Persea
palustris.
syltaonynn
Swamp forest.
Maritime Shrub Swamp
Locations
Wet dune swales and depressions on barrier islands.
Hydrology
Palustrine, seasonally flooded or saturated to intermit-tently
exposed.
Vegetation
Open to dense canopy of shrubs or small trees. Pet-sea
borbonia is the most typical dominant, although some
areas are dominated by Cornusfoeming. Occasional larger
trees such as Pinus taeda or Acer rubrum may be present.
Vines, particularly Smilax spp., Toxicodendron (Rhus)
radicans, and Berchemia scandens, often form dense tan-gles
above or among the shrubs. The sparse herb layer may
contain Osmunda cinnamomea, 0. regalis var. spectabilis,
Woodwardia virginica, Onoclea sensibilis, or Thelypteris
palustris var. pubescens. Clumps of Sphagnum may be
common.
Associations
Usually surrounded by maritime evergreen forest or
maritime deciduous forest. Occasionally may border on
dune grass, marsh, or interdunal pond communities.
Distinguishing Features
Distinguished by its occurrence in maritime nontidal
wetlands and its dominance by wetland shrubs or small
trees.
SyIIollyms
Maritime swamp forest, bay forest.
Interdune Pond
Locations
Depressions in active or relict dune areas on barrier
islands.
Permanently flooded to intermittently exposed. (Some-times
described as water table windows connected to the
local groundwater system [Kling 19861.)
Vegetation
Varies with depth of water. Deep-water areas may
have various floating or submerged aquatic plants, in-cluding
Azolla caroliniana, Ceratophyllum muricatum
(echinatum), Limnobium spongia, Riccia fluitans, Ric-ciocarpus
natans, Spirodela polyrrhiza, Wolfiella
gladiata Cfloridana), Utricularia gibba (biflora), Lemna
gibba, and Hattonia inflata. Shallow-water and intermit-tently
exposed areas have various freshwater marsh spe-cies,
such as Leersia oryzoides, Eleocharis baldwinii,
Typha angustifolia, Sacciolepis striata, Setaria magna,
Hydrocotyle ranunculoides, Bidens frondosa, Triade-nun
(Hypericum) walteri, Lycopus rubellus, Boehmeria
cylindrica, Thelypteris palustris var. pubescens, Zi-zaniopsis
miliacea, Cladium mariscus ssp. jamaicense,
Typha latifolia, Fimbristylis castanea, Juncus spp., and
Polygonum spp. Some pond margins have a border of
shrubs and trees such as Salix nigra, Acer rubrum, Nyssa
biflora, Rosa palustris, Cephalanthus occidentalis, and
Decodon verticillatus. Some have been invaded by the
aggressive weed Phragmites australis (communis).
Associations
Small areas, surrounded by dune grass, maritime wet or
dry grassland, maritime shrub swamp, maritime swamp
forest, maritime evergreen forest, or maritime deciduous
forest.
Distinguishing Features
Distinguished from maritime wet grasslands by having
standing water all or much of the year and by vegetation;
may be distinguished from the inland small depression
ponds by their location on barrier islands. Distinguished
from tidal freshwater marsh by the lack of fluctuation in
water levels.
syfl0nynI.s
Dune marsh, dune swale, sedge.
Geographicall Distribution
Maritime forests occur all along the Atlantic Coast of the
United States. The distribution is not continuous. Forest
cover is interrupted by bays and inlets, by narrow barrier
island segments too unstable to support forest growth, and,
increasingly, by urban development. Adjacent maritime for-ests
are often floristically similar to one another and show
strong floristic affinity with nearby mainland forests. On a
8 BIOLOGICAL REPORT 30
finer scale, subtle floristic differences have been noted with to the Appalachian oak forest region. In southeastern
respect to the relative abundance of plant species in nearby Massachusetts, Rhode Island, New York, and New
forests or on adjacent islands. The cumulative effect of these Jersey, the barrier island forest vegetation fits into
subtle floristic changes becomes evident when the maritime the northeastern oak-pitch pine region. The
forest flora of Cape Cod, Massachusetts, is compared with transitional zone from the Delmarva Peninsula to
that of Cape Canaveral, Florida. The extreme locations are North Carolina can be considered part of the
quite different floristically, although the shifts in species southeastern oak-pine forest, but northern beach
composition are transitional and without sharp discontinui- grass (Ammophila breviligulata) and deciduous
ties. oaks remain dominant.
Godfrey (1976a:8) described the floristic gradient
along barrier islands of the Atlantic Coast of the United
States (Fig. 1.1): as
The region from Maine to New Hampshire provides
a meeting ground for typically southern species and
those of the boreal north. In southeastern Maine,
spruce and fir trees mingle on sand dunes with pitch
pines and oaks. In general, the Maine barriers are
part of the northern hardwoods region; those of
northern Massachusetts and New Hampshire belong
From North Carolina to northern Florida and the gulf
coast, the barrier island vegetation is part of the
southeastern evergreen oak-pine subunit of the
oak-hickory and southeastern pine forest. The
presence of sea oats (Uniola panicdata) and live
oak (Quercus virginiana) distinguishes this
vegetation from that found inland. In south Florida,
the flora of the Caribbean plays an important role in
the vegetation, while on the gulf coast there is a rich
coastal grassland.
State
Alabama
conoecticut
Delaware
Florida
Georgia
Louisiana
Maine
Maryland
Massachusetts
Mississippi
New Hampshire
New Jersey
New York
Noah Carolina
Rhode Island
South Carolina
zz?s
Virginia
18states
N”“,r” TotaI
islands acreage
5
14
2
28,200
2,362
10,100
80 467.710 :.NY fr
15 is
;
21
if35;600
41,120
I%%
37:600
?%
2%
146:450
14gFI
Beach
PO01
h
295- 1,60-5,152-
9 Cape Canaveral
Miami Beach
-.-..- ._._
/Timbalier Island
Isle Dernieres
I%. 1.1. Composite Ioc&on mq, of barrier islands of the Atlantic coast of United States (U.S. Fish a4 Wildlife Service 1990).
~?J~OLOOY OF bhHIlhE FORESTS OF THE SOUTHERN ATLANTIC COAST 9
Godfrey recognized three major barrier island sections
that could be distinguished geographically and floristically
as follows: (1) northern section (Maine to New Jersey), (2)
transition or central section (Delmarva Peninsula), and (3)
southern section (North Carolina to Florida). This report
primarily addresses the southern section of the reef islands
on the southeastern coast of the United States.
Godfrey further subdivided the southern section into
four subsections based primarily on geomorphology.
According to this subdivision, the Outer Banks of North
Carolina extend from near the northern boundary of the
state with Virginia south to Beaufort Inlet. The Outer
Banks are readily exposed to oceanic storms and exhibit
relatively high rates of barrier island retreat (Fig. 1.1).
West and south of Beaufort Inlet to Cape Romain, South
Carolina, the barrier islands are closer to the mainland,
are generally more protected from oceanic storms, and
support more stable dunes and more extensive maritime
forest cover. The Georgia Embayment, south of Cape
Romain, is characterized by low wave energies except
during hurricanes. Here, the Sea Islands occupy the most
protected section of the south Atlantic Coast. These
islands typically consist of Holocene beaches attached to
older Pleistocene beach ridges, and the oldest portions
have remained stable long enough to develop fertile soils
that support vigorous maritime forest cover. The north-ern
Atlantic Coast of Florida above Jacksonville appears
to represent an extension of the Sea Island system. Be-tween
Jacksonville and Cape Canaveral, maritime for-ests
are scattered along a narrow barrier island system.
Tropical species, including wild coffee (Psychotriu ner-vosa),
bloodberry (Rivina humilis), and naked wood
(Myrcianthus fragrans), begin to appear as shrubs and
small trees at Canaveral National Seashore. These and
other tropical species increase in abundance, height, and
species diversity farther south (A.F. Johnson, Florida
Natural Areas Inventory, personal communication).
South of Cape Canaveral, quartz sand beaches are re-placed
by increasing concentrations of carbonate sands,
and the sand ridges are replaced by limestone. The bar-rier
islands and beaches of Florida have become so
completely modified by urban development and intro-duced
exotic species such as Australian pine (Casuarina
equisetifolia) that their predevelopment characteristics
cannot be determined. In the Florida Keys, south of
Miami, the maritime forest containing Virginia live oak
(Quercus virginiana) is completely replaced by tropical
evergreen forest and mangrove swamps.
Barrier Island Origins
Maritime forests of the southeastern United States
develop almost exclusively on barrier islands or coastal
sand ridges. Is this distribution pattern simply a fortui-tous
circumstance of geography, or are there certain
characteristics associated with barrier island microcli-mates,
hydrology, soils, and other factors contributing to
development of that particular forest cover termed
“maritime forest”? It is beyond the scope of this report
to review the history of geological controversies con-cerning
the origin of barrier islands; however, to appre-ciate
the discussions of plant succession, fauna1 distribu-tion,
and community stability within maritime forests, a
basic understanding is necessary.
A barrier island is a narrow strip of deposited sand located
some distance offshore from the mainland. Barrier islands
form along seacoasts throughout the world wherever there
is an adequate supply of sand-size sediments, a low, sloping
coastal plain, and a wave-dominated energy regime with
tidal ranges of less than 3 m (S. R. Riggs, East Carolina
University, personal communication; Bascom 1980). Bar-rier
islands and maritime forests on them are geologically
ephemeral features. Barrier islands are formed and main-tained
by changing sea level in three possible ways. First,
when sea level remains relatively stable for some time,
barriers may prograde seaward with a series of parallel beach
ridges if there is a net surplus of sand, or they may migrate
landward by shoreface erosion, overwash, and inlet migra-tion
processes if there is a net deficiency of sand. Second,
when sea level is rising relative to land, landward migration
processes dominate but at significantly increased rates.
Third, when sea level is falling relative to land, the barrier
island progrades seaward, leaving a series of parallel beach
ridges, ultimately stranding the former barriers as a series of
sand ridges above and behind a new barrier island system.
Thus, the net retreat or advance of the shore is dependent on
the availability of sand, as well as on changes in sea level.
Three different explanations are plausible for the origin
of the southeastern barrier islands. Otvos (1970) presented
evidence suggesting some Gulf of Mexico Coast barriers
formed by emergence of submarine bars (Fig. 1.2). Hoyt
(1967) suggested that most Atlantic Coast barrier islands
originated by submergence of relict dune ridges (Fig. 1.3),
whereas Fisher (1968) thought that they formed by progra-dation
of sandspits entrained by headlands (Fig. 1.4). It
became evident to Pierce and Colquhoun (1970) that these
different explanations may not be mutually exclusive.
Schwartz (1970) attempted to synthesize the explanations
into a single conceptual model; thus the engulfed ridge of
Hoyt became a “primary” barrier island, while Fisher’s
breached spit and Otvos’s emergent bar became “secon-dary”
barrier islands.
The barrier islands of the soutbeastem United States are
now thought to represent complex features in which primary
barrier islands are modified by numerous processes to pro-duce
complex secondary barriers. Pierce and Colquhoun
IO BIOLOGICAL REPORT 30
Laguna Padre Island We
-1Om
5 lokm
(c)
Sea level
1
2
1 5 IOkm
a 1. Pleistocene
a 2. Holocene beach and eolian complex
[2771 3. Holocene brackish lagoonal, bay-sound, and
estuarine sediments
m 4. Holocene open-marine subtidal foreshore
sediments
B 5. Holocene alluvium
:--&&;5
d!FY=F km Vertical exaggeration=100
5
Fig. 1.2. Cross-sections of Gulf
coast barrier islands. (a) =
Padre Island; (b) =Galveston
Island; and (c) = Pine Island
(from Otvos 1970; used with
permission of Geological
Society of America).
Fig. 1.3. Formation of barrier islands by submergence.
1. Beach or dune ridge forms adjacent to shoreline.
2. Submergence floods area landward of ridge to
form barrier island and lagoon (from Schwartz 1971
after Hoyt 1967; used with permission of Geological
Society of America).
ECOLOGY OF MARITIME FORESTS OF THE 5iovn-r~~ ATLAN~C COAST 11
Barrie; island
Fig. 1.4. Development of barrier islands (indicated by dashed
lines) through breaching of complex spits (from Schwartz
1971 after Fisher 1968; used with permission of Geologi-cal
Society of America). Numbers l-5 indicate a series of
prograded beaches.
(1970) consider the Outer Banks of North Carolina to have
started as a primary barrier along a topographic high zone
formed by an older barrier island during a temporary stand-still
associated with a previous Pleistocene sea level high.
As the sea level rose during the last 5,000 years, the modem
shoreline intercepted the older barrier and inundated the
low-lying land behind, detaching it from the mainland. The
present configuration of the North Carolina Outer Banks
evolved by the modification and migration of this primary
barrier and associated headlands and by formation of secon-dary
barriers by spit progradation across shallow open bays
on the Continental Shelf. Only about 40% of the present
barrier consists of a modified primary barrier, and the re-mainder
is of secondary origin (Fig. 1.5).
The Sea Islands of Georgia were described as com-pound
barriers of relatively recent (4,000-5,000 years)
Holocene barriers welded onto a core of older Pleisto-cene
ridges (Fig. 1.6)(Johnson et al. 1974). Different-age
portions of barrier islands can be distinguished on the
basis of their soils. For example, Sea Island has poorly
developed soil because of insufficient time for forma-tion;
on the other hand, St. Simon’s Island has more
mature soil to a depth of more than 2 m in places
(Johnson et al. 1974).
South toward St. Augustine, Florida, Amelia and Little
Talbot Islands are similar to the Sea Islands of Georgia. The
modem sands of Little Talbot Island are welded onto the
older Pleistocene core of Big Talbot Island.
The “drumstick” shape of the Sea Islands was inter-preted
by Hayes (1979) as a response to the relatively
great tidal amplitude in the Georgia Embayment. Inter-action
of waves on the major ebb-tide deltas (formed by
strong tidal currents through the inlets) leads to long-shore
drift and formation of curved beach ridges at the
tips of the islands.
Florida has the longest coastline in the coterminous
United States. The Atlantic coast north of Miami consists
of sandy beaches fronting a chain of barrier islands
(Figs. 1.7 and 1.8). The sands of the beaches north of
Cape Canaveral were derived by southerly longshore
sediment transport of quartz sands originally weathered
from Piedmont rocks in Georgia and the Carolinas (Giles
and Pilkey as cited in Johnson and Barbour, 1990). Like
the barrier islands to the north, the Florida barrier islands
seem to occupy locations determined by geological
events of the Pleistocene (Johnston and Barbour 1990).
From St. Augustine to Boca Raton, the modern barriers
are perched on an underlying coquina ridge known as the
Anastasia Formation. South of Boca Raton, the beach
sediments are composed of a mixture of quartz sand and
fragmented molluscan shell hash. The Pleistocene Anas-tasia
Formation grades southward into Pleistocene
oolites, a series of limestone units that occur at Miami
and southward and form the substrate of the keys. Along
the Florida Keys, the southern evergreen maritime forest
is replaced by mangrove islets and palm-pine scrub.
12 BIOLOGICAL REPORT 30
Banks
Hatteras Island
Cape Hatteras
m Secondary barrier
m Eroded and modified primary
m Primary barrier with frontal progradation
Fig. 1.5. Types of barrier islands-forming
the Outer Banks of
North Carolina (from Pierce
and Colquhoun 1970).
ECOLOGY OF MAREME FORESTS OF THE SOUTHERN ATLANTIC COAST 13
Wilmington Island,
L - Y” island
St. Catherine’s
Georgia
Jeckyil island
Holocene barrier island
Pleistocene barrier island
(Silver Bluff)
0L_-t!-Y
Km
Fig. 1.6. Geologic age of the barrier islands
(Sea Islands) of Georgia (modified from
Johnson et al. 1974 after Hoyt 1968).
14 BIOL~CICAL REPORT 30
Fii. 1.7. Areas of natural vegetation on barrier
islands of the Atlantic coast of Florida
from Duval County south to Volusia
County (from Johnson and Barbour 1990).
Several of these are state parks (SP), state
recreation areas (SRA), national monu-ments
(NM), national seashores (NS), and
national wildlife refuges (NWR).
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 15
Lake Worth inlet
S.L. Worth Inlet
Hillsborough Inlet
Fig. 1.8. Areas of natural vegetation on barrier
islands of the Atlantic coast of Florida
from Brevard County south to Dade
County (from Johnson and Barbour 1!200).
See Fig. 1.7 for site label.
FLOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 17
.“.’..:j/ ‘1
The Maritime Environment
18 BIOLOGICALREPORT 30
It is widely recognized that the biological communi-ties
almost exclusive to barrier islands owe their charac-teristic
structure to some factor or combination of factors
related to their maritime environment. Special environ-mental
conditions associated with barrier-island envi-ronments
typically include exposure to potentially toxic
levels of salt; exposure to strong winds, shoreline ero-sion,
and ocean overwash during storms; low levels of
plant nutrients in the soil; low and unpredictable supply
of freshwater; and unstable soil substrate that is subject
to wind or water erosion. Along the southeastern coast
of the United States, the proximity of the barrier islands
to the warmer waters of the Gulf Stream results in a
northerly shift in the frost line and winter temperatures
that are somewhat higher than inland at a given latitude.
The proximity of the barrier islands to the sea tends to
dampen seasonal temperature extremes. Barrier islands
also tend to be geologically unstable. Inlets open and fill,
and entire islands slowly migrate before the advancing
sea. Fire frequency may not be directly related to condi-tions
of the maritime environment but can exert a signifi-cant
impact on island biota.
Climate
On the barrier islands, geological processes determine
the types of habitat available, whereas climate sets
broad limits on such critical environmental conditions as
temperature extremes, solar energy input and day length,
storm exposure, and availability of fresh water.
TRF - Tropical forest zone
TBEF Temperate broad-leaved
evergreen forest zone
SMHF Southern mixed hardwood
forest zone
TRFEBEF Transition subzone of
TRF to TBEF
TBEWRF - Transition subzone of
TBEF to SMHF
The barrier islands of the Atlantic Coast between Vir-ginia
and the Florida Keys extend almost 1,600 km along
a roughly north-south axis. The climate ranges from tem-perate
to subtropical; most of the area is best described as
warm temperate (Eastern U.S. road map).
South of Cape Hatteras, the maritime climate is influ-enced
by the warmer water of the Gulf Stream, whereas
north of the Cape, the nearshore zone is influenced to a
greater extent by colder water moving south from the
North Atlantic Ocean with the longshore Virginia Cur-rent.
Biologists have long recognized this natural bound-ary
in their distinction between “Virginian” and “Caro-linian”
biotas. Northeastern North Carolina represents a
transition or tension zone between these two biotas.
Many species of plants, as well as marine and terrestrial
animals, reach their northernmost or southernmost range
limit here and may exist as pairs competing for the same
habitat. The presence of this transition zone may account
for the greater diversity among plants and vertebrate
animals along the northern barrier islands of North Caro-lina
compared than in other locations along the southern
barrier island system (Otte et al. 1984).
Another biotic effect of climate is a greater northerly
range of southern and subtropical species along the bar-rier
island chain than at comparable latitudes inland. The
effect has been noted for maritime forests in New York
(Greller 1977) and Florida (Greller 1980). In Florida,
Greller (1980) mapped the distribution of three major
upland broad-leaved forest types (Fig. 2.1). These were
identified as tropical forest (tropical), temperate broad-leaved
evergreen forest (evergreen), and southern mixed
hardwood (hardwood). The tropical forest was domi-nated
by evergreen and drought-deciduous tropical taxa
Fig. 2.1. Zones and subzones of broad-leaved forest in
Florida.
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 19
(gumbo limbo [Burseru simaruba], wild tamarind
[Lysiloma latisiliqua], mastic [Mastichodendron
foetidissimum], and stoppers [Eugenia spp.]), and was
associated with a hot to very warm, subhumid to humid
climate. The evergreen forest, dominated by live oak
(Quercus virginiana) and palmetto palm (Sabal paf-metto),
occurred under warm to very warm, subhumid to
humid climatic conditions, The hardwood forest was
dominated by southern magnolia (Magnolia grandi-flora),
American beech (Fagus grandifolia), pignut hick-ory
(Carya glabra), flowering dogwood (Cot-mu flor-ida),
American holly (I&x opaca), and other taxa
commonly found in the coastal plain forests of the South-east.
The hardwood forest occurred in association with a
warm temperate and humid climate.
The boundaries between these climate regimes and asso-ciated
forest types correspond best to the average daily
minimum temperature of the coldest month (T&. The
boundaries defined by Greller (1977, 1980) (Fig. 2.2) were
tropical (Tmin = 12’ C), evergreen (Tmin = 1O.S’ C), and
hardwood (Tmin = 5.5” C). The range limits of several
indicator tree species in each of these forest types closely
followed the appropriate isotherms (Fig. 2.3). Furthermore,
the Tmin isotherm boundaries bend sharply to the north
immediately along the east coast of Florida. Each successive
forest type extends much farther to the north along the east
coast than along the west coast of Florida. This trend of
southern plant species reaching a more northerly limit im-mediately
along the coast than they do inland, appears to
extend northward at least as far as Cape Hatteras, North
Carolina.
A comparison of climatological records for selected
coastal locations along the southeastern Atlantic Coast
(Table 2.1) indicates the range in climate regimes for
maritime forests between Virginia and south Florida.
Mean percentage (of maximum possible) sunshine and
mean annual relative humidity vary little across the lati-tudinal
gradient between Norfolk, Virginia, and Miami,
Florida. Mean annual percentage (of maximum possible)
sunshine is within It3% of 65%, and mean relative hu-midity
is within z!z4% of 83% at all six locations (Ruffner
and Blair 1977, USDC-NOAA 1974).
Mean annual precipitation ranges from 1,135 mm/year at
Norfolk to I,5 19 mm/year at Miami. The intervening loca-tions
precipitation of 1,334 + 45 mm/year. Maximum pre-cipitation
occurs in July or August at all locations except in
Miami where it occurs in June (Ruffner and Blair 1977,
USDC-NOAA 1974).
Latitudinal differences of 6.1 to 8.4 km/s in mean
annual wind velocity are probably insignificant; how-ever,
the recorded maximum wind velocity of record was
highest at Jacksonville, Florida (87.5 km/s), but lowest
at nearby Savannah, Georgia (47.3 km/s). Prevailing
winds are from the west from Savannah northward, from
the northwest in north Florida, and from the east at
Miami (Ruffner and Blair 1977, USDC-NOAA 1974).
Temperature variables form the most conspicuous
gradient along the latitudinal axis between Norfolk and
Fig. 2.2. Isotherms of .5.5”C,
lOS”C, and 12°C mean daily
minimum temperature of the
coldest month (T& in Florida
(from Greller 1980; used with
permission of Torrey Botanical
Club).
20 BIOLOGICAL REPORT 30
-Cocco/obs diversifolia (northern limit
Sahel palmetto (northern limit)
- - --. Ouercus elba (southern limit)
Fig. 2.3. Limits of the distribution of a
tropical taxon (Coccoloba diversifo-lia),
a temperate zone evergreen spe-cies
(Sabul palmetto), and a temperate
zone deciduous hardwood (Quercus
albu) in Florida (from Greller 1980;
used with permission of Torrey Botani-cal
Club).
Miami. The mean annual temperature is 15.4” C at Nor-folk
and 24.0” C at Miami. Frost-free days range from
256lyear at Norfolk to 3 13lyear at Savannah to 365lyear
at Miami (Raffner and Blair 1977, USDA-NOAA 1974).
Given the ranges in climate variables noted pre-viously,
it seems reasonable to assume that growing
season, length of exposure to freezing temperatures, and
hurricane exposure may constitute the major climatic
factors corresponding to variations in maritime forest
biota. North of Cape Hatteras, the shoreline tends to face
east and northeast, whereas south of that location the
shore faces east, south, or southeast. Storm effects tend
to be greatest when storm winds are onshore. Winter
storm winds tend to come from the west and north,
whereas summer winds come from the west and south.
Along the Virginia and northern North Carolina coasts,
storm damage often results from northeasters during
spring months, while coastal residents south of Cape
Hatteras tend to be more concerned by the threat of
hurricanes from the southeast in late summer or autumn.
Table 2.1. Climatological data for selected Adantic Coast locations of the southeastern United States (Ruffner and Blair
1977 and USDC-NOAA 1974).
Precipitation
Temperature (” C) (mm)
x x x Month Wind
Location Annual (Jan.) (July) FFDa Sunshineb Annual of maximum Humidityd Directione Velocityf Maximumg
Norfolk, Va. 15.4 5.0 26.1 256 62 1,135 July 19 SW 7.6 55.8
Buxton, N.C. 16.8 8.3 25.6 296 63 1,384 Aug. 83 s 8.4 51.8
Charleston, S.C. 19.2 10.0 21.2 294 66 1,323 July 86 SW 6.3 51.0
Savannah, Ga. 19.1 11.1 27.2 291 63 1,308 Aug. 85 SW 6.1 47.3
Jacksonville, Fla. 20.8 13.3 28.3 313 62 1,295 July 85 NW 8.2 87.5
Miami, Fla. 24.0 19.5 27.8 365 67 1,519 June 81 E 6.6 53.3
aFreeze-free days.
b Annual possible percentage of sunshine.
iMonth of maximum precipitation.
Relative humidity (percent at 0 100 local time annual mean).
y Direction of prevailing wind.
Mean annual velocity (km/s).
gMaximum velocity (km/s), highest recorded.
E&OLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLAN~C COAST 2 1
Hurricanes form from tropical cyclones in the Atlan-tic,
Caribbean, or Gulf of Mexico (Simpson and
Lawrence 1974). Off the East Coast of the United States,
hurricanes tend to follow the warmer, less dense air
above the Gulf Stream. Since the Gulf Stream ap-proaches
closest to shore along the east coast of Florida
and again off Cape Hatteras, North Carolina, these two
areas serve as focal points for hurricane landfall
(Fig. 2.4).
The effects of a major hurricane on forest trees were
observed following Hurricane Camille, which struck the
Gulf of Mexico Coast in 1969, and were described by
Touliatos and Roth (197 1:288). Most of the direct dam-age
to trees from hurricanes is caused by high-velocity
wind. Camille came ashore with winds of over 89 m/s
and a record storm surge as high as 6.7 m. Wind effects
were evident for more than 160 km inland. Poorly an-chored
trees were uprooted, and well-anchored trees
were stripped of their leaves. Secondary effects included
salt-aerosol damage to foliage and flooding of root sys-tems
by brackish water.
Touliatos and Roth (197 1) concluded that a tree’s
ability to withstand hurricane winds was dependent on
the strength of the wind, the size and shape of the crown,
the extent and depth of the root system, the antecedent
soil moisture content, and the shape of the bole. They
assessed the degree and type of damage among 20 com-monly
occurring coastal native and ornamental trees. In
terms of resistance to breakage, uprooting, salt damage,
and subsequent susceptibility to insect attack and dis-ease,
live oak (Quercus virginiana) and palm (Sabal
palmetto) consistently exceeded all other species. Live
oak was described as having “exceedingly strong and
resilient” wood (Fig. 2.5). “Palm trees,” they noted,
“offer little surface to the wind because they have almost
no laterally extended crown. This characteristic makes
them a fairly wind-resistant tree, despite their close and
small root structure” (Touliatos and Roth 1971:288).
Common shallow-rooted trees, including dogwood
(Cornusflorida), water oak (Quercus nigru), sweet bay
(Magnolia virginiana), and red maple (Acer rubrum),
were among the least resistant to hurricane damage.
Hurricane Camille’s effects on forest canopy de-scribed
by Touliatos and Roth were confirmed by the
author of this report for Hurricane Hugo, which struck
Charleston, South Carolina, in September 1989. I had
visited the area in August to compare current vegetative
cover on the Isle of Palms with the described vegetation
(Coker 1905). Since 1905, the Isle of Palms has under-gone
intensive urban development, but much of the for-est
canopy had been left intact. Prior to the storm, many
residential streets and lawns were deeply shaded by live
oaks. Tall cabbage palms and loblolly pines were also
abundant canopy trees. In November, after the storm,
shrub vegetation that had been present in the interdune
area between the beach and the first line of homes had
been washed away or buried under sand. Almost all
large pines were broken off about a meter above the
ground.
Falling pine trees were a major cause of roof damage
in Hurricane Hugo; roof damage then led to increased
water damage to the inside of the houses. The storm
surge of up to 5.5 m flooded the lower floors of most
homes and resulted in irreparable damage to possessions
aProbability of occurrence (%);
ail hurmanes\great hurricanes
* Less than 1% occurrence
\
\
Fii. 2.4. Hurricane probability at numbered stations along the
Atlantic and gulf coasts of the United States. The probability
(expressed in percent) that a hurricane (winds exceeding
30 meters per second or 73 miles per hour) or a great hurricane
(winds exceeding 56 meters per second or 125 miles per hour)
will occur in any 1 year in an 80-km segment of coastline.
(Modified from Simpson and Lawrence 1971 as cited in U.S.
Department of Interior 1978.)
22 BIOLOGICAL REPORT 30
Fig. 2.5.7.G live oak (Quercus virginiana) near Monck’s Comer, South Carolina, was defoliated by the winds of Hurricane Hugo
in September 1989. The photo, taken in May 1990, shows new growth originating along the surviving branches (photo by author}.
on the ground level. Some pines were simply uprooted
and tipped over, resulting in structural damage to foun-dations
and service lines. In contrast to the pines
(Fig. 2.6), palms and live oaks remained. The surviving
oaks were stripped of their leaves and leafy branches,
and the palms stripped of most of their mature fronds.
The nearly closed evergreen forest canopy of August
now resembled more that of late autumn in a deciduous
forest. These observations about the different survival of
live oak and palm following damage by Hurricane Hugo
were confirmed for the uninhabited Bulls Island, South
Carolina (J. Nelson, University of South Carolina, per-sonal
communication).
I visited the Isle of Palms again in May 1990. Half a
year after Hurricane Hugo struck, rebuilding was well
under way, but some cleanup was still in progress. A
contractor, who removed and burned fallen and damaged
trees, estimated that 1.5 million cubic yards of wood and
branch debris had been removed from the Isle of Palms
and Sullivans Island (an area encompassing 1,024 ha of
forested land), and he noted that most of the debris was
from pine trees. The typically greater frequency of live
oak and palm within the canopy of southern barrier
island forests may be related, at least partly, to the greater
ability of these two species to survive storm damage.
Maritime forests and their sandy substrate are ulti-mately
dependent for their origin and maintenance on
changes in sea level. Sea level appears to respond to
long-period oscillations in climate. The present barrier
island system is thought to have assumed approximately
its current location and configuration about 5,000 years
ago, concomitant with a marked decline in the rate of
sea-level rise from about 0.3 m/century to 0.1 m/century
(Fig. 2.7).
At present, many scientists believe that the rate of
sea-level rise may soon increase relatively rapidly to a
level equal to or exceeding that existed before to the
origin of the present barrier island system. Any signifi-cant
increase in the rate of sea-level rise has obvious
implications for maritime forests. If rising sea level
hastens the process of barrier island migration, will mari-time
forests be able to keep pace? In more practical
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLAWIC COAST 23
Fig. 2.6. Loblolly pine (Pinus rue&) forest in Francis Marion National Forest, South Carolina, showing damage caused by Hurricane
Hugo in September 1989 (photo by author).
terms, barrier island managers are already recommend-ing
that the lowest portions of barrier island segments,
which are subject to overwash and flooding, be identified
and that further development in such locations be dis-couraged
(Cantral 1988). Because maritime forests typi-cally
occupy the highest, most stable portions of barrier
islands, one result of such a policy may be to increase
development on the few remaining maritime forests.
Oceanic Salts
The growth-inhibiting effect of salt has been thought
to be a major ecological factor governing floristic zona-tion
on barrier islands. Wells and Shunk (1938) reported
that the dominant woody plants fronting the ocean (wax-myrtle
[Myrica cerifera], yaupon [Zlex uomiroria], and
live oak [Quercus virginiana]) were all more salt tolerant
- Shells
+ Oolites
* Coralline algae
A Salt-marsh peat
l Beachrock
150 1 ,,, /,/,,I,I,, I, I,, , ,,,,I, I, I,, I,, I ,,I,
0 5 10 15 20 25 3 0 35
Time (thousands of years ago)
Fig. 2.7. Depths and ages of sea level indica-tors
from the Atlantic Continental Shelf
of the United States. The solid line is the
inferred sea level curve for the past 35,000
years; the dashed line indicates range in
sea level estimates inferred from the fossil
record (adapted from Milliman and Em-ery
1968; used with permission of Sci-ence).
D
24 BIOLOGICALREPORT 30
than loblolly pine (Pinus rue&z), a tree that usually with surfaces, its concentration in the atmosphere de-occurs
in greater abundance at some distance behind the creases. Farther back from the beach, the maritime forest
beach. Wells (1939) subsequently described a “salt- canopy gradually assumes the more uneven surface of a
spray-climax community” dominated by live oak mainland forest as individual tree height becomes more
(Q. virginiana) on Smith Island, North Carolina. He an expression of the genetic potential of the species
believed that live oak dominated the canopy in the Smith rather than a growth response to an inhibitory environ-
Island maritime forest because its salt tolerance gave the mental factor. Greatest salt damage to plants typically
slower-growing live oak a competitive edge over faster- occurs during the spring or early summer, just as new
growing but less salt-tolerant trees. buds are breaking.
The term “salt spray” has been extensively used to Plant leaves may become necrotic and die if subjected to
describe the salt aerosol that is blown over barrier islands excessive salt exposure. Wind-driven aerosols tend to con-from
the sea by onshore winds. Unless used in a direct centrate along the edges of leaves. Small, simple, smooth-quote,
the term “salt aerosol” will be used throughout edged leaves having a thickmesophyll, tough epidermis, and
this report to identify this material. thick cutin seem to withstand salt-aerosol impact better than
The salt-aerosol explanation of vegetative-cover larger, thinner lobed or compound leaves. Trees and shrubs
zonation was tested experimentally by Oosting and Bill- with small, salt-resistant leaves dominate the maritime forest
ings (1942). They evaluated the correlation between canopy nearest the sea. Less salt-tolerant hickories, sweet-plant
zonation and the environmental parameters of soil gum, maples, and lobe-leaved oaks generally increase in
moisture, soil salinity, soil temperature, air temperature, relative abundance with increasing distance from the beach
evaporation rate, salt-aerosol input, and relative humid- (Boyce 1954).
ity. Of these parameters, only salt-aerosol input corre- Salt ions appear to enter the leaves through cracks in
lated with the plant zonation pattern. the epidermis caused by vigorous bending and brushing
Convincing evidence about the toxic effects of salt together of twigs during high-wind conditions. Boyce
aerosols on vegetation was provided by Boyce (195la, (1954) has shown that in many types of leaves, excess
195 1 b; 1954). He experimentally investigated the origin, salt is translocated to the leaf tip. The resulting V-shaped
atmospheric transport mechanism, salt-deposition pat- yellowed or necrotic area with the apex of the V origi-tern
into the vegetation, and mode of entry into plant nating at the leaf midrib, constitutes a diagnostic charac-tissues
of ocean-derived salt aerosol. He also studied the teristic of damage from salt. Salt may accumulate in a
translocation and physiological effects of salt after it had leaf until it is killed; the dead salt-laden leaves then fall
entered the plants. from the tree. As a result, only portions of the affected
Boyce (1954) showed that maximum salt-aerosol im- plants rather than the entire plant are killed.
pact on vegetation occurs under conditions of strong Proffitt (1977) measured salt inputs at various loca-onshore
wind. Salt spray, propelled into the air after the tions and elevations within the maritime forest on Bogue
plunge of a breaking wave, becomes an aerosol entrained Banks, North Carolina (Fig. 2.8) but found no consistent
in the wind. The entrained aerosol flows with the wind seasonal pattern in salt deposition (Fig. 2.9). Proffitt used
and is deposited according to wind patterns determined the field data from his study to develop regression equa-by
the shape and texture of the underlying surface. Salt tions for predicting the atmospheric mineral inputs at any
is deposited when aerosol droplets fall on surfaces; salt- location where the topography is known. These equa-aerosol
concentration is greatest close to the ocean or tions were as follows: for chloride, y = 21.9x - 5.48; for
land surface. Vegetation along the windward edge of the calcium, y = 0.40x + 0.98; and for magnesium, y = 1.1 lx
maritime forest intercepts most of the salt. Unhardened - 0.26; where y represents the atmospheric inputs in
developing branches derived from terminal buds may grams per square meter per year and x is the topographic
grow into the space above the canopy, where they are index for the site (Fig. 2.10). Proffitt (1977) demon-killed
by salt desiccation. Terminal buds nearest the strated an inverse relation between maritime forest can-ocean
rarely complete their development. Death of the opy height and chloride input (Fig. 2.8). The correlation
terminal bud or branch produces a hormonal change in between measured canopy height above mean sea level
shrubs and trees, which results in growth stimulation to and measured chloride inputs during one year at six
previously repressed lateral buds. Continued loss of ter- locations across the barrier island yielded a correlation
minal growth, together with development of lateral buds, coefficient of -0.87 (P = 0.05).
produces the “espalier” or wind-sculpted appearance in Seneca and Broome (1981) found reasonable agree-the
maritime forest canopy. Close to the ocean, the ment between measured salt input into the forest canopy
maritime forest canopy is kept low and of uniform height and values predicted using the Proffitt equations at an-by
the effects of salt aerosol. As salt is lost by impact other site on Bogue Banks (Fig. 2.11). Proffitt also
lEco~o0~ 0~ MARITIME FORTH 0~ THE SOUTHERN ATLANTICCOAST 2.5
reported a relationship between canopy-species commu-nity
structure and the effects of salt-aerosol. Species
diversity and species evenness were lowest in the area of
maximum salt impact.
Exposure to salt aerosol is a major agent that regulates
both the height and species composition of the maritime
forest canopy. Both of these effects attenuate rapidly as salt
impact diminishes away from the seawardedge. This feature
of maritime forest structure also suggests that the forest type
termed “maritime forest” originates as a result of progressive
loss of canopy species from an existing, more diverse forest
with a floristic composition similar to that on the adjacent
mainland. If this scenario is correct, then mixed-hardwood-pine
barrier island forests are continuously transformed into
maritime forest as rising sea level and beach erosion cause
the zone of salt-aerosol impact to shift toward the mainland.
a. Canopy height above mean sea level
5 IO-I
.a, 8 -
+
6-
“IL I I I I
57 124 168214 264 396 564 680 879
Distance from ocean (m)
b. Atmospheric contribution of selected salts
120
100
80
60
-
‘;L
x
‘E
40
=
J!?
2
/
20
+ Cl (First year)
l cl (Second year)
= MI
A Ca
2 -
.___A
l- -___f-_ Et
I I I I I I I I 1
57 124 168 214 264 396 564 680 879
Fig. 2.8. Canopy height
and atmospheric
contribution of se-lected
salts of the
maritime forest can-opy
at various dis-tances
from the
ocean, BogueBanks,
NorthCamlina(from
Proffitt 1977).
Distance from ocean (m)
26 BIOLOGICAL REWRT 30
Station 879
16
12
8
Fig. 2.9. Chrono-logical
pattern of
chloride deposi-tion
into the
maritime forest
canopy at two lo-cations
on Bogue
Banks, North
Carolina (from
Proffitt 1977).
I975 1976
Collection periods
Atlantic Ocean
Fii. 2.10. Reference points
for calculations of min-eral
inputs from salt
aerosols at Bogue
Banks, North Carolina
(from Profftt 1977).
Topography index at a station = CIA + 0.078 BID
A= Distance of the station from the ocean (m)
B= Elevation of gage above mean sea level (m) x 100
D= Distance of the station from the sound (m)
I%OLOGY OF MARITIME FORESTS OF THE SOUTHERN AILANTIC COAST 27
Highway and parallel cut
A Highway cut only
Predicted - Proffitt 1977
d I I I , I , I
A B c D E F G H
Site location
Fig. 2.11. Percent of salt spray collected at the foredune (A), ocean side of forest (B), leeward edge of the barrier forest (C), ocean
side of the central barrier forest (D), ieeward edge of central barrier forest (E), ocean side of the barrier forest near sound (F),
interior of forest (G), and sound side of barrier forest (I-I) for various representative transects (from Seneca and Broome 1981).
Soil Formation and Mineral
Cycling
Soils of maritime forests are typically one of two gen-eral
types. Forested dune ridges consist of sandy soil,
whereas interdune swale wetlands may contain accumula-tions
of peat. Maritime forest soils tend to be highly
permeable, acidic, deficient in plant nutrients, and poorly
developed because of their secondary origin from well-leached
ocean sediments, geologically recent origin, and
relatively high regional precipitation.
An orderly process of soil formation and stabilization
on maritime dunes was described by Chapman (1976).
Newly formed sand dunes progress through four stages as
they develop from “embryo dunes” to “yellow dunes,”
then “gray dunes,” and finally “mature vegetated dunes.”
Embryo dunes are formed when sand is freshly depos-ited
on an accreting beach, when migrating dunes reform
following destabilization, or when fresh sand is swept
from the beach to form a berm along the leading edge of a
maritime forest on an eroding beach. Initially, the embryo
dune is devoid of vegetation, its soil is undeveloped, and
no soil profile is apparent. Given sufficient time, however,
sea oats and other grasses and herbaceous plants may
become established. This vegetative cover helps retain
nutrients, soil moisture, and dune stability. After a vegeta-tive
cover develops, the dune is called a yellow dune.
Yellow dunes also lack a distinctive soil profile.
Koske and Polson (1984) found that the phosphate
concentration in yellow dune soils on Rhode Island was
typically two orders of magnitude lower than in agricul-tural
soils. Under the condition of low phosphorus avail-ability,
a phosphate deficit zone develops around the roots
of grasses and other plants. Root hairs are apparently
unable to bridge this gap unaided; however, plants of
American beachgrass (Ammophila breviligulata) infected
with the zygomycetan mycorrhizal fungus Gigaspora sp.
are able to grow very well. Laboratory studies demon-strated
that this and other vesicular-arbuscular mycorrhi-zae
assist in phosphorus uptake and appear necessary for
significant growth of dune grasses. Fungal mycelia also
serve to bind sand grains together and help retain soil
moisture.
American beachgrass (Ammophila breviligulata),
waxmyrtle (Myrica cerifera), and beach pea (Luthyrus
japonicus) are common plants in the yellow dune zone; all
are associated with nitrogen-fixing bacteria (Godfrey
1976a). Nitrogen fixation by endosymbiotic bacteria is
28 BIOLDGICALREP~RT 30
probably a major source of nitrogen on barrier islands.
Haines (1976) reported that the amount of nitrogen deliv-ered
annually to the Georgia coast by rainfall was about
0.3 g/m2, an amount well below the calculated require-ments
of coastal plants.
Development of a soil microflora enhances nitrogen
and phosphorus availability (Koske and Polson 1984). As
these essential plant nutrients accumulate in the dune
ecosystem, growth by woody species is promoted and
organic matter begins to accumulate in the soil, giving it a
gray color. This is the “gray stage” in dune development.
Shrubs and dwarf trees dominate the vegetative cover of
gray dunes (Chapman 1976). If they remain stable long
enough, gray dunes may mature into maritime forest. Art
et al. (1974) reported that on Fire Island, New York, forest
can form on siliceous sands within 200-300 years.
As vegetative cover increases on mature forested
dunes, a soil profile develops as organic acids are leached
downward. The uppermost soil horizon is the litter or duff
layer and consists primarily of dead leaves, twigs, and
other plant materials. Beneath the litter, the soil is ashy
white because most of the humic substances have been
leached into the sand to a depth of several centimeters,
where they accumulate to form a tan or orange layer.
Because moving sands have buried soils repeatedly, often
a series of soil horizons can be seen in the exposed face of
eroding dunes, demonstrating the instability of some bar-rier
islands (Koske and Polson 1984).
Although the mycorrhizae and endosymbiotic nitro-gen-
fixing bacteria of the soil microfloraplay an important
role in the process of dune stabilization by stimulating
vegetative cover (Koske and Polson 1984), comparable
studies of the microflora of mature maritime forest soils
are lacking. The mycorrhizae (Gigaspara sp.) that pro-mote
phosphate uptake in beach grass do seem to have
specific host requirements and are associated with several
tree species, including oaks (Koske and Polson 1984).
Waxmyrtle with its nitrogen-fixing bacteria is a common
component of maritime forests. It is therefore highly prob-able
that these microflora play an important role in the
cycling of phosphorus and nitrogen in mature maritime
forest soils, as well as during soil development.
The pattern of mineral cycling on barrier islands is quite
different from the pattern in forests that cover rocky soils
(Art et al. 1974). In most mainland forests, minerals lost
in runoff are replaced by weathering and decomposition
of the soil’s parent rock. Mineral-deficient quartz sand is
the primary parent material of barrier island soils. Mari-time
forest soils have low water-holding capacity and low
cation-exchange capacity. Soluble minerals released into
the soil are transported quickly downward into the ground
water unless intercepted by organic matter, fungal myce-lia,
or plant rootlets near the soil surface. Most maritime
forest plants have their roots concentrated in the upper 30
cm of the soil (Art et al. 1974). At any given time, most of
the minerals in a maritime forest are contained in the form
of living or dead biomass. Continued survival of the eco-system
may depend on the ability of the microflora inhab-iting
the rhizosphere to sort rapidly and retain such criti-cally
important plant nutrients as phosphorus and nitrogen,
while simultaneously allowing potentially toxic levels of
chloride to pass into the ground water for dilution and
dispersal.
If barrier island soils are inherently deficient in miner-als,
then where did the minerals now contained in the
biomass come from? Possible sources include excrement
from migratory birds, transfer from estuarine sources by
animals that graze in the salt marsh but seek shelter (and
defecate) on high portions of the island, wind transport of
ocean-derived detritus (dry sea wrack) into the dune sys-tem,
and atmospheric inputs.
Art et al. (1974) attempted to measure the meteorological
contribution of cations to themaritime Sunken Forest on Fire
Island, New York. Although the Fire Island maritime forest
is composed predominantly of deciduous species of trees
and is therefore floristically quite different from typical
maritime forests of the Southeast, there are enough similari-ties
in soil origin and growth form of the forest to consider
this work the best model for understanding cation cycling in
a maritime forest. Art et al. (1974:6 1) concluded that cation
sources other than meteorological were insignificant and
that the Fire Island ecosystem was “nearing a steady state
[in which] meteorological inputs balance losses to ground-water.”
This pattern of nutrient cycling was similar to that
inferred for some tropical moist forests. Both forest types
have highly weathered soils, low mineral input from weath-ering,
and large proportions of their cations held in living
biomass. Both depend on rapid circulation of nutrients be-tween
soil and biomass.
Interactions between meteorological inputs of nutrients
and primary production apparently are instrumental in the
development and maintenance of the forest cover on the Fire
Island dunes. Vegetation is the major interceptor of mete-orological
nutrient inputs to the ecosystem. Living vegeta-tion,
litter, and humus constitute the major sink for nutrients.
Thus, a potential positive feedback system develops in
which increases in vegetative biomass result in greater cap-ture
and retention of minerals from the atmosphere, thereby
producing still greater biomass. The growth-stimulating po-tential
of increased nutrients is countered by the growth-re-tarding
effects of toxic salt aerosols. Maritime forest-growth
response at any given location or time would seem to result
from the ambient tension between these two contrasting
effects of salt aerosols.
The forest canopy on Fire Island is dominated by several
deciduous species, such as sassafras (Sassafras albidum)
ECOLOGY OF M.MWME FORESTS OF THE SOUTHERN ATLANTIC COAST 29
and shadbush (Amelanchier canadensis), as well as the
evergreen American holly (Zlex upaca) (Art et al. 1974). The
deciduous species lose all their leaves over a short period in
the fall; mineral recycling then begins and continues in
spring. In contrast, the canopy of southern maritime forests
tends to be dominated by evergreen species. The Fire Island
climate might be described as mild and temperate, while that
of the Southeast coast is hot and humid. The warmer, wetter
southern climate provides an extended season during which
rapid decomposition and mineral cycling can occur. Monk
(1966a) noted that evergreen species tend to lose their leaves
continuously rather than seasonally. The litter from ever-green
trees tends to be tough, waxy, and aromatic and thus
moderately to strongly resistant to decomposition through
insect milling followed by fungal decay. This vegetative
adaptation so commonly found in the southern maritime
forest may help to ensure a continuous, albeit low, supply of
mineral nutrients.
Two general systems of mineral uptake in relation to
tree growth form were described by Hillestad et al. (1975).
Live oaks have a shallow, spreading root system about
equal in diameter to that of the crown. The crown serves
as a high-surface-area collector of meteorologically de-rived
nutrients that are diverted by rainfall directly into a
dense, shallow root zone. Art et al. (1974) reported that the
salt-aerosol-sculpted canopy at Fire Island exhibited an
extremely large ratio (9.5: 1) of branch-to-canopy surface
area. In contrast to live oak, pines have sparse, shallow root
systems but deep taproots. This growth form leads to a
large root surface area in contact with a large section of
the soil profile, allowing pines to scavenge nutrients that
percolate through the groundwater. Pine canopies tend to
be more sensitive to salt-aerosol damage than those of
oaks. Because oaks are more resistant to salt damage, they
can better exploit minerals carried with the salt aerosol,
whereas pines are better adapted to exploit soil nutrients
at sites protected from salt aerosol. Both life forms and
their associated nutrient-capture systems reduce nutrient
losses over the entire forest gradient.
Cation retention is affected by soil-water acidity (God-frey
1976a). Maritime forest soils provided with calcium
or magnesium tend to be less acid and probably retain
mineral nutrients longer than soils in which calcium and
magnesium cations are in lower concentration. Important
sources of calcium and magnesium for maritime forests
are the carbonates (aragonite) from mollusk shell frag-ments
and other biogenic carbonates carried by the wind
from the beach. Available cations increase in a southerly
direction along the Atlantic coast as the proportion of
limestone-derived carbonates in beach sand increases
(Godfrey 1976a).
The second major soil type in maritime forests is peat
or sandy peat (Brown 1983; Bumey and Bumey 1987).
Feat soils accumulate in interdune swales when the swales
arc intercepted by the freshwater table or flooded by
brackish water from the estuary. Swale ponds are initially
temporary bodies of water. Freshwater ponds become
seasonal and finally permanent as rising sea level pushes
the freshwater lens higher. Eventually, any trees in the
swale may be killed by flooding. Organic matter (leaves,
branches, stumps) tends to collect in these low, wet depres-sions
between forested dunes. Pond sediments are often
very anaerobic and charged with hydrogen sulfide, result-ing
in reduced oxidative decomposition. Pond sediments
typically consist of unconsolidated, coarse woody debris
and leaves at the surface. Humification of this material
produces a fine-grained, sticky black mud. Beneath this
are coarse wood fragments and an indurated surface that
represents remains of a soil profile predating pond forma-tion.
Beneath this layer, the soil consists of fairly clean
sand. Cores drawn from the peat and sandy peat sediments
of freshwater ponds have yielded pollen and microfossil
evidence from which pond origins and recent vegetative
events in the surrounding maritime forest can be recon-structed
(Brown 1983; Bumey and Bumey 1987).
Hydrology
The hydrological regime on barrier islands is distinctive
(Fig. 2.12) (Art et al. 1974). Precipitation provides the
only natural source of fresh water. Typically, the barrier is
underlain by permeable sediment containing salt water.
Under these conditions, fresh water tends to float as a lens
over the underlying salt water. Under ideal geological
conditions, the freshwater lens can be modeled by using
the Ghyben-Herzberg lens principle (Ward 1975). This
predicts that for every meter of free water table above
mean sea level, there will be 40 m of fresh water in the lens
above the saltwater aquifer. The freshwater lens can be
quite deep below elevated ridges on the barrier but short-ens
abruptly to zero depth at the island and saltwater
interface (Fig. 2.13).
The sea islands of Georgia receive an average annual
precipitation of 1,308 mm (Table 2.1), an amount that
appears to be typical for barrier islands of the Southeast.
Floyd (1979) estimated that a major portion of the average
annual precipitation of 1,143 mm at Nags Head, North
Carolina, was lost through evaporation, runoff, and dis-charge
of ground water to the ocean or bay by lateral
movement. Only 25% of the precipitation was available
for percolation into the zone of saturation where it could
become part of the groundwater supply.
Water in the freshwater lens is usually very low in
dissolved salts, considering the periodic pulses of salt
aerosol delivered to the vegetative cover (Proffitt 1977).
Apparently, excess salt is rapidly diluted by precipitation
30 BIOLOGICALREP~RT 30
c3
Condensation
Evapotranspiration
-c- 0~ -1 ~ -1 Precipitation
Evapotranspiration
Seawater intrusion
~,,1,1,,.,,~,“~\,,,~,,1...,1,\,~,,.~.,.,~,.,.,,~,~~~~~~,,,,~,.,.~,,,,~,.,,1~~.,..,.,~.,,,,~,.,~I,*..,.,.~II~.~.~ . ~. .~ .. ~. . . . . . . . . . . . . . . . . . . . . . . . ..r...r....~......,,,,,,,,,,;,,,,;,,~,,,,~,,,,,,,,,
~~~,,\\\,,.~,,~,1~~\111\~~.1..~~~..~.~..~~~.~.~~.~.~ \,*1,1\.~.~.~.~.\.,\~~~.,~,~.~~~~~~,,~,~,.~..,~,,~ .,~.,,,\,\,~,\,,,.,~, .~r .,.,.,1,.~.,,,,,,,,,~..,\,1,,,,...~.,~,,,,1...,\~,,~,,11. ,1,,1,1..~,,~l~s~,~~,,.l.,.,.~, Shallowartesian aqui,_fe,,r., (,,..,Il,\.,~~.,L,,1I1>.~1,,,...,.,,.1~,,,~,,.,,.1,.,1~,,,1,,.,.,.,.,,~.,,,\,,~~,.,,\,,\,1,1*,,,1,1>1.\,,~,,,.,,~.<\,\1,.,,..~,~I.,~.,,.~,,~1,.1.,1~.1.~1,,,~1~1.1.,
and flushed from the system. Precipitation entering the soil
near the interior of the watershed is rapidly drawn down-ward
to the bottom of the freshwater lens. Counter-flow
along the contact with salt water brings excess fresh water
back to the surface, where it seeps into the bay or ocean
(Fig. 2.13).
Rapid dispersal of salt below the root zone was demon-strated
by Proflitt (1977). He buried 2.3 kg of rock salt just
below the litter layer on a forested slope on Bogue Banks,
North Carolina, and measured chloride concentrations in
the soil a& various depths and distances from the salt burial
site for a period of 3 months. Soil salt content in the root
zone (O-30 cm deep) at the source remained about two
orders of magnitude above the background level. Lateral
salt transport near the surface was minimal, since chloride
concentrations were never found above background in the
root zone at monitor stations 0.6 m from the salt burial site.
After 47 days, the chloride concentration had returned to
near background in the root zone at the salt burial site, and
movement of chloride was mainly downward, concentra-tions
exceeding background by one order of magnitude at
greater depths.
Evapotranspiration rates are unknown for maritime
forest. Is surface moisture lost more rapidly from unvege-tared
sandy soils or from forested dunes? How effective
are the various canopy surface patterns in absorbing and
holding precipitation? What are the cumulative effects of
destroying maritime forest while simultaneously pumping
water from the freshwater lens to serve the needs of barrier
island development? Given the high permeability and low
cation-exchange capacity of barrier island soils, what is
Fig. 2.12. Hydro-logic
cycle of a
typical Holocene
barrier island
(from Missimer
1976; used with
permission of
SCiWX?).
Fig. 2.13. Idealized diagrammatic cross section of a barrier
island, showing water-flow pattern in the freshwater lens
(from Art et al. 1974).
~OLOGYOF~%WTIME~ORESTSOFTHESOUTHERNATLANI~CCOAST 31
their potential for becoming contaminated by septic tank
seepage? What effect does septic tank seepage or disposal
of wastewater by spraying have on soil microflora and
mineral cycling?
On most developed barrier islands, the remaining mari-time
forest is the primary watershed and source of public
water supply. At what point does transfer, by pumping of
groundwater to the surface, speed up saltwater intrusion?
Excess pumping and the cutting of canals and marinas
along the freshwater lens and saltwater margin may lead
to loss of hydrostatic head in the freshwater lens and thus
result in saltwater intrusion at the groundwater surface
(Winner 1975; Ward 1975).
The potential interrelation between surface groundwa-ter
and maritime forest cover on barrier islands appear to
be numerous, but information about them is scarce. Further
research on the role of vegetation in influencing the hy-drology
of barrier islands is needed.
Wetlands
Several types of wetland habitat may be associated with
maritime forest. Wetlands are usually associated with
topographically low areas between dunes and form when
the groundwater table rises and intercepts low-lying soils.
Temporary rain pools formed in this manner may develop
into semipermanent freshwater ponds. Shallow ponds sup-port
growths of willow (Salti spp.), gums (Nyss~ sylvuticu
and Liquidambar styraciflua), ash (Fraxinus americana),
or other wetland trees and thus resemble the deciduous
hardwood swamps found on the mainland. Deeper ponds
support submersed vegetation. These kinds of freshwater
wetland are often called maritime forest swamps or swale
ponds. Both types are a “water table window” (Bensink
and Burton 1975).
Ponds of brackish water are formed when the ends of
swale ponds are captured by an expanding salt marsh,
along closed ocean inlets, or by tidal flooding (Bensick and
Burton 1975). Long, narrow brackish ponds of the first
type may grade into narrow “finger salt marshes” toward
their lower ends. Larger, more open brackish ponds are
often called “salt ponds.” Odum and Harvey (1988) clas-sified
these pond types, using the wetland classification
system of Cowardin et al. (1979), as palustrine emergent,
palustrine shrub/shrub, palustrine forested, estuarine
emergent, and estuarine shrub/shrub.
Bumey and Bumey (1984) reported palynological evi-dence
illustrating the pattern of development of freshwater
ponds at Nags Head Woods, North Carolina. Radiocarbon
dating of the oldest organic sediments in ponds indicated
a recent origin of less than 400 years ago. The pollen
percentages at all levels exhibited a near-constant back-ground
of the same species of flowering trees and shrubs
that inhabit the area now. Pollen from bottomland trees and
shrubs increased steadily, whereas pine pollen declined
from bottom to top in sediment cores. This palynological
pattern is consistent with the hypothesis that the hydric
forests that now occupy dune swales developed quite
recently from upland forest in response to a rising water
table. During the initial stage in wetland development,
waterlogged soil was colonized by fast-growing herba-ceous
plants such as Mexican tea (Chenopodium am-brosioides)
and false nettle (Boehmeriu cylindricu). These
wetland plants were replaced by freshwater aquatics such
as species of Typha, Nymphaea, Myriophyllum, Lem-naceae,
Utricularia, and Potamogeton.
Water quality of maritime forest ponds is variable, even
among ponds near one another (Kling 1986). Variability
in water-quality characteristics among ponds is probably
related to the fact that at any given location the various
ponds are usually in differing stages of development.
Ponds vary in hydroperiod, solar exposure, and degree of
exposure to direct inputs of atmospheric salts. Based on a
comparison of ion ratios, Kling (1986) concluded that the
water in the Nags Head Woods ponds more closely resem-bled
that of the local groundwater than diluted seawater or
typical river water in the region. A. Cole (North Carolina
State University, personal communication) confirmed low
salinity and absence of water chemistry variability of
freshwater ponds of similar age and origin in the Buxton
Woods, North Carolina.
Freshwater ponds in maritime forest were described by
Odum and Harvey (1988) as generally having slightly
higher ionic concentrations than typical inland freshwater
ponds (Table 2.2). Interdunal ponds tend to be circumneu-tral
in pH and poorly buffered. When dense populations of
aquatic vegetation deplete the water of bicarbonate
through intensive photosynthesis on bright days, pH can
increase to about 9.0; when decaying vegetation releases
organic acids into the water, pH can decline to about 4.5.
Fresh ponds typically do not exhibit excessive amounts of
nitrogen or phosphorus and are not normally described as
eutrophic. Anaerobic conditions may exist in the peaty
sediments of the ponds throughout the year and may
extend to the pond bottom during warm weather.
Freshwater ponds often provide the only dependable
source of water for animals on barrier islands. The associ-ated
freshwater wetlands also expand habitat diversity.
Major groups of animals such as frogs, salamanders, water
snakes, turtles, aquatic birds, and aquatic mammals are
largely excluded from barrier islands without freshwater
ponds. When such ponds are present, however, many of
these animals provide a varied and more dependable food
source for nonaquatic inhabitants. Hillestad et al. (1975)
described an oscillating pattern of predator-prey relation-ships
related to perturbations in the wetland communities
32 BIOLDGICAL REPORT 30
Table 2.2. Mean values of physical-chemical parameters
for five freshwater ponds in the Nags Head Woods,
North Carolina (modified from Kling 1986).
Parametera Mean Range
PH 6.8 6.2-7.2
Conductivity(pS/c m) 207.6 112.0-381 .O
HC03- 46.6 13.8-81.4
Cl- 26.3 21.4-38.0
so4-- 3.6 0.02-6.3
Na+ 18.0 12.0-35.3
Ca++ 10.9 2.9-19.5
Mg++ 4.3 2.40-7.44
K+ 1.7 1.1-3.3
Nt.bW~~ 3.9 0.0-8.0
Nfb-Wgn) 13.7 9.2-18.5
H2P04--P(&b) 31.4 5-l-80.6
5.5 3.6-8.3
Secchi(cm) 54.0 40.0-70.0
02 6.6 2.5-8.3
aMilligrams per liter unless indicated otherwise.
on Cumberland Island, Georgia. When the water table is
high, certain prey species such as frogs, insects, and mos-quito
fish are provided with ample food and breeding
habitat and thus, predation pressure tends to be relatively
low. When the water table falls and water levels are low,
the prey animals concentrate in shallow water, and the
habitat advantage shifts in favor of predators such as
snakes, herons, and alligators. Prey species are again fa-vored
when the water level falls below the bottom of the
ponds. Then, predators are reduced in number or tempo-rarily
eliminated, while prey species find refuge in alliga-tor
holes or crayfish burrows or under damp vegetation.
The abundance of prey populations quickly increases with
the return of higher water levels.
Some observers (Mayes andList 1988) indicatedconcern
over possible ecosystem-damaging effects of periodic
drought conditions on maritime fresh ponds, whereas others
(Hillestad et al. 1975) suggested that water-table oscillations
may be necessary to maintain these pulse-stabilized aquatic
systems. Without perturbations such as drought and tire,
shallow-water wetlands would rapidly till with organic mat-ter
and develop toward a shrub or swamp forest. When the
shallow bottom is exposed to the atmosphere and solar
drying, aerobic decomposition is accelerated, releasing nu-trients
that can later support wet-season productivity.
The biota of freshwater ponds in Nags Head Woods,
North Carolina, was inventoried by a multidisciplinary
team of researchers. Their surveys were carried out during
a drought phase in the local climate and served to assess
the ability of the pond biota to survive drought.
The algal flora of the Nags Head Woods ponds was
dominated by desmids, euglenoids, and periphytic diatoms
(Bellis 1988). Seventy-two algal taxa representing the
seven major algal groups normally present in fresh water
were reported from rather few collections. The ecological
significance of the algae in these ponds is as yet poorly
understood; however, several nitrogen-fixing cyanobacte-ria
such as Nostoc commune and Anabaena azollae, an
endosymbiont of the mosquito fern (Azolla carolinianu),
were among the most frequent algae in several ponds.
Periphytic diatoms in the ponds included Pinnularia
braunii, P. latevittata var. domingensis, Gomphonema
gracile, and Eunotia curvata. These taxa were described
by Patrick and Reimer (1966) as indicators of waters with
low dissolved mineral content and relatively low pH. A
variety of euglenoid taxa (Euglena, Trachelomonq
Phacus) occurred abundantly among the often-anaerobic
organic debris.
The algal flora of the Nags Head Woods freshwater
ponds was dominated by motile unicells (Bellis 1988).
Algae exhibiting this morphology typically form very
resistant cysts or spores when environmental freshwater
ponds consisted of taxa that commonly occur in similar
environments on the mainland and seemed adapted for
survival during episodic droughts.
Vascular plants in the Nags Head Woods ponds consisted
of 40 aquatic or emergent taxa and 3 wetland shrub taxa
(Davison 1988a). Other vascular taxa associated with pond
margins included 22 taxa of ferns, herbs, shrubs, and trees.
Pond water levels were extremely low during the vas-cularplant
survey (Davison 1988a). Differences in species
composition and diversity among ponds strongly corre-lated
with differences in pond size and depth gradient.
Despite individual differences among ponds, certain gen-eral
patterns were evident. Where pond margins were
exposed to sunlight, they were invaded by opportunistic
seedlings. In several ponds, the open water surface was
completely replaced by a vegetated “quaking bog.” Wet-land
species growing on exposed pond bottoms and along
pond margins shaded by forest canopy included false
nettle (Boehmeria cylindrica) and lizard’s tail (Saururus
cernuus). Vascular plant opportunists in fully exposed
areas were dominated by graminoids (Leersia oryzoides,
Eleocharis baldwinii), Polygonum spp., and pennywort
(Hydrocotyle ranuncutoides). Deeper portions of the
ponds were reduced to small pools of open water during
the drought; they were completely covered by floating
aquatics, dominated by duckweeds (Spirodefla polyrhiza,
Lemna spp., Woljfia columbiana), mosquito fern (Azolla
caroliniana), and frog’s bit (Limnobium spongia).
Prolonged lowering of the water level permitted estab-lishment
of saplings of loblolly pine (Pinus taeda), redbay
(Persea borbonia), Carolina willow (Safix caroliniana),
black gum (Nyssa sylvatica), and red maple (Acer rubrum)
(Davison 1988a). The latter three species can survive
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 33
seasonal flooding and may become established around the
pond margins after the water table returns to normal.
The microinvertebrate fauna of the Nags Head Woods
ponds was surveyed by MacPherson (1988), who re-ported
70 taxa from a few collections limited to a single
season (spring). Taxonomic richness was greatest among
Diptera (18 taxa), Coleoptera (15 taxa), and Odonata (11
taxa). Amphipods, isopods, and a clam (Sphaerium) rep-resented
the most abundant microinvertebrates present.
Most of the microinvertebrates were associated with
mats of floating or emergent vegetation, a habitat also
dominated by dragonflies and beetles. Benthic microin-vertebrates
were less abundant and included clams,
leeches, and worms. Surveys of aquatic and wetland
vertebrates in the Nags Head Woods ponds included fish
(Schwartz 1983, 1988), amphibians and reptiles
(Braswell 1988), birds (Cooper 1988), and mammals
(Webster 1988).
Schwartz (1983, 1988) proposed multiple possible
origins of fish that now inhabit ponds along the northern
Outer Banks of North Carolina. Marine fish may be
transported into ponds from the ocean or from the estuary
during overwash events. Freshwater fish remain on the
islands in ponds formed from relict river channels, or
they may gain access by overwash transport from estu-aries
that became much less saline in recent times. The
fish with a saltwater affinity are generally absent from
ponds in the Nags Head Woods; this may be related to
the apparent lack of recent washover.
Schwartz conducted intensive fish surveys on the
Nags Head Woods ponds in 1983 and again in 1987. Five
fish species were found in each survey; however, only
three of the species in the second survey were the same
as those reported in the first survey. Fish reported in both
surveys were mosquito fish (Gambusia holbrooki),
bluegill (Lepomis macrochirus), and largemouth bass
(Micropterus salmoides). Species reported in only one of
the surveys were golden shiner (Notemigonus crysoleu-cas),
black crappie (Porno.& nigromaculatus), rainwater
fish (Lucania parva), and pumpkin seed (Lepomis gib-bosus).
Schwartz suggested that the species missing in
the second survey (rainwater fish and pumpkin seed)
may have been extirpated through predation by the large-mouth
bass or other carnivores. He believed the new
residents were recent accidental introductions.
A survey of amphibians and reptiles of the Nags Head
Woods by Braswell (1988) indicated that availability of
freshwater habitat resulted in greater herpetofaunal diver-sity.
The herpetofauna associated with the Nags Head
Woods ponds appeared to be the most diverse of any
barrier island of the Atlantic Coast. Of the 41 species
reported, 23 species were directly dependent on the fresh-water
pond habitat.
Freshwater ponds enhance habitat quality for some ver-tebrates
as well. A listing of breeding birds near Nags Head
Woods (Cooper 1988) showed the greatest species richness
in the pine-dominated forest. Greatest abundance of birds,
however, was found in a gum swamp (interdunal forested
swale) along the margins of a series of fresh ponds.
Webster (1988) reported that mammalian diversity was
greater on Currituck-Boclie Island (including the Nags Head
Woods) than on any other barrier island in North Carolina
or adjoining states. Mammalian diversity was somewhat less
in the Nags Head Woods itself than in the larger area.
Webster (1988) attributed this reduced mammalian richness
to a more limited range of habitats in the Nags Head Woods.
The freshwater ponds were frequented by muskrats, rac-coons,
otters, deer, and bats (Webster 1988).
Fire
Anthropogenic and natural fires have been reported on
barrier islands from early in the European colonial period
until the present. Since the land-clearing and hunting prac-tices
of the aboriginal inhabitants of the islands involved
the use of fire, it is probably safe to assume that barrier
island biotic communities have been influenced by tires
caused by humans throughout much of their presumed
approximately 5,000-year existence.
In recent history, residents of barrier islands have used
fire to improve grazing land, remove unwanted vegetation,
maintain open vistas, create wildlife habitat, and eliminate
unwanted insects and snakes. The use of fire for these and
related purposes is deeply ingrained in the southern agri-cultural
tradition (Davison 1983; Turner 1985; Bratton
1985, 1986a; Turner and Bratton 1987; Bratton and Dav-ison
1987).
Natural fires initiated by dry lightning do not occur with
uniform frequency along the southeastern coast but seem
to have a gradient of increasing frequency from north to
south (author’s observation). Summer thunderstorms oc-cur
almost daily along the coasts of Georgia and Florida
and dry lightning is common. Staff at national wildlife
refuges at Canaveral and Merritt Island, Florida, recorded
some of the highest lightning frequencies in the United
States. In contrast, Cape Hatteras National Seashore expe-riences
fewer summer thunderstorms and virtually no
lightning-initiated fires (Bratton 1986a).
In forested areas, fire intensity varies with the litter
deposition pattern (Williamson and Black 1981). Early
seral plants such as pines and shrubs may be inferior
long-term competitors; however, these plants exhibit fire
tolerance and even fire facilitation, characteristics that
may give them a short-term advantage in environments
where fires occur fairly frequently. Williamson and Black
(1981) measured the air temperature at various distances
34 BI~LOGICALREIQRT~O
above the litter layer in burning forests of several types and
discovered that fires in pine forests consistently produced
a higher temperature at any given level above the ground
than fires in a live oak forest. In the seedling zone and up
as high as 0.5 m above the soil, the temperature in a live
oak stand averaged about 175” C, whereas the temperature
in the pine forest at the same level averaged about 290” C.
Williamson and Black (1981) concluded that maximum
temperatures of fires were high enough under pines to
eliminate the otherwise competitively superior oaks in
areas near mature pines.
Davison (1983) noted repeated fires on a portion of
Cumberland Island, Georgia; she suggested that this pine-dominated
woodland is maintained by fires of natural
origin. The nutrient-poor soils of the site prevent the oaks
from growing fast enough to form a dominant canopy
before the conjunction of the climatic conditions and fuel
accumulation result in fire. The significance of fire as a
disturbance that maintains vegetative cover on Cumber-land
Island has since been questioned by McPherson
(1988:1), who concluded from studies of the shrub-forest
and marsh-forest boundaries that “succession to oak-pal-metto
(Quercus spp. and Serenoa repens) forest is control-led
by soil moisture.” Fire played only a minor role in
community dynamics.
Davison (1983) reported that recovery of maritime
forest on Cumberland Island, Georgia, during the year
after an intense fire in 1981 did not involve a change in
species composition. No new species appeared after the
fire and none was lost. Only the apparent age distribution
of individuals was altered by the fire.
Oak forests and pine forests differ in the way in which
they carry a fire (Davison 1986; Davison and Bratton
1986; author’s observations). Closed-canopy maritime
oak forests tend to have a dense evergreen canopy with
sparse understory and herbaceous vegetation. Shading
also promotes moisture retention in the litter layer. Un-der
these conditions, fires tend to be smoldering ground
fires; crown fires are rare. Fires often originate outside
the oak forest and enter it from adjacent marsh or pine
forest.
Pine-dominated forests are usually drier and provide a
better quality fuel that allows intense and fast-moving
fires. Along the coasts of Georgia and Florida, a dense
understory of palmetto beneath short, scattered pines pro-motes
intense crown fires.
Canopy trees of the maritime forest appear to be well
adapted to fire. Live oak is protected from fire by its
thick, ridged bark, while cabbage palm is protected by
its sheathing leaf bases. The terminal bud of cabbage
palm is surrounded by woody, flame-resistant leaf peti-ales.
Aboveground portions of understory trees and
shrubs such as dwarf palmetto (Subal minor), waxmyrtle
(Myrica cerifera), American holly (IIex opaca), sparkle-berry
(Vaccinium arboreum), and redbay (Persea bor-bonia)
we less resistant to fire, but all have underground
or surface structures from which burned individuals re-generate
sprouts. Loss of aboveground portions of these
plants through fire stimulates hormonal release of latent
buds. Rapid regrowth and recovery follow as the sprouts
use nutrients stored in underground roots and rhizomes
in an open environment temporarily freed from intense
competition for solar energy. Thus, it is clear that fire has
been an impartant factor in organizing forest cover pat-terns
on barrier islands since long before the present.
ECOLOGY OF MARITIME FORESTS OF THE SOUTHERN ATLANTIC COAST 35
Flora of Maritime Forests
36 BIOLOGICAL RFPORT~~
Introduction
Extensive information about the vascular flora of the
Atlantic Coast barrier islands has been amassed since the
beginning of this century. Typically, this information was
presented as taxonomic lists of species in particular sites.
Woody trees, shrubs, and vines ate usuahy dominant. Un-derstory
shrubs and herbs are sparse in the shade provided
by the dense evergreen canopy characteristic of most mari-time
forests. Many floristic studies also described vegetation
zonation, and some authorities attempted to relate the sev-eral
distinctive vegetative cover zones into a successional
sequence. Quantitative studies of plant community structure
and function are extremely rare and simply not available for
most barrier island and coastal dune forests.
Latitudinal Gradient in Floristic
Composition
The maritime forests of the south Atlantic Coast of the
United States do not seem to represent a single well-defined
plant community that can be described by characteristic
taxa. The available floristic data suggest that maritime for-ests
actually consist of a northern and a southern forest
assemblage that overlap to produce a diversity maximum at
about 35” N latitude along the North Carolina coast
(author’s observation). Evidence in support of this concept
was derived by the author by expanding and modifying a list
of the range limits of common barrier island phtnt species
along the Atlantic Coast (Art 1971). Floristic lists were
included in the analysis that were not available to Art in
1971, especially for southern barrier islands. The resulting
data (Fig. 3.1) summarize floristic lists from 40 different
reports covering 32 barrier-island forest locations between
25” N (southern Florida) and 42” N latitude (Cape Cod,
Massachusetts). This summary contains no weighting for
relative abundance of the local flora. Fifty taxaof commonly
encountered barrier-island forest trees and shrubs are listed.
The species are arranged in order of their first southerly
occurrence along a south-to-north line.
Of the 50 taxa listed, only red maple (Acer rubrum)
occurs throughout the range of the survey. Some taxa listed
toward the top of Table 3.1 (live oak [Quercus virginianal,
palmetto palm [Sabalpalmetto], laurel oak [Quercus lauri-folia
= Quercus hemisphaerica], &bay [Perseu borbonial,
etc.) seem to represent a southern assemblage. Taxa listed
near the bottom (white oak [Quercus albal, bayberry
[Myrica pennsylvunica], pitch pine [Pinus rigida], beach
plum [Prunes marifimfz], etc.) seem to represent a northern
assemblage. Finally, there is a large assemblage of plants
distributed widely along the central Atlantic Coast. This
group includes water oak (Quercu m&a), loblolly pine
(Pinus rue&), yaupon (1lex vomitoria), American beauty-beny
(Callicarpa americana), toothache tree (Zanthoxylwn
clava-herds), and sweetgum (Liquidambar styraciflua).
Figure 3.2 is a plot showing the number of taxa (of the
50 listed) occurring at l-degree intervals of latitude. The
pattern is one of maximum species diversity near the middle
(35” N) of the geographic region. A dendrogram based on
Jaccard’s Index of Similarity (Fig. 3.3) also supports the
concept of overlapping plant assemblages. The pattern is one
of greatest similarity among locations between Georgia
(31” N) and North Carolina (36’ N). A second cluster of
similar vegetative locations extends northward from Vir-ginia
(37’ N to 42’ N). The third cluster (25” N to 30” N)
consists entirely of Florida records, and is least similar to all
other locations.
The maritime forests of the southeastern Atlantic states
(Virginia to Florida) consist of a discontinuous chain of
forests that is seldom more than 1.5 km wide but nearly
1,600 km long. This narrow island chain extends generally
along a north-south axis from a subtropic to a temperate
climate. Much of the gradual variation in floristic composi-tion
along the major axis can be accounted for as repre-senting
a differential response of individual species to the
climatic gradient.
Zonation
On a local scale, the floristic expression of a maritime
forest appears to be influenced by another smaller scale
gradient, the proximity to direct ocean influence. Although
relative exposure to salt is generally considered the major
factor controlling zonation, the actual process may be more
complex and involve interactions between water supply,
nutrient cycling, sand blasting, sand migration, storm expo-sure,
and other factors.
The earliest botanical descriptions of barrier islands
(Johnson 1 !NO; Kearney 1900; Coker 1905) consisted pri-marily
of botanical inventories that were made during brief
visits to particular islands. These early botanists were in-trigued
by the conspicuous zonation of vegetative commu-nities.
The following description of the Isle of Palms, South
Carolina, by Coker (1905: 136) is typical:
The island is about four and one-half miles long and
one mile across at its broadest part. The time at my
disposal being limited, I did not attempt to study the
entire island, but confined myself to the western half.
Within this small area, however, there is as great a
diversity of ecological conditions as is generally
found over a much more extended region, From the
few struggling and half-buried halophytes of the
beach one may pass over the dunes with their palms,
then across a narrow marshy strip and into a dense
forest of oaks and pines, with trees over forty feet in
height-_and all within a distance of three hundred yards.
38 BIOLOGICALREFORT 30
Table 3.1. Characteristic plant communities of the barrier islands of the southeastern United States (Oosting 1954).a
I. Sand Strand Vegetation
1. Treeless (open)
a_ Inner Beach-Croton punctatus, Cenchrus tribuloides, (Cakile ea’entula, Spartina patens, Physalis maritima)
b. Outer Beach-Salsola kali, Euphorbia polygonifolia (Fimbristylis castanea, Spartina patens)
c. Dune Beach-Uniola paniculata (Strophostyles helvofa, Oenothera humtjka, and any of others from inner or outer beach)
2. Trees and Shrubs (closed)
a. Thicket-Alex vomitoria (Myrica cerifera, Juniperus virginiana)
b. Thicket Woodland-Persea borbonia (and forma pubescens) (Juniperus virginiana, many lianas including Ampelopsis
arborea, Parthenocissus quinquefolia,Vitis spp., Smiiax spp., Toxicodendron radicans, numerous ericaceous shrubs,
especially Vaccinium arboreum)
c. Woodland-Quercus virginiana (Carpinus caroliniana, Ilex opaca, Mow rubra, Quercus laurifoha, Bumelia lycioides,
Zanthoxylum clava-herculis, Osmanthus americanus)
II. Marsh Vegetation
1. Salt Marsh-Spartina alterntjlora, Salicomia virginica, (Suaeda linearis, Borrichia frutescens, Spergularia marina,
Limonium carolinianum, Distichlis spicata, Kosteletzkya virginica)
2. Creek Marsh-Juncus roemerianus
3. Dune Marsh-various species
4. Tidal Flat-Scirpus americanus, Paspalum distichum (Fimbristylis castanea, Spartina patens)
*List is of plant communities. Community dominants are listed first. Taxon names in parentheses indicate common asoeiates.
Coker (1905) completed his observations with detailed
descriptions of vegetative cover types, identified as upper
beach, dune, fresh marsh, forest, hammock, salt flat, and salt
marsh.
Almost half a century later, Oosting (1954) summa-rized
the information about the vegetative cover along
maritime strands in the southeastern United States. Oost-ing
revised the earlier list of vegetative cover associa-tions
for Ocracoke Island, North Carolina by Kearney
(1900) by expanding it to include “those species repeat-edly
found elsewhere along the Atlantic coast from New
England to Florida in similar zones” (Oosting 1954:230).
This list (Table 3.1) assigned names to characteristic
plant communities on barrier islands. Different names
have been assigned to these communities, (see Chapter
1) but most would probably agree with the species group-ings
presented by Oosting (1954).
Although the relative abundance of particular plant
species may vary from site to site on the barrier islands
along the Atlantic Coast, the same fundamental life
zones occur in essentially the same arrangement at each
site. From ocean to estuary, these zones are ocean beach,
dunes, maritime forest, and salt marsh. Shackleford
Banks, North Carolina, is a specific example of this
zonation pattern on forested and unforested por-tions
(Fig. 3.4). Godfrey (1976b) produced a generalized
Degrees north latitude
F’ii. 3.2. Number of taxa (of the 50 shown in Fig. 3.1)
occurring at l-degree latitude intervals from 25”N to
45”N.
fi m
F; 70-
?
2
E 60-
8
aki
50-
I - L----l- -
Fig. 3.3. Taxonomic similarity (Jaccard’s
In&x) for assemblages of maritime for-est
trees arid shrubs.
transect diagram of the physiographic and ecological consist of drowned relict portions of mainland ridges or
zones of a typical barrier island (Fig. 3.5). In the same resulted from accretion along stable shorelines. The
study, Godfrey also observed that the proportion of a more northerly barriers seem to be affected to a greater
typical barrier island covered by forest increases be- extent by the destabilizing effects of ocean washover and
tween New England and the Sea Islands of Georgia (Fig. dune migration. The characteristic taxa in a given zone
3.6). He attributed this increase to geologic diffcrenccs along a northeastern barrier are typically replaced by a
in barrier island origin and processes. The southern bar- visually similar but floristically dissimilar assemblage
rier islands (sea islands) are more stable because they along a southeastern barrier (Fig. 3.7).
L_~. 2_0_0_ -m_----_-J
Fore
Inner
beach
Back
Sound
marsh forest
Fore
dune
Back
Sound
marsh
Fig. 3.4. Transect diagrams showing genera lized physiography of forested and unforested portions Of Shackleford *ds, North
Carolina (From Au 1969).
40 BIOLOGICAL REPORT 30
Fig. 3.5. The basic physiographic and ecological zones of a typical barrier island (the diagram indicates the zonation on typical barrier
beaches, and does not imply that every barrier resembles the drawing).
1. Northern coast barrier
\
2. Central and southern coast barrier
\
3. An accreting barrier
4. The “sea island” type of seacoast barrier
Fig. 3.6. Typical barrier island profiles found
along the east coast of the United States.
1) A northern coast barrier where dune
building is more significant than over-wash.
Well-developed dune lines exist
close to the beach, and are often scarped
if the beach is retreating. The barrier is
made up of dunes on top of earlier over-wash
deposits. Where enough protection
exists, it is vegetated by dune grasses,
shrubs, and woodlands. 2) A central and
southern coast overwash barrier. Regular
overwashes create a broad, generally
sloping barrier that is made up primarily
of overwash strata and terraces with
dunes scattered<