Biological Services Program FWc.jOE^-^1 /37 FWS/OBS-81/37 March 1982 THE ECOLOGY OF BOTTOMLAND HARDWOOD SWAMPS OF THE SOUTHEAST: A Community Profile Fish atxJ Wildlife Service U.S. Department of the Interior The Biological Services Program was established within the U.S. Fish and Wildlife Service to supply scientific information and methodologies on key environmental issues that impact fish and wildlife resources and their supporting ecosystems. The mission of the program is as fonows: • To strengthen the Fish and VJildlife Service in its role as a primary source of information on national fish and wild- life resources, particularly in respect to environmental impact assessment. • To gather, analyze, and present information that will aid decisionmakers in the identification and resolution of problems associated with major changes in land and water use. • To provide better ecological information and evaluation for Department of the Interior development programs, such as those relating to energy development. Information developed by the Biological Services Program is intended for use in the planning and decisionmaking process to prevent or minimize the impact of development on fish and wildlife. Research activities and technical assistance services are based on an analysis of the issues, a determination of the decisionmakers involved and their information needs, and an evaluation of the state of the art to identify information gaps and to determine priorities. This is a strategy that will ensure that the products produced and disseminated are timely and useful. Projects have been initiated in the following areas: coal extraction and conversion; power plants; geothermal , mineral and oil shale develop- ment; water resource analysis, including stream alterations and western water allocation; coastal ecosystems and Outer Continental Shelf develop- ment; and systems inventory, including National Wetland Inventory, habitat classification and analysis, and information transfer. The Biological Services Program consists of the Office of Biological Services in Washington, D.C., which is responsible for overall planning and management; National Teams, which provide the Program's central scientific and technical expertise and arrange for contracting biological services studies with states, universities, consulting firms, and others; Regional Staffs, who provide a link to problems at the operating level; and staffs at certain Fish and Wildlife Service research facilities, who conduct in-house research studies. Cover design by Graham Golden FWS/OBS-81/37 March 1982 THE ECOLOGY OF BOTTOMLAND HARDWOOD SWAMPS OF THE SOUTHEAST: A COMMUNITY PROFILE by Charles H. Wharton Institute of Ecology University of Georgia Athens, Georgia 30602 and Wiley M. Kitchens Edward C. Pendleton U.S. Fish and Wildlife Service National Coastal Ecosystems Team 1010 Gause Boulevard Slidell , Louisiana 70458 Timothy W. Sipe Department of Biology Wabash College Crawfordsville, Indiana 47933 Performed for National Coastal Ecosystems Team Biological Services Program Fish and Wildlife Service U.S. Department of the Interior Washington, D.C. 20240 DISCLAIMER The findings in this report are not to be construed as an official U.S. Wildlife Service position unless so designated by other authorized documents. Fish and This report should be cited as follows: Wharton, C.H., W.M. Kitchens, E.C. Pendleton, and T.W. Sipe. 1982. The ecology of bottomland hardwood swamps of the Southeast: a community profile. U.S. Fish and Wildlife Service, Biological Services Program, Washington, D.C. FWS/OBS-81/37. 133 pp. PREFACE This report is one in a series of community profiles whose objective is to synthesize extant literature for specific wetland habitats into definitive, yet handy ecological references. To the extent possible, the geographic scope of this profile is focused on bottomland hardwood swamps occupying the riverine floodplains of the Southeast whose drain- age originates in the Appalachian Moun- tains/Piedmont or Coastal Plain (see study area Figure 1). References are occasion- ally made to studies outside this area, primarily for comparative purposes or to highlight important points. The sections detailing the plant associations and soils in the study area are derived from field investigations conducted specifically for this project. In order to explain the complexities of the ecological relationships that are operating in these bottomland hardwood ecosystems, this report details not only the biology of floodplains but also the geomorphological and hydrological compo- nents and processes that are operating on various scales. These factors, in concert with the biota, dictate both the ecologi- cal structure and function of the bottom- land hardwood ecosystems. We have utilized the ecological zone concept developed by the National Wetlands Technical Council to organize and explain the structural com- plexity of the flora and fauna. The information in this profile will be useful to environmental managers and planners, wetland ecologists, students, and interested laymen concerned with the fate and the ecological nature and value of these ecosystems. The format, style, and level of presentation should make this report adaptable to a variety of uses, ranging from preparation of environmental assessment reports to supplementary or topical reading material for college wet- land ecology courses. The descriptive materials detailing the floristics of these swamps have been cross-referenced to specific site locations and give the report the utility of a field guide hand- book for the interested reader. The senior author wrote the original manuscript and accepts the responsibility for all statements, theories, and figures not credited otherwise. The co-authors extensively revised, reorganized the for- mat, and contributed parts of the manu- script, especially Chapters 3 and 4. Any questions or comments about or requests for this publication should be directed to: Information Transfer Specialist National Coastal Ecosystems U.S. Fish and Wildlife Service NASA/SI idell Computer Complex 1010 Gause Boulevard SI idell, LA 70458 Team m CONTENTS Page PREFACE iii FIGURES vi TABLES X ACKNOWLEDGMENTS xii INTRODUCTION 1 CHAPTER 1. MODERN AND PALEO-GEOMORPHOLOGY OF FLOODPLAINS 4 Introduction 4 Origin and Dynamics of Floodplains 4 Features of Modern Floodplains 8 Paleo-Geomorphology 12 CHAPTER 2. HYDROLOGY 16 Alluvial Rivers 16 Blackwater Rivers 17 Spring-fed Streams 17 Bog and Bog-fed Streams 20 Flooding Duration and Frequency 20 CHAPTER 3. PHYSICOCHEMICAL ENVIRONMENT 21 Chemical Characteristics of Rivers 21 Physicochemical Characteristics of Floodplain Soils 23 CHAPTER 4. FLORA OF BOTTOMLAND HARDWOOD COMMUNITIES 31 Introduction 31 The Anaerobic Gradient 31 Plant Responses to Anoxia-related Stresses 32 Plant Community Patterns in the Floodplain 37 Disturbance and Succession in Bottomland Hardwood Plant Communities .... 76 Primary Productivity of Floodplain Forests 80 CHAPTER 5. FAUNA OF BOTTOMLAND HARDWOOD ZONES 84 Fauna of Zones II and III 84 Fauna of Zone IV 90 Fauna of Zone V 91 The Use of Bottomland Hardwood Zones by Fish 92 Trophic Relationships 95 Summary of Faunal Utilization 100 CHAPTER 6. COUPLING WITH OTHER SYSTEMS 103 Natural Couplings with Headwater Tributaries and Estuaries 103 Coupling with River Deltas 104 Chemical Coupling with the Uplands 104 Modifications of River and Floodplain 105 Chemical Coupling with the Water Table and Atmosphere 107 Coupling Via Faunal Movements 107 Summary 107 REFERENCES 109 APPENDIX: SOILS OF SOUTHEASTERN FLOODPLAINS 127 FIGURES Number Page 1 Major river floodplains and their associated bottomland hardwood cominunities within the Carolinas, Georgia, and Florida 2 2 Landsat image of the floodplain of the Oconee River, GA, showing how large alluvial rivers that drain the Piedmont form extensive tracts of bottomland hardwoods below the fall line 5 3 Point bar and meander formation in floodplains 6 4 Diagram of an idealized alluvial floodplain with various depositional environments 9 5 A meander bend and cross section showing levee and ridge and swale topography so common on modern and relict surfaces 10 6 Aerial photomosaic of the lower Roanoke River indicating flood- plain features characteristic of rivers approaching the coast 11 7 Sea level changes between the Sangamon interglacial period and modern times 13 8 Development of present-day relict and modern floodplain surfaces ... 14 9 Hydrographs of four types of southeastern floodplain rivers and streams 17 10 Hydrographs of an alluvial river showing the possible effects of an increase in floodplain width on water levels, between upstream and downstream 17 11 Two photos showing drydown and inundation of the floodplain in the Congaree Swamp National Monument 18 12 Relation of flood discharge of Oconee River (GA) to distance downstream 19 13 Diagram of the water budget for Creeping Swamp (NC), July 1974- June 1975 19 14 Indication of where Tertiary limestones lie at or near the surface, often giving rise to spring-fed rivers and contributing heavily to the Suwannee-Santa Fe system 19 VI FIGURES (continued) Nunber ^^^^ 15 (A) An idealized sequence of NWTC bottomland hardwood zones from a water body to an upland, along a moisture continuum. (B) One-half of a floodplain from mid-alluvial river to bluff, indicating modification of the idealized sequence by the intrusion of a natural levee between Zones II and III. (C) Further modification of the idealized sequence by inclusion of an abandoned river channel 24 16 Diagrammatic scheme of the relationship of bottomland hardwood zones to soil types on a large alluvial river floodplain (Congaree Swamp National Monument, SC) 27 17 Dense surface mat of minute rootlets 28 18 A remarkable example of multiple-trunked stool ing of Ogeechee tupelo at Sutton's Lake (Apalachicola River, FL) 34 19 Oak displaying buttressing, common among bottomland hardwoods 35 20 A windthrown diamondleaf oak on a small blackwater creek flood- plain illustrates the large diameter of the root crown of bottomland hardwoods 36 21 The correspondence between alluvial floodplain microtopography and forest cover types 38 22 Microtopographic relief on a small blackwater creek floodplain .... 38 23 A liverwort zone, the upper boundary of which indicates that high water has stayed at that level for at least 16% of the year 49 24 The outermost swale along the Roanoke River (NC) 49 25 Ogeechee tupelos on the Apalachicola River measuring 1 m DBH (diameter at breast height) 50 26 Scenic Ebenezer Creek is a unique variant of Zone II 50 27 Drydown in water tupelo on Ebenezer Creek 51 28 The outermost backswamp on an alluvial floodplain 51 29 Tidal forest with characteristic interwoven mat of large roots close to the surface 52 vn FIGURES (continued) Number Page 30 Three shoreline dominants, sweet bay, red cedar, and cabbage palm, characteristic of the tidal zone of an alkaline blackwater river 53 31 A small grove of virgin cypress preserved on Lewis Island 54 32 Large overcup oaks occupy depressions (Zone III) in the dominantly Zone IV floodplain of the Congaree Swamp National Monument 57 33 Sweetgum giants in the Congaree Swamp National Monument 61 34 Many Zone IV bottomland hardwoods on Coastal Plain alluvial river floodplains have an understory of Sabal minor 61 35 The floodplain of the blackwater Canoochee is somewhat anomalous in the co-dominance of the diamondleaf oak with either spruce pine or loblolly pine 62 36 Normally an upland species, the loblolly pine also grows in Zone V areas of many floodplains 65 37 An old levee ridge near Cedar Creek supporting Zone V vegetation ... 66 38 Narrow, long ridges between swales on the lower Roanoke flood- plain (NC) of probable late Pleistocene age have an almost diagrammatic zonation of Zone V hardwoods 67 39 An example of understory species taking advantage of the dry environment of a floating log in order to exist in Zone II 75 40 Cross-sectional transects (aspect is looking downstream) of nine southeastern rivers and floodplains, indicating zones (I-V) and major vegetational and natural features 77 41 A windthrown bitternut hickory on the floodplain of the Murder Creek Special Management Area 79 42 Aerial view of relict rice fields on former bottomland hardwood forests that are presently managed for waterfowl 80 43 The effect of a gradient of flooding on productivity as compared with a regional level that might be expected in the absence of standing or flooding water 81 44 Organic matter production in ecological zones 83 vn 1 FIGURES (continued) Number Page 45 Comparison of the bottomland hardwoods and characteristic fauna at different reaches of a blackwater river near the coast 85 46 Floodplain pools (Zone III) on a Piedmont alluvial floodplain are concentration centers for detrital decomposition and teem with vertebrate life 86 47 Snails of the genera Viyipara are often abundant in shallow aquatic zones (Zones II) above tidal influence 88 48 Six endemic species (and Lampilis splendida) of unionid clams recorded for the Altamaha River, GA 89 49 The crop contents of a wild turkey killed in April in the Arkansas River bottomlands 93 50 Two-way traps with wire mesh wings, set in a small Coastal Plain blackwater creek and in shallow drainways on the floodplain, revealed that 21 fish species, used the floodplain extensively .... 94 51 Largely terrestrial food chains involving detritus, granivory, frugivory, and herbivory 96 52 Largely aquatic detrital food chains 97 53 Common invertebrates of southeastern rivers 99 IX TABLES Number Page 1 Acreages of bottomland hardwoods (oak-guni-cypress and eln- ash-cottonwood) and other forest wetland classes in the south Atlantic States 3 2 Comparative sediment losses and land-use practices 7 3 Changes in levee height in upper, middle, and lower reaches of typical southeastern floodplain rivers 10 4 Physicochemical data summarized for Georgia rivers in water year 1577 22 5 Mean inorganic constituents in selected Georgia coastal plain rivers and in the "world average river" 22 6 Physicochemical characteristics of floodplain soils by National Wetland Technical Council Zones 25 7 Comparison of some floodplain soil nutrients to those of Coastal Plain, Piedmont, and mountain soils 29 8 Percentage of organic matter in selected wetland and upland soils 30 9 Trees, shrubs, vines, and herbs characteristic of southeast- ern bottomlands and the floodplain zones in which they most frequently occur 39 10 Response of mature bottomland hardwoods, some upland species, and some levee species to varying lengths of time of inundation during growing season 43 11 Dominance types of Zone II 45 12 Dominance types of Zone III 56 13 Dominance types of Zone IV 58 14 Dominance types of Zone V 63 15 Forest dominance type distributions across the nine floodplain transects illustrated in Figure 40 69 2 16 Net primary productivity (g dry wt/m /yr) for bottomland hardwood communities, compared with other wetland and upland environments 82 Number TABLES (continued) Page 17 Some environmental factors affecting the fauna of the bottom- land hardwoods and their relative importance in each bottomland hardwood zone '■^^ A-1 Soil analysis for floodplain dominance types, Zone II, alluvial rivers A-2 Soil analysis for floodplain dominance types. Zone II, black- water rivers A-3 Soil analysis for floodplain dominance types. Zone II, tide- influenced sections (tidal forests) of spring-fed blackwater and alluvial rivers ^^^ A-4 Soil analysis for floodplain dominance types. Zone III, alluvial rivers ^^- A-5 Soil analysis for floodplain dominance types, Zone IV, alluvial rivers A-6 Soil analysis for floodplain dominance types. Zone IV, black- water rivers A-7 Soil analysis for floodplain dominance types, spring-fed rivers (Zones II and IV) A-8 Soil analysis for floodplain dominance types. Zone V, alluvial rivers 130 131 131 132 XT ACKNOWLEDGMENTS The authors reserve their most pro- fuse thanks to those individuals (and their organizations) who devoted time to assist in the field: L.L. Gaddy, H. Boyd, A.E. Radford, J.K. Lynch, L. Peacok, J. Moore, R. Sharitz, M.K. Brinson, R. Sniffen, C. Gresham, J. Pinson, H.M. Leit- man, A. Redmond, P. Grover, N. Eichholz, R. Kautz, G. Bass, and N. Young. The special logistical support of the Florida Game and Fresh Water Fish Commission through Grey Bass, Norman Young, and Neal Eichholz is particularly acknowledged. We deeply appreciate the efforts of those who aided in species identification: S. Gilchrist, W. Heard, R. Guard, T. Wilkins, R. Godfrey, University of Georgia Herbarium. Those who supplied other as- sistance were R. Jones, C. Coney, J. Cely, D. Shure, H.L. Ragsdale , R. Sharitz, J. Adams, G. Guill, F. Parrish, and A. Benke. We are grateful to the following who critically read all or portions of the manuscript: Jay Benforado, Robert Huffman, Robert Meade, Helen Leitman, J.M. Lynch, Roger T. Saucier, Arville Touchet, and Eugene Turner. We acknowledge the very special help of Mark Brinson, Steven Forsythe, John Coffin, and Parley Winger. We extend our sincere appreciation to Linda C. Lewis, Gladys V. Russell, and Kathy B. Waggoner of the Institute of Ecology, and to Elizabeth Krebs and Daisy Singleton of the U.S. Fish and Wildlife Service, for their dedication to the many typing provided for the publica- this manuscript. We are also to Kate Benkert and Hal Rienstra, and Wildlife Service, hours of tion of indebted U.S. Fish Finally, for their profound valuable assistance, thanks are expressed to Gaye Farris, U.S. Fish and Wildlife Service, for her extra- ordinary editorial efforts. xn INTRODUCTION Bottomland hardwoods occupy the broad floodplains that flank many of the major rivers of the Southeastern United States as they flow through the Piedmont and Coastal Plain to the sea. These forests and their fauna comprise remarkably pro- ductive riverine communities adapted to a "fluctuating water level ecosystem" (Odum 1969) characterized and maintained by a natural hydrologic regime of alternating annual wet and dry periods. The bottomland hardwood communities support recognizably distinct assemblages of plants and animals that are associated with particular landforms, soils, and hydrologic regimes. The fluctuating hydro- logic regime dictating the ecologic func- tioning of modern floodplains is rela- tively recent, perhaps originating around 18,000 years ago in the late Pleistocene period, when changes toward present strong seasonal climates began (Martin 1980). Many floodplain species are traceable to Tertiary times, and others originated as far back as the Mesozoic. Apparently, rivers and their floodplains have served as refugia for numerous relict life forms which found the dynamic conditions there suitable. Plants such as tupelo gums, and animals such as alligators, turtles, gar, bowfin, sturgeon, and amphibians (Siren) survive essentially unchanged on modern floodplains as relicts from the Age of Dinosaurs. Ironically, in the face of massive land use of surrounding uplands, floodplains today remain some of the last refuges not only for floodplain species but also for upland species. Because the floodplains occupied by bottomland hardwoods are transitional in the aquatic continuum between permanent water and terrestrial uplands, they are elusive to classify. The scheme of Cow- ardin et al. (1979) used here classifies bottomland hardwoods as forested wetlands in palustrine and estuarine ecosystems. Other terms or categories which have been used to classify this community include "seasonally flooded basins or flats" (Shaw and Fredine 1956); "mixed bottomland hard- woods and tupelo-cypress swamps" (Stubbs 1973); "oak-gum-cypress" and "elm-ash-cot- tonwood" (Boyce and Cost 1974); and "deep swamps," "narrow stream margins," and "broad stream margins" (Forest Service Re- source Bulletins 1970, 1972, 1974, 1978). The extent and distribution of bot- tomland hardwoods in the Southeast are indicated in Figure 1 and Table 1. Diverse classification schemes and the inclusion of other categories of forested wetlands make difficult precise calculation of the area! extent of the community; however, acreages appear to be equal in the four States (North Carolina, South Carolina, Georgia, and Florida) focused upon in this profile. The latest U.S. Forest Service Forest Surveys, standardized in 1930, have been used in preparing Table 1, which gives combined acreages of the two forest types occurring in each of the Forest Sur- vey's physiographic classes. Each forest type, oak-gum-cypress or elm-ash-cotton- wood, is dominated singly or in combina- tion by these species (Boyce and Cost 1974). Table 1 also includes the acreage of forested wetlands other than bottomland hardwood floodplains. Cypress and willow strands, where water spreads out and moves downslope through a wide forest of cy- press, are not included as bottomland hardwoods; similarly, bays, pocosins, and cypress ponds are excluded from this com- munity profile. Small drains, defined as poorly drained narrow strands lacking a well defined stream, include many tiny headwater branches and drainways. Although not specifically floodplains, they are certainly important in filtering drainage from the uplands into the larger systems. Their acreage is large but they are ex- cluded from our calculation of bottomland floodplain acreage. Large swamps such as the Okefenokee and Dismal Swamp have been excluded as well. CHOWAN ROANOKE PAMLICO ESCAMBIA YELLOW CHOCTAWHATCHEE CHATTAHOOCHEE APALACHICOLA ST MARKS OCHLOCKONEE SOPCHOPPY WAKULLA WACISSA AUCllLA _,^: if> Figure 1. Major river foodplains and their associated bottomland hardwood comriunities within the Carolinas, Georgia, and Florida. Inset indicates physiographic provinces within the study area. Table 1. Acreages of bottomland hardwoods (oak-gum-cypress and elm-ash- cottonwood) and other forest wetland classes in the south Atlantic States*^ (Data courtesy Noel Cost, Southeastern Forest Experiment Station, U.S. Department of Agriculture Forest Service, Asheville, NC. ) States Forested wetlands classes FL GA NC SC Total bottomland hardwoods on floodplains 1,149,891 1,435,453 1,618,135 1,110,343 Total other 2,160,906 1,623,052 565,942 717,138 Cypress strands 268,988 9,730 3,037 11,635 Cypress ponds 594,857 200,641 2,581 56,298 Bays & wet pocosins 564,359 186,088 221,307 217,650 Willow heads & strands 39,492 8,405 - 1,098 Marl flats & forested prairies 2,743 - 2,745 _ Small drains 589,291 1,154,151 328,722 398,629 Other hydric 101,176 64,037 7,550 31,828 Inventory dates: Florida, 1980; North Carolina, 1974; South Carolina, 1978; Georgia, primarily 1972 but includes 1981 survey of southwest Georgia. This community profile has been pre- pared in part to provide information for management decisions. Like most natural communities, bottomland hardwoods have felt the impact of man. Unfortunately, the absence of uniform treatment of data and the screening of it as indicated above in publications such as Boyce and Cost (1974), Langdon et al. (1981), and Turner et al. (1981) make it difficult to use earlier survey figures to calculate this impact in terms of loss of bottomland hardwoods on southeastern floodplains over time. Losses of bottomland hardwoods in areas outside the specific study region have been severe, none more so than the floodplains of the ^'ississippi River drainage. There conversion of forest to agriculture, primarily soybeans, has re- duced by 60% the areal extent of the hard- wood community; by 1995, only a projected 3.9 million acres will remain intact, down from 11.8 million acres in 1937 (MacDonald et al. 1979). Although losses of flood- plains bottomland hardwoods in the Caro- linas, Georgia, and northern Florida have been much less extensive, few areas in the Southeast have escaped some direct or potential impact of man. Besides conversion to agriculture, another impact on bottomland hardwoods has been conversion to tree-farm monoculture. Numerous examples occur along the flood- plains of the Gcmulgee and Oconee Rivers, GA, where the higher elevated bottomland hardwood communities are logged, the stumps bulldozed into windrows, and the terrain prepared for pine (or other) monoculture. Floodplain rivers have also been subject to severe impacts, including con- struction of impoundments, diversion canals, channelization, dredging, and shortening of channels. These alterations change the hydroperiod and may permanently alter the ecology and functioning of the floodplain. It has not always been recognized that the entire bottomland over which flooding occurs is a functional part of the wetland system and must be considered as a unit when making resource decisions. Because the bottomland hardwoods in the study area still retain their ecological functions and value, environmental manag- ers have the opportunity to consider and weigh management alternatives. This pro- file provides information to aid them in this task. CHAPTER 1 MODERN AND PALEO-GEOMORPHOLOGY OF FLOODPLAINS INTRODUCTION The complexities of hydroperiod, cli- mate, soils, and watershed characteristics have produced an often bewildering mosaic of vegetative zones and associations in the bottomland hardwood community. To understand better the complex relationship between hydroperiods and the bottomland hardwood community, one must first con- sider the geomorphology of the floodplain itself. The biota of the floodplain cannot be wisely interpreted or managed, nor can the impact of man's modifications be cor- rectly evaluated without understanding watershed-dependent floodplain hydrology and geomorphology. The biota alone pro- vide too narrow a viewpoint. The energy of flowing water and the sediment load of river flows are ulti- mately responsible for the geomorphic landforms on southeastern floodplains. This chapter discusses processes of water and sediment distribution on floodplains and landforms characteristic of both modern and ancient environments. ORIGIN AND DYNAMICS OF FLOODPLAINS The flows and sediments carried by a river are responsible for the origin, character, and maintenance of the flood- plains and their forest cover. The gently sloping coastal plains of the Southeastern United States provide an ideal environ- mental setting for floodplain formation. Erosional and depositional processes cul- minate in a sinuous river channel within a broad flat plain bounded by valley walls or bluffs--the floodplain (Figure 2). Striking examples (Figure 2) of floodplain formation occur along rivers in the study area at the fall line, the abrupt transition between the Piedmont and the Coastal Plain. The excess energy of river flows in passage over the bedrocks and red clays of the Piedmont begins to dissipate at the fall line, where the river first encounters the easily erodible sedimentary strata of the Coastal Plain. This dissipation results in deposition of alluvium (sands, silts, and clays) which in turn is reworked by riverine processes into meanders. The residual energy of the flowing water is expended by this lateral meandering which serves to widen the river valley. Rivers adjust their slopes by meandering until they reach a nearly steady-state, with sediment load balanced with water velocity and volume. As the floodplains widen out, more sediments can drop out from overbank deposition. Remark- ably broad, flat floodplains are the result of these processes. Mountain and Piedmont rivers, on the other hand, al- though they form floodplains when topogra- phy and soils permit, still retain much potential channel energy. First order variables that determine the behavior of water and sediments in- clude climate, geology, soils, land-use, and vegetation (Morisawa and LaFlure 1979). Variables that describe the chan- nel are velocity, slope, flow-depth, plan-form (shape from an aerial view- point), and width (Gregory 1977). Other variables are meander length and meander belt width (Blench 1972), discharge (vol- ume of water/unit time), and roughness of river bed (bottom presence of trees, cobbles, dunes, etc.) (Leopold and Wolman 1957). These variables are extremely interdependent. The dominant depositional/erosional processes on floodplains are: (1) point bar deposition, (2) overbank deposition, and (3) sheet or gully erosion (scour) and redeposition on floodplain surfaces in a sequence of floods (Sigafoos 1964). The point bar is built on convex banks of streams or river meanders by lat- eral accretion (Figure 3). Since deposi- Figure 2. Landsat imaae of the floodplain of the Oconee River, GA, showing how large alluvial rivers that drain the Piedmont form extensive tracts of bottomland hardwoods below the fall line, which in this photo runs diagonally from lower left to upper right. Milledgevil le, GA, is marked M. Photo courtesy Georgia Department of Natural Resources. tion on the convex bank keeps pace with erosion of the opposite concave bank, the bulk of the sediment remains stored in the floodplain (Leopold and Wolman 1957). Though much slower than point bar formation, vertical accretion by overbank deposition also builds most southeastern floodplain surfaces. Overbank deposition results from high water losing its veloc- ity and dropping sediments as it traverses the floodplain, usually by sheetflow or overflow channels. The amount of sediment deposited can vary widely. For example, a single flood in the Atchafalaya River Basin, LA, caused sediment accumulations of up to 46 cm (18 inches) over portions of this vast flood plain. Accumulations ranging from 0.3 cm (.125 inches) to 3.8 cm (1.5 inches) have been documented for floods in the Potomac River Basin, VA (Sigafoos 1964). Average deposition, how- ever, ranges between 0.3 m (1 ft) and 0.6 m (2 ft) in 200 to 400 years (Wolman and Leopold 1957). The bed of the chan- nel, as well as the surface of the flood- plain, accumulates sediment deposits dur- ing floods. Channels are also created and maintained by these overbank flows and accommodate the excess discharge of flood waters. On forested floodplains, local ero- sion by overbank flows may produce rapid recycling and overturn of deposited sedi- ments. Surface erosion or scour is fol- lowed by deposition of comparable magni- tude, and the floo'dplain becomes a spatial (1) INITIAL UNIFORM CHANNEL DIRECTION OF FLOW - (2) m <234 -^ gg^ POOL TO POOL - 5 TIMES WIDTH POOL TO POOL - 5 TIMES WIDTH ^=--RIFFLE CUT BANK Figure 3. Point bar and meander formation in floodplains. Unstable stream flow in uniform river channels (1) forms pools and riffles (2). A meandering channel (3) develops and eventually exhibits erosion on the concave banks of meanders as well as deposition and point bar formation on the convex banks (4). (After Muller and Ober- lander 1978, courtesy of Random House, Inc.) mosaic of erosional and depositional sur- faces superimposed over material deposited earlier by point bar accretion (Sigafoos 1964). The features resulting from these processes are detailed in the followinc se.ctions. Sediment Sediment source is provided by con- tinual erosion of the landscape throughout geological time. In the Piedmont and mountains, this source is igneous and metamorphic rocks (granites, schists, and gneisses) which decay or weather under the influence of rainwater. Released compo- nents are sands, silts, and clays, which are transported by sheet-wash or gully- wash into streams. In the Piedmont this decay produces a soil horizon (vertical layer) of saprolite (decomposed rock) up to 9.2 m (30 ft) thick. The weathered Piedmont saprolite little by little washes downstream. Pied- mont sands are transported as bedload rolling along the bottom of stream chan- nels, whereas silts and clays are carried as suspended matter in the water column. Silts and clays, the principal components of surface floodplain soils, settle out or form overbank deposits during floods. A reduction of particle size downstream v\h)/ result from the weathering of silt and clay particles that remain in place for long periods of time between episodes of downstream transport (Curry 1972). Land uses in uplands profoundly af- fect the quantities of sediment entering a stream. Before the 19C0's, sediment inputs to southeastern river systems were minimal according to some observers. River pilots recalled that as late as 1912 the Tennes- see River was relatively clear even after heavy rains (Ellis 1936). According to one report, the turbid Altamaha River of coastal Georgia was once a relatively clear stream, and as late as the 1840's it was possible to determine on which tribu- tary (Ocmulgee or Oconee) rains were fall- ing since much of the Oconee drainage was in agriculture (Lyell 1849). Sediirent losses from forested uplands are usually modest (2.5 cm or 1 inch/ 16,000 years; Soil Conservation Service 1977). However, the losses after the forest is removed can be quite substantial (Table 2). When forest cover was reduced from 80% to 20% in the Potomac Basin, sediment loading increased eight times (Patrick 1972). The Soil Conservation Service (1977) reported losses from crop- lands (some of which were once forested floodpTain) of 38.4 tons/acre/year of top- soil to the Obion-Forked Deer River (TN). It is difficult to distinguish wheth- er sediments are derived from natural or culturally accelerated sources (Strahler 1956). Several investigators have at- tempted to estimate losses from the up- lands by measuring the thickness of sedi- ment layers in coastal floodplains. Soils surveys indicate a loss of 15.2 cm (6 inches) of topsoil from the South Carolina Piedmont in 150 years. Between 1910 and 1934, one Georgia Piedmont watershed lost 218 tons/km2/year, but by 1974 this rate was reduced by 86% (30 tons/km^/year) (Meade and Trimble 1974). Happ (1945) concluded an average Piedmont upland soil loss of 9.4 cm (3.7 inches) since earliest settlement. Although agriculture has heavily accelerated the loss of soil from uplands, 90% of the sediment from accelerated Pied- mont erosion remains on hill slopes and in stream bottoms (Trimble 1979). In fact, the composition of alluvial sediments and their rate of deposition in some flood- plains do not reflect a marked change in rate due to agriculture. In South Caro- lina, both Coastal Plain rivers (black- water) and rivers originating in the Piedmont (alluvial) have three terraces (Pleistocene) above the present floodplain that are similar in type of landform, slope, particle size, and composition of sediments to those of the present (Holo- cene) floodplain (Thorn 1967). For example, the quartz sands of the present point bars of the Little Pee Dee River (a blackwater stream) are similar to those of the higher terraces. In the Great Pee Dee River (a Piedmont, or alluvial, stream), the three older terraces also have the same composi- tion of silts and sands as does the pres- ent floodplain. Water and sediment supply are not continuous but result from discrete cli- matic events (Harvey et al. 1979). The largest portion of the total load of many rivers is carried by high flows on the average of once or twice a year. As flow variability increases and as size of watershed decreases, a larger percentage of sediments is carried by less frequent flows. In many basins 90% of sediment is moved during floods recurring at least once every 5 years (Wolman and Miller 1960). Piedmont streams carry 10 times the sediment of Coastal Plain streams at the same discharge rate during floods (Meade 1976). Slope and Meandering River systems are remarkably dynamic. Changes of slope (elevational gradient) which cause rivers to flow can be due to (1) crustal uplifting or downwarping responses of the coast to the removal of the Pleistocene ice mass, (2) scour or erosion which steepens headwaters, or (3) Table 2. Comparative sediment losses and land-use practices (Happ et al. 1940). Land use practices Sediment loss (tons lost/acre/year) Oak forest Bermuda grass Cotton (contour plowed) Cotton (down slope plowed) Barren abandoned field 0.05 0.19 69.33 195.10 159.70 deposition which flattens downstrean reaches. Deposition by tributaries often increases river slope in the region imine- diately downstream of their junction with the main river. This increase in slope results frotn elevation of the river chan- nel bed due to sediment accretion at the juncture of the two streams. Basic features of swamp rivers and streams are their sinuous meanders (Figure 3). Meandering is one way the river accom- modates slope. Meanders lengthen the path of the water, adjusting the energy of the flow to a uniform rate of energy loss per unit of stream length (Leopold and Lang- bein 1966). In other words, the flow rate is stabilized as the slope of the stream is made more uniform and less steep by increasing the distance that the water traverses in its vertical descent from the Piedmont to the coast. The process can be compared to a skier whose rate of descent is slowed by meandering: the consistency and resistance of the snow determine the meander path which he follows for his particular weight. The path of the water bends as uniformly as possible (conforming to a sine-generated curve), minimizing bank erosion and the expenditure of energy (Leopold and Langbein 1S66). After the water rounds the meander curve, it flows straight until centrifugal and inertial forces are diminished to the point where gravity can turn the water back downs lope and the process is repeated. The greater the volume, velocity, or density (water containing sediment weighs more than water without sediment), the greater the force and the further the river travels before it is turned back to form the next meander loop. It follows that meander length and radius are closely related to the width of the river. The width of meanders is a function of water volume, velocity, and density. The development of meanders occurs at bankful (flood) stages (Thorne and Lewin 1979). Meanders occur at fairly consistent intervals of 7 to 15 times the width of the channel (Dury 1977). River slope adjusts naturally to the velocity required to transport the load of water and sediment supplied by the drain- age basin. As slope increases, a stream must cut down (degrade) or develop a sinu- ous course. A meandering reach is stabler than a straight reach (Yang and Song 1979) because it more closely approaches uni- formity in the rate of work over the various riverbed irregularities than does the straight channel (Leopold and Langbein 1966). In addition, the meandering river and its floodplain apparently present the most efficient geometry to accommodate the mean and extremes of flow variability that have occurred throughout its history. The types of sediment through which meanders pass are important. In a river system, meanders will occur where bank material is comparatively uniform. Mean- ders move slowly if banks are cohesive (silt or clay) and rapidly if banks are easily erodible (sands, silty sands). Depending on the type of sediment load, meander wavelength can vary ten-fold at a given discharge (Schumm 1969; Thorne and Lewin 1979). FEATURES OF MODERN FLOODPLAINS Although floodplains appear flat and featureless from the air, in reality they are characterized by diverse topographic features as a result of the continual, dynamic reworking of their sediments by rivers. The low topographic relief of the floodplain landscape is deceptive; a mat- ter of inches in elevation may produce quite distinct ecological zones (see Chap- ter 4). The following sections detail the origin and the dynamic nature of the major geomorphic features of the floodplain (Figure 4). Channels The river channel processes, as dis- cussed earlier, create and maintain the floodplain. As a precursor to developing meanders, basically unstable flows in the stream create a series of pools (scoured areas) and riffles (areas of redeposition) through erosion of the stream bank. Ero- sive forces continue to act in meanders. Scouring occurs on the concave bank of a meander; conversely, scoured material is deposited on the opposite convex bank (Figure 3). Good basic references on channel geomiorphology are Leopold et al. (1964), Allen (1965) and Schumm (1971) as well as overviews by Thorne (1977) and Winger (1981). The stream channel morphology (width, depth, slope, and meander characteristics) 8 Figure 4. Diagram of an idealized allu- vial floodplain with various depositional environments. RC = river channel; K = direction of meander movement; L = natural levee; P = point bar deposits (alternating ridge and swale topography); B = back- swamps; C = channel fill deposits (former ox-bow lake); R = ridge (former natural levee around adandoned channel); OC = overflow channel; S = swale deposits. depends on long-term patterns of flow (Blench 1972). High flow regimes are principally responsible for the formation of channels, while low flows are responsi- ble for only minor adjustments in channel morphology (Keller 1977). The process of channel formation is greatest at bankful stage. Therefore, the annual to biennial flood intervals are more important than periodic catastrophic flooding (Wolman and Miller 1968). When flows exceed the capacity of the river channel, the entire floodplain becomes the channel, and addi- tional physical factors come into play. Since current velocities are a func- tion of the slope of the water surface, it is not surprising that water velocities over the floodplain during overbank flows are comparable to the mean velocities of natural channels and transport sands and silts (Wolman and Leopold 1957). In fact, during overbank flows when the waters leave the meandering channel, the flood- plain surface becomes the high water channel with the current directed straight downslope (valley-axial direction) short- ening the path of flow and increasing the slope of the water surface (Carlston 1965). Only the structure of the intact forest impedes the ravaging flov/s and pre- vents catastrophic scouring of the flood- plain surface and valley walls. Under certain conditions, specialized channels are formed. Termed braided and anastomosing channels, they are character- ized by the main river channel dividing into numerous interconnected channels. Braiding results from a change in grade or slope so abrupt that coarse sediment, usually sand, is precipitously deposited. Braiding, however, can occur at any point in a stream where large deposits of coarse sediments occur. For example, large amounts of sand brought down by a channel- ized reach of a tributary (Flat Branch) of the Alcovy River (GA) have been deposited on the main stream floodplain, causing the main flow to braid. Braiding may also occur at river confluences (e.g.. Little Pee Dee and Lumber River, SC). Braiding is often a temporary phenomenon. When a river divides and rejoins on a vegetated floodplain and the channel configurations are relatively unchanging, it is better termed an anastomosing stream. Both Four Hole Swamp (SC) and parts of the Chipola River (FL) are examples of anastomosing channels. natural Levees During periods of overbank flow, as waters spread out over the floodplain, water currents abruptly slacken, and sus- pended sands and silt are deposited as a levee Immediately adjacent and parallel to the channel . Natural levees (Figures 4-6) are best developed on concave stream banks. They occur along straight reaches, they are usually higher on one on the other. Some large rivers mile-wide levees; however, the river in the Southeast has levees 30 m (98 ft) and ICO m (328 ft) also may although side than may have average between wide. Natural levees slope gently to flood basins and backswamps. The height Table 3. Changes in levee height in upper, middle, and lower reaches of typical southeastern floodplain rivers. Figure 5. A meander bend and cross sec- tion showing levee (L) and ridge (R) and swale(S) topography so common on modern and relict surfaces. Pioneer and suc- cessional plant species anchor the newly formed sandy ridges leading to even more deposition at each high water episode. Normal helicoidal currents (C) conduct sediments up the bai- slope. of levees diminishes as rivers approach the coast because the stream's energy to move sand also decreases downstream (Table 3 and Figure 6). Some deeply incised streams with headwaters arising from clay- rich soils (Tallahala River, MS) have barely discernible levees. Kany black- water streams have unimpressive levees, apparently due to lack of sufficient gradient in the outer Coastal Plain. A breach (or crevasse) in the levee may pro- duce an alluvial fan-shaped feature termed a crevasse-splay deposit which spreads out over the floodplain (Allen 1965). Floodbasins, Flats and Backswamps The term floodbasin specifically applies to vast underfitted floodplains (floodplains developed under a signifi- cantly higher flow regime than at present) where channel meanders may occupy only a portion, or belt, of the floodplain width. Along southeastern rivers that are not markedly underfitted, the floodplain between the natural levees and high valley wall is generally called ambiguously a "backswamp" or more succinctly a "flat" Levee height (ft) Rivers Upper Middle Lower Roanoke (NC) Great Pee Dee (SC) Apalachicola (FL) 13.0 6.0 5.6 6.0 3.0 4.5 0 <2 where elevational relief is limited to shallow depression basins and almost imperceptible rises. The term backswamp also may be applied specifically to peat- forming environments occupying relict channels along the outer rim of the flood- plain. Floodbasin, flat, and backswamp sedi- ments are composed of fine silt and clay particles. Acid backswamps are environ- ments where deposition is minimal and are constantly wet, having water tables at or near the surface. Most of the floodplain, however, dries out annually. The fine clays tend to dry and crack in polygonal patterns, allowing oxygen to enter. Depos- its in these areas vary with frequency of flooding, proximity to the channel, sedi- ment load, flow velocity, and substrate texture. In many eastern floodplains devoid of significant relief, overbank deposits may be coarse and layered. Coarse layers represent the rise to maximum stage of an individual flood; the alternating fine layers represent recession of flow (Allen 1965). Point Bars and Ridge and Swale Topography Host deposition occurs along the main channel of the swamp stream. Materials are eroded from concave sides of channel meanders and redeposited on convex bends to form point bars (Figure 3). Small ridges formed on the point bar by deposi- tion of bed load material during floods form a temporary natural levee on the con- vex side of meanders. The crests of these ridges may stand higher than natural lev- ees on the concave side. As the river bed moves laterally and downstream (Figure 5), d series of ridges' forms with intervening 10 1- •— 1 +-> i_ C3 •— < O a) cn .c oj - — " — • s_ > — - CD > 4- ^ o o S- OJ XI J <-> ' - ■^ -^ "^ ^ J= s- ■r- s_ c 3 +-> +-> S_ E (/5 O C_3 o > o ^^9. c c o O) I ^ ° t/l o "i^ "> ^^ s- -o .^ >, 10 "^ ist S- OJ >— o) I— aj +J D. > ID OJ 2 O) ■- ■!-> H- ro OJ O I — I/) O O +J fO X (O i- 'r— *^ E S- SB I 5s" (/> 0) (J S- 3 O I/) 0) ItJ--^-— ■ o ■a •I- CLcn s- •>- U_ ro Lf) "O T3 •o ° <= s C 3 ro O s_ 10 t/l <_) ■•- ai -o S- I— c ro S- ITJ 4-> ro 3 E ^ XJ 0) u -.- J3 C S_ .— >, I— ■< —1 11 depressions or swales. Vegetation quickly invades and stabilizes each point bar ridge, encouraging further deposition. Ridges are composed priiriarily of sands. Silts and clays are deposited mainly in swales, forming a sticky clay subsoil (gley), sometimes called "blue mud." Mate- rial eroded from the concave side of one meander loop is deposited on the convex side of the next downstream meander. The floodplain is thus "reworked" to the depth of the deepest part of the channel. Sedi- ments are resuspended by powerful bottom- flowing crosscurrents (Figure 5). Meanders migrate because of the constant erosion (undercutting) and/or slumping of the con- cave bank laterally and downslope. Meander migration is slow (<3 m or <10 ft/yr) in small southeastern rivers, particularly those with forested banks. On the other hand, meanders in India's huge Kosi River moved 750 m (2,460 ft) in 1 year (Wolman and Leopold 1957). Dune Deposits Aeolian dunes form when strong winds blow exposed sand from point bars or other sources onto the floodplain. Dunes 13.7 m (45 ft) high sometimes are formed by the deflation (wind removal) of point bar sands and other bare areas of the flood- plain (Allen 1965). Several linear series of large dunes occurring on the east side of the Altamaha River (GA) floodplain are of probable aeolian origin (Bozeman 1964). So extensive are these dunes that the pro- posed Big Morter-Snuffbcx Project (Soil Conservation Service) recommended that they be artifically joined to create a huge levee to block off part of the flood- plain and divert water from eventually flowing into certain tidal river distribu- taries. Aeolian dunes and those associated with the relict braided stream channels (e.g.. Little Pee Dee floodplain, SC; Thorn 1967) probably were formed by gale-force Pleistocene winds blowing across the unvegetated part of the floodplain from the southwest. Dune chains are more likely to be formed where discharge varies widely and the floodplain is not heavily vege- tated (Allen 1965). Discharge is thought to have varied miuch more during the Pleis- tocene when strong seasonality developed. Scour Channels, HuniP'Ocks and "Mini-Basins" Scour channels, hummocks, and mini- basins are microtopographic features producing only slight elevational and drainage changes; however, their effect on plant species distribution is often marked. Scour channels are small v/aterways within the floodplain generally formed during high water as flows seek shortcuts: for example, cuts or chutes across bends, or tributary connections to the main channel. A high percentage of sand is present in the scour channels (and on the adjacent floodplain as well) because scour channels are areas where sheet flow may carry a substantial bed load of sand across the floodplain flats. Hummocks are small "islands" left after years of erosion by scour channel currents. Usually the curved channels in hummock terrains are weblike, weaving around the bases of trees which rriay be "stooled," often bearing ferns and shrubs on swollen bases. The top of hummocks may bear trees characteristically found in areas of higher elevation, although in som.e cases trees such as tupelo gums and cypress form hummocks themselves. Minibasins are shallow depressions that sometimes occur between tree bases. Some may be created by swirling water; others are of ambiguous origin. They are frequently filled with rainwater. Any detritus trapped in them is rapidly decom- posed by frequent fluctuations between dry and moist conditions. This is in contrast to areas around the drier, raised tree bases where detrital accumulations tend to increase floodplain floor elevations. In addition, rriuch of the aerobic-anaerobic nutrient cycling is accomplished in rain- filled minibasins (Wharton and Brinson 1979b) (see Chapter 3). PALEO-GEOMORPHOLOGY It is now recognized that Pleistocene ice age climates and hydrology strongly influenced both terrestrial and aquatic landforms. Glacial and interglacial peri- ods during the Pleistocene produced dra- matic changes in climate, precipitation, and sea level. Increased precipitation and more intense, frost action during glacial advances caused considerable down- slope movement and subsequent transloca- 12 tion and deposition of surface materials (Whitehead and Barghoorn 1S62). The dimin- ished relief and aggraded (filled in) valleys of today's Piedmont are evidence of the tremendous impact of the lateral migration of soils during that era (Eargle 1940). Many common Piedmont soils are underlain by organic deposits as much as 3.7 m (12 ft) thick representing downslope transport and deposition; in one study in South Carolina more than 50% of the sur- face was underlain by these soils (Eargle 1940). Pollen dating of the soils indi- cated they were deposited more than 35,000 years ago, possibly during the waning of the first of the two periods of Wisconsin glaciation, the so-called Altonian sub- stage. Some floodplains in the Southeast apparently were affected by the climatic changes associated with continental gla- ciation. One striking feature reflecting these past climatic regimes is the dramat- ic discrepancy between the size of the floodplain and the size of the present- day river. Today many streams are too small (in terms of discharge volume and meander dimensions) to have produced such wide floodplains. Such streams are de- scribed as "underfitted" (Dury 1977). This phenomenon is common in alluvial rivers and may occur in coastal blackwater streams (Wharton 1977). A growing body of evidence indicates that the geomorphology of underfitted stream floodplains can be explained by the sequence of different hydrologic regimes resulting from prehis- toric climates (Fisk 1947, 1951; Schumm 1971; Dury 1977; Froehlich et al. 1977; Mycielska-Dowgiallo 1977). Floodplain width is a function of sediment deposition and redistribution by meandering during periods of greatest stream discharge, coupled with periods of relatively high sea level. Increased dis- charge over that of the present was prob- ably due to increased precipitation. An- cient flow regimes can be determined through studies of ancient paleochannels in present floodplains (Schumm 1971; Dury 1977; and others). Dury calculated, from ratios of former to present channel bed- widths and meander wavelengths, that dis- charge 12,000 years ago was 18 times greater than that at present. Sediment delivery rates were three times those of today. This increased- discharge was at least in part due to a pluvial (rainy) period of much greater rainfall occurring 18,000 to 10,000 years ago (Thorn 1967; W.G. Kclntire, Louisiana State University Center for Wetland Resources, Baton Rouge; personal communications). Discharge for many streams subsided about 10,000 years ago. Runoff decreased to one-seventh of its magnitude 2,000 years earlier (Dury 1977). Streams and rivers began to assume their underfit characteristics at this time. Climatic changes, coupled with the more subtle influences of change in grad- ient brought about by lowered sea levels (Figure 7) or tectonic rebound of the land, formed another characteristic geo- morphic feature of southeastern flood- plains — the floodplain terrace. Increased flow volume or, in some cases, an in- creased gradient, changed the hydrologic regime and created a new floodplain sur- face, often lower than the old one. De- creased flow volume or increased sediment sometimes reversed the sequence, filling the floodplain back up with new sedi- ments. In any event, steplike terraces resulted, many of which are remnants of prehistoric surfaces. This sequence of alternating high (degrading) flows and lower (aggrading) flows is diagrammed in Figures 7 and S. Because precipitation generally has declined into modern times, THOUSANDS OF YEARS BEFORE PRESENT 60 40 20 5 100 g o 500 Figure 7. Sea level changes between the Sangamon interglacial period (S) and modern times (M) covering two periods of Wisconsin glaciation, the Altonian (AS) and the Woodfordian substages (WS), and a warmer interglacial period, the Farmdalian (F). Periods of entrenchment (E) occurred during glacial buildup. (A) represents periods of alluviation when alluvial river valleys were filled with sediments. (Mod- ified from Saucier and Fleetwood 1970.) 13 Figure 8. Development of present-day relict and modern floodplain surfaces. (1) Sanga- mon (S) interglacial stage. (2) Entrenchment (scouring) during waxing Altonian sub- stage. (3) Farmdalian (F) interglacial with substage alluviation filling in the former entrenched valley. (4) Entrenchment during waxing Woodfordian substage. Remaining Farmdalian (F) deposits are also known as Deweyville (or Terrace I). (5) Post-Wood- fordian alluviation forming modern day Holocene floodplain surface (M) about 4-5,000 years ago. Drawings highly modified from Saucier and Fleetwood (1970). 14 and rivers have made drastic changes in their courses, the floodplain has become a succession of relict surfaces, each bounded by terraces older than those closer to the river. Their importance lies in the hydrologic control they still exert over the modern floodplain. At least three terraces can usually be found in southeastern floodplains. The Holocene terrace is usually the most re- cent; such terraces are known as "first bottoms." The next lowest terrace is known as the Terrace I, Deweyville, or in South Carolina "second bottom," and is distin- guishable en many southern floodplains, including the Altamaha (GA), Pee Dee (SC), Cape Fear (NC), the Pearl and Pascagoula (MS), and the Sabine, Trinity, and Brazos (TX) (Gagliano and Thom 1967). Terrace I sediments were deposited during a fluvial period 17,000 to 36,000 years ago with flows that were five to seven times great- er than at present forming giant meander scars or a braided topography of sandy bars and fossil dunes (Pee Dee River, SC) (Thom 1967). In South Carolina, Terrace I lies 1.5 to 3.0 m (5 to 10 ft) higher than the modern floodplain and 1.5 to 6.0 m (5 to 20 ft) below a still higher Pleistocene fluvial or river terrace known as Terrace II (Gagliano and Thom 1967). Another floodplain terrace classification scheme for the Ouachita River of Arkansas and Louisiana combines three terraces into a "Deweyville sequence" lying between the original pre-Wisconsin glacial sediments deposited in the Sangamon interglacial period (the Prairie Terrace) and the modern Holocene floodplain (Saucier and Fleetwood 1970). Prehistoric floodplain surfaces still function in the modern hydrologic regime. Some are inundated by present high water, and relict channels, ridges, and swales bear vegetation associations indistin- guishable from those on their recent ana- logues. The Pleistocene has left its imprint in many other ways. The mouths of numerous rivers at the coast (Roanoke, Chowan, NC; Escambia, Choctawhatchee, FL) are narrow, drowned floodplains entrenched during the Woodfordian phase of Wisconsin glaciation. The sediments of Terrace I may Tiave provided much of the sands for the barrier islands of the gulf and Atlan- tic coasts (Thom 1967). Other ancient floodplains have been variously used by man. Along the Waccamaw (SC) the relict ridges support roads and pine plantations while the swales bear bog vegetation. Along the Roanoke (NC) row- crop agriculture occupies most of the ridges of the higher Pleistocene terraces (3 m or 10 ft above MSL). 15 CHAPTER 2. HYDROLOGY Water is the driving force of the bottomland hardwood community. As has been shown, water plays a crucial role in forming and maintaining the floodplain by transporting and redistributing sediments within the system. The rivers and their floodplains are fluctuating water level ecosystems. Their high flows are brought about by winter-spring rains (peak flow is in the summer in Florida). Their low flows correlate with high evapotranspiration during late summer and dry fall months (Wharton and Brinson 1979a). Sources of water to bottomlands include precipitation and runoff from mountains and Piedmont (alluvial rivers), groundwater from local convective and storm-front rainfall (lower Coastal Plain blackwater streams), under- ground aquifers (spring-fed alkaline streams), continuous seepage from sand aquifers (bog and bog-fed streams), and tidal flow. Before development, when intact for- ests with thick, organic soil layers cov- ered the landscape of mountains and Pied- mont, almost all water to alluvial streams was derived via subsurface (ground water) flow. Today exposed subsoil horizons in the Piedmont lead to surface runoff which is now the primary source of water to tl^ese streams. On the flat Coastal Plain terrain, surface runoff occurs only spo- radically except when the soils are satu- rated (water table at or near the sur- face); therefore, rainfall in the Coastal Plain reaches blackwater streams via ground water (base flow) in fall and by ground water and surface runoff in winter and spring. Base flows become the primary source to streams during low water or drought conditions. Surficial aquifers (Hawthorne, Creta- ceous) may contribute markedly to base flow. During a fall drought Thompson and Carter (1955) computed base flow discharge from the Tuscaloosa formation to minor Georgia streams ranging from the rainfall equivalent of 28 cm (11 inches) to as much as 102 cm (40 inches) per year from this Cretaceous aquifer. blackwater rivers become visibly clearer in the fall because their flow is derived largely from ground- water base flow. It is reasonable to assume that rivers recharge the shallow aquifers at high water in the flat Pleistocene depos- its, but it is yet to be proven how much rivers contribute to the deeper aquifers. The net contribution of alluvial rivers to the principal limestone aquifer is thought to be insignificant (Stringfield and LeGrand 1966). Surface streams and swamps may recharge valley aquifers (Wharton 1970; Bedinger 1980). ALLUVIAL RIVERS Alluvial rivers in the Southeastern United States originate in the mountains and Piedmont and form huge swamps at the junction of the Piedmont and Coastal Plain. Most of these rivers have periods of sustained high flow resulting from the cumulative effect of many tributaries and distant rainfall (Figure 9A). Generally, the annual high winter-spring runoff water overflows the floodplain features. Pat- terns of river discharge vary in different sections of the watershed. For example, discharge peaks are higher in the Apalach- icola River (FL) in the comparatively nar- rower upper section with high levees and steeper gradient, as compared with the flatter stage hydrograph approximately 48 km (30 mi) downstream where the water spreads out over a much wider (5 x) and flatter floodplain (Figure 10). Differ- ences in wet (flooded) and dry stages can be dramatic (Figure 11). Discharge volumes may cease to rise and sometimes even fall as the water flows through the floodplain toward its mouth (Figure 12). Evapotrans- piration after March leafout and surficial aquifer recharge may help account for some oi^ this flow reduction (Mulholland 1979; Brown et al. 1979). 16 A. ALLUVIAL C. SPRING FED O'ND'J'FMAMJJAS' MONTH Figure 9. Hydrographs of four types of southeastern floodplain rivers and strearrs. Figure 10. Hydrographs (1974-75) of an alluvial river (lower Apalachicola River, FL) showing the possible effects of an increase in floodplain width on water levels, between upstream (solid line. River Mile 126) and downstream (dashed line. River Mile 68). (After Leitman 1978^ BLACKWATER RIVERS Blackwater rivers and tributary streams originate in the Coastal Plain and receive most of their discharge from local precipitation. These streams have nar- rower, less well-developed floodplains and reduced sediment loads compared to those of alluvial rivers.. The waters are rela- tively clear, but highly colored (coffee- colored) due to the presence of organics (humic substances) derived from swamp drainages. A hydrograph of a blackwater stream (Figure SB) is characterized by irregular discharge peaks that are due almost wholly to frontal or local weather events. Summer flooding, as well as more typical winter-spring flooding, may result from local storms. Unlike that of larger alluvial streams (Figure 9A), the hydro- graph of a smaller blackwater stream may register dry periods during which dis- charge may dwindle to near zero. Many blackwater streams are coastal plain tributaries to alluvial rivers. Water levels in some of these streams may be controlled by the discharge levels in the main river creating a "water dam" effect (Wharton and Brinson 1979a). Ground-water seepage, or base flow, is a particularly important component of the discharge of blackwater streams. A study (Winner and Simmons 1977) of a small North Carolina Coastal Plain blackwater stream (Creeping Swamp, N'C) (Figure 13) resulted in a water budget in which over- land runoff accounted for 17.75 cm or 6.99 inches (17%) and base flow runoff for 21.69 cm or 8.54 inches (20%) of the total precipitation of 107.29 cm or 42.24 inches. Evapotranspiration accounted for 65.81 cm (25.91 inches) (61%) of the rain- fall. A negligible 2% seeped underground and was lost to the watershed. SPRING-FED STREAMS Spring-fed streams, characterized by clear, alkaline flow issuing principally from underground aquifers are common in northwest Florida and in other areas underlain by Tertiary limestone aquifers (Figure 14). The discharge hydrograph of a predominantly spring-fed stream (St. Marks River, FL). (Figure 9C) is quite flat compared to those of alluvial and 17 Figure 11. Two photos showing drydown (upper) and inundation (lower) of the floodplain in the Congaree Swan.p National Monument (SC). Photo courtesy of U.S. National Park Service. 18 20Q 100 :^ 50 a> 20 / / ■ — 3/ 2 .. EVAPO PRECIPITATION TRANSPIRATION 107 cm (61%l (100%) B C DOWNSTREAM Figure 12. Relation of flood discharge of Oconee River (GA) to distance downstream. (A) = Piedmont station (Greensboro); (B) = fall line station (Milledgeville, drainage 3000 mV-) just above the the upper Coastal Plain; station at Dublin; (D) = Ocmulaee (Mt. Vernon, 338 km2 or 5150 mi2). area 7770 km^ or Oconee swamps in (C) = downstream junction with the 13 drainage area, (1) = 2-year flood, (3) = 10-year flood, (5) = 50-yedr flood, (Wharton 1980.) (2) = 5-year flood, (4) = 20-year flood, (6) = 100-year flood. OVERLAND RUNOFF H7%l DEEP GROUND WATER OUTFLOW (2%1 Figure 13. Diagram of the water budget for Creeping Swamp (NC), July 1974-June 1975 (after Winner and Simmons 1977). Figure 14. Dotted areas indicate where Tertiary limestones lie at or near the surface, often giving rise to spring-fed streams such as Florida's Chipola, Wa- kulla, Wacissa, and St. Marks and contri- buting heavily to the Suwannee-Santa Fe system. The dashed line is the inner mar- gin of the Coastal Plain. (Adapted from Stringfield and LeGrand 1966.) 19 blackwater streams. In Figure 9C, the highest flows are only about twice the lowest flow. Although most of the base flow of the stream arises from the uniform discharge of the spring, hydrographs may also indicate local rainfall (Rosenau et al. 1977). Spring-fed streams are influenced by surface and ground-water fluctuations. During flood stages of the Suwannee River (FL), the flow from Falmouth Spring is reversed, and the darker waters of the Suwannee flow into the spring. At the other extreme, these streams may go dry annually, leaving an exposed bed as does, for example, the Alapaha River (FL); or the entire river channel, bed and all, may disappear as the river drops into under- ground corridors (lower Aucilla River, FL). BOG AND BOG-FED STREAMS Two additional swamp stream types occurring on the Coastal Plain of the Southeast are bog and bog-fed streams. Bog streams have limited distribution and generally occupy the linear depressions or swales between adjacent sand ridges and reworked Coastal Plain relict dune depo- sits. An example is White Water Creek in Georgia, located in Cretaceous residual dune sands. Many bog streams occur within the Florida Panhandle area. Bog streams are characterized by a steady lateral seepage from the surrounding sand ridges. Therefore, substrates of these systems are constantly wet and support fire-resistant, bog-type vegetation. The linear nature of these streams precludes any significant watershed interception of rainfall beyond that falling directly on the stream. Bog-fed streams, on the other hand, flow intermittently due to discharge from expansive bog-filled depressions. This intermittent discharge occurs only after significant runoff from rainfall exceeds the water storage capacity of the bog. The depressions which feed these streams are areas of internal perched drainage underlain by clay aquicludes (impervious soil layers that retard the downward move- ment of groundwater). These basins are not incised by streams, water tables gen- erally occur at the surface, and excess flow from precipitation discharges readily into the receiving bog-fed stream. The streams receive little or no sediment load; therefore, few have floodplains and most resemble shallow ravines. Their hydrographs exhibit extreme fluctuations in response to rainstorms, with little or no base flow (Figure 9D). Examples are the New and Sopchoppy Rivers in Florida, which drain giant shrub bogs and bay swamps in the Bradwell Bay Wilderness Area (FL). Typically the streams flood rapidly and drain gradually due to the baffling effect of '■.heir dense bay vegetation. FLOODING DURATION AND FREQUENCY Flooding on alluvial floodplains de- pends on the size and slope of the water- shed, which, together with soil and slight elevation differences, help explain the variability in forest communities on vari- ous floodplains. The duration of flooding also directly relates to watershed drain- age area. Bedinger (1980) concluded that drainage areas in the mid-West with less than 776 km^ (300 mi^) have fast runoff characteristics, with flooding occurring 5% to 77o of the year. Flooding occurs in drainage areas ranging between 12,950 and 18,130 km2 (5,000 and 7,000 rniO. Flood- plains for rivers with watersheds exceed- ing several tens of thousands of square miles are inundated from 18% to 40% of the year. Flood peaks are significantly lower in basins with lake and wetland areas (Carter et al. 1978). Steep watersheds with dense clay soils have "flash" inundations of compara- tively short duration. Rivers with intact floodplain swamp forest slow down the rise and fall of floodwaters (Wharton 1970). Flood heights diminish markedly as soon as alluvial (Wharton 1980) and blackwater (Benke et al. 1979) rivers top bankful stage and begin to utilize their flood- plain swamps. Leitman (1978) showed the importance of local rainfall in maintaining saturated conditions at several locations on the Apalachicola floodplain where residual water is often perched in backswamps 1 to 2m (3 to 7 ft) higher than river level. Water levels in these floodplain pools and sloughs rise from local rainfall indepen- dently of river stage. 20 CHAPTER 3. PHYSICOCHEMICAL ENVIRONMENT The physicochetnical environment of floodplains (including both aquatic and soil environments) is a function of the interactions or processes occurring in the water column, in soil, and at the soil- water interface. These processes are facilitated by the prolonged periods of flooding (inundation) which saturate the soils and the subsequent periodic inter- vals of drydown which de-water the soils. This cyclic wet/dry regime imparts a unique chemical environment that has pro- found effects on nutrient cycling and the character and adaptations of the flood- plain biotic communities. CHEMICAL CHARACTERISTICS OF RIVERS The chemical composition of flood- plain rivers and streams reflects water sources, headwater origin, and the compo- sition of geological formations through which rivers flow to the coast. Of the three major chemical classes of world rivers (rock-dominated, precipitation- dominated, and evaporation-dominated; Gibbs 1970), floodplain rivers fall into two: rock-dominated and precipitation- dominated. Alluvial rivers are rock- dominated rivers whose inorganic chemical load is derived from the products of weathering and leaching of the parent rocks and soil in the mountains and Piedmont. Concentrations of inorganic ions are typically higher than total organic carbon (TOC) concentrations (Table 4). Blackwater rivers arising in the Coastal Plain, on the other hand, are precipitation-dominated. Rainfall, which represents most of the water input to these streams, contains relatively low concentrations of dissolved inorganic solids (specific conductance). A compari- son of river data in Georgia (Wharton and Brinson 1979a) indicated that alluvial rivers usually were higher than blackwater rivers in nitrogen, phosphorus, calcium, and magnesium (the latter two constituents increasing water hardness) (Table 4). Blackwater rivers were more acidic (lower pH) and characterized by high concentra- tions of total organic carbon and low con- centrations of dissolved inorganics. Distinctions among blackwater streams can be explained by their different ori- gins within the Coastal Plain (Wharton and Brinson 1979a) (Table 4). The waters of the Satilla River, arising in the lower coastal plain of Georgia, were soft, acidic and highly organic, while the chem- istry of the Ogeechee, Canoochee, and Ochlockonee Rivers reflect the input (increased hardness, pH, and nutrients) from geological formations at their head- waters. In Florida many rivers clear dur- ing low flows, and pH approximates that of subsurface aquifers (pH = 7.7). During high flows, however, surface leachates add organic acids and lower the pH to 4.0. The blackwater Santa Fe River typifies a phenomenon especially evident in many Florida rivers. In its swampy headwaters the Santa Fe has a pH of 5.3. In the cen- tral section, with swamp drainage during high flow and alkaline ground water drain- age during low flow, the pH is 6.4; in the lower river, fed by artesian springs, the pH rises to 7.4. The distinction between blackwater and alluvial river water chemistry is best reflected in the difference in the ratios of inorganic to organic constituents. The high concentrations of organic matter in 'blackwater rivers result in a 1:1 ratio of dissolved inorganics to total organics whereas the predominance of inorganic com- ponents in alluvial rivers leads typically to a 1C:1 ratio (Beck et al. 1974). The magnitude of the organic load affects the concentrations of some of the inorganic load constituents. For example, only those inorganic ions such as iron and aluminum, which form complexes with the dissolved organic matter (COM), are pres- ent in greater concentrations in black- water streams (Table 5). Additionally, since the bulk of the dissolved organic constituents are organic acids (humic and fulvic), the waters of blackwater streams are considerably more acidic (low pH) and highly colored than alluvial streams. 21 Table 4. Physicochemical data summarized for Georgia rivers in water year 1977 (Wharton and Brinson 1979a). Figures in parentheses are numbers of streams for which data were averaged. Rivers TOC (mg/1) pH Hardness (Ca, Mq) (mg/1) Specific conductance (umho/cm) Total nitrite + nitrate-N (mg/1) Total phosphorus (mg/1 ) Mountain river 1-8(2) 6.6(2) 3.5 16(2) <0.C4(2) <0.02(2) Alluvial river-Piedmont 2.9(6) 6.9(6) 12-18(4) 48-83(6) 0.14 -0.50(3) 0.06(3) Alluvial river-Coastal Plain Ocmulgee-Oconee 4.1(2) 7.2(2) 18-33 68-122 0.09 -0.38 0.08(2) Flint (Newton) 5.5 7.5 30-50 84-144 0.34 -0.63 0.06 Altamaha (Everett City) 7.9 6.6 13-33 60-191 0.02 -0.55 0.07 Blackwater river Ogeechee (Oliver) 8.1 6.9 12-28 47-104 <0.18 0.C4 Canoochee (Claxton) 12.2 5.6 6-11 35-57 <0.08 0.06 Ochlockonee 9.0 7.0 11-56 49-327 <1.70 0.32 Satnia-Suwannee 21.7(2) 4.9(2) 5(2) 40-59 <0.06 0.04(2) Table 5. Mean inorganic constituents (ppm) in selected Georgia coastal plain rivers and in the "world average river" (modified from Beck et al. 1974). Rivers pH HCO, CI SO, Na Mg Ca SiO, Al Fe Mn Satilla 4.58 2.6 6.1 Ohoopee 6.25 8.8 5.9 Ogeechee 6.28 30.2 5.4 Altamaha* 6.80 29.6 4.0 World avera river ige 58.4 7.8 0.8 3.70 1.00 0.74 1.0 2.73 0.86 1.00 1.2 3.42 0.74 1.01 2.6 3.95 1.29 0.96 11.2 1.32 6.6 0.41 1.05 0.06 8.69 11.1 0.04 0.11 0.01 6.78 12.3 0.03 0.12 0.03 6.25 10.3 0.24 0.08 0.01 6.30 2.30 4.10 15.00 13.1 0.67 * Represents conditions in the Georgia coastal plain below the confluence with the Ohoopee River. 22 In both blackwater and alluvial sys- tems organic iriatter represents the link between the river and its floodplain. Most of this organic matter is in the dissolved form termed dissolved organic matter (DOK) or dissolved organic carbon (DOC), com- posed principally of humic substances leached from soil, peat, and leaf litter. For example, up to 95% of the total organic matter in the Altamaha River was DOK (Reuter and Perdue 1977). Total organic matter averages around 15 mg/1 (Windom et al. 1975), ranging up to ICO mg/1 in waters leaching peat deposits (Malcolm and Durum 1976). These materials are often chemically and biologically inert (i.e., refractory) with concentra- tions changing principally in response to discharge additions or dilutions. A small proportion of humic substances flocculate in fresh water and can be seen as "silts" on white sand bars, or are rolled as bed load particles (J.H. Reuter, Department of Geophysical Science, Georgia Institute of Technology, Atlanta; personal communica- tion). PHYSICOCHEMICAL CHARACTERISTICS OF FLOOD- PLAIN SOILS The alternation of inundation of floodplains during extended high flow per- iods of the river with drydown periods during low flow conditions produces a spec- trum of soil types across the floodplain. These soil types are associated with elevational gradients which in turn dic- tate flooding frequency and duration: the hydroperiod. Differences in elevation and hydroperiods are the basis of a system of classifying the environmental and biotic zonation that result from this continuum of fluctuating water levels and soil mois- ture. A system of six zones, developed by the National Wetlands Technical Council (NWTC) (Larson et al. 1901), provides a convenient framework for portraying the relationship between the bottomland hard- wood community and environmental factors necessary for effective management consid- erations. Throughout the remainder of this report these zones will be referred to as either ecological or bottomland hardwood zones. Briefly, the classification generally corresponds to the following broad geomor- phologic floodplain features: Zone I: river channels, oxbow lakes, and permanently inundated backsloughs Zones II-V: the active floodplain including swales (II and III), flats and backswamps (IV), levees, and relict levees and terraces (V) Zone VI: the floodplain-upland tran- sition to terrestrial ecosystems Examples of floodplain zonation are depicted in Figure 15. An idealized floodplain proceeds sequentially from the river channel to the surrounding uplands (Zone I-VI) along a gradually increasing elevational gradient (Figure 15A). The presence of natural levees interrupts this sequence (Figure 15B); depending on eleva- tion, the levee may be characteristic of Zones II, III, IV or V. Accordingly, levees are generally excluded from the NWTC zonal concept. Other geomorphic fea- tures (Figure 15C) contribute further to the complexity of zonation patterns on most southeastern floodplains (see Chap- ter 4). Flooding produces and regulates the chemical properties of floodplain soils by (1) continually depositing and replenish- ing minerals, including essential nutri- ents on the floodplain (the mineral sub- sidy); (2) producing anaerobic conditions in the soils; (3) importing particulate and dissolved organic matter (POM, DOM); and (4) removing or exporting accumula- tions of organic detritus (principally degraded leaf litter). The degree to which these processes operate in the six zones is determined by the hydroperiod (Table 6). An example of the relationship of floodplain soil types to bottomland hard- wood zones is illustrated in Figure 16 for an alluvial river floodplain, the Congaree River (SC). The bulk of the floodplain floor is Tsw Caw silty clay loam, support- ing principally Zone IV forest. However, variations in microrelief or subsurface water table height can make differences in surface soils even in small quadrats. Reynolds and Parrott (1980), found spe- cific soil differences on a 1-ha plot coincidental with different patterns of tree distribution and postulated (from 28 wells) that water table differences ac- counted for the numerous soil differences 23 < ^ CD " 'O OJ c C C C ro O -t- 4_ O ■— 'O o '^ 4- n3 ■— o a. T3 O O — cr i) C ■-. 0) — ' OJ C CD <0 l/l O N |5s N O) +-> ■0^4- S- 4-> O § ' 4- 00 u O (U -r- s_ -o ' U OJ •!- • 3 O) 4- "^ 4-> 2 ■o > qj -a JD -r- >, OJ E ■(-> Ol u_ -§ . i. ,— 3 Q. Ll- >> E O -o . +-> -I- XI 1/) O) ' c o *- o r^ £:=§ :*; r- fo S +-> .i; +-> ^ Ol l" '^ s- •■- 3 O e; Ol E; O ■I- S- U. ro Ol ^ > CI ,_ T3 ■o 01 lO 4- C » jQ 24 O) r— o fO +-> "U ■o X 1- 2 +-> OJ , o ■1— *-> 4-> 3 4-> woo 4-> S- c c re 13 •<- >— X , o .— . >- -o -o re O s_ cu 0) o •M H- re -M +-> O i- re re ■o Q- o -a s- o 3 ^ CL c 3 ■M E 3 +-> s- S- c OJ c re QJ O O I— ■>- */) Q- 4- e i_ c o re &^ -o o a> O 4-> LT) Ol c .— •*-> in c C c re re OJ CM o >. c >» 3 3 I/) >>TD -rr -o cr 1 O re re c E c CO en LO surface; C ■actions s ,e with d i/i 4- •r- re u ■M c re o O J3 c >> O i- XI (/I S- o re (U ■t-> (/I u u re O O) u c «c E re O re XI •r— o ■*-» i- 0) ~D re C c o re u OJ 5 >>+-• s- o . — re o s- .— s_ 4- cn re 3 c +-> en CD o re c ^ (/) re S_ 3 . o CSJ s- t— I cn re E re Q. .— O t/i O) >^ re E .— s- E 4- C +J 3 3 O o C -M tn (/I OJ re +-> re C VI XI 1/) (D re c o i/i E s- re E s_ o C7) 0) oi cn c Q.T3 c c ■.- CD 3 E +-> &- S- O OJ re Q. 3 S^ LO "O C/1 TH cn .-t &« o Ln i/i ■-- >> CNj re E re A CD o .— >^ (U +-> u re t/> CD CD c £ S_ CD s- -o TD +-> +-> > c 1- cn^ S- .— o 1 c cn OJ s_ •-- s- •.- 3 ■«-> re 4-» re s_ o c (U re - -O -Q <«-> o re c l/l >> E re o . — s. *^ (U «3 ■!-> 3 la u CM CO S -u ■.- •-I ■* CJ) (O o O CO c c s- CVJ OsJ O LT) > (O (O cn cn ocj >*' O -M C UD CM E t/> 'O "X) re CD .— 4- C "O ■r- O O ^ ^ -^ o re re -M ' — xa 3 re 4- o o o 1- c " re «— o u •<- s- .. -r- +-> I— re +J '— o 4-»corere Erere ■«— r- o 3 ■'- 3 -r; -^ ■f-Q..i(:> u-^> re t/lE03 -r-US C • -ore'— ere' — (D ■OO' — ■ — "3 ' — ' — o^ c CO > re 3^ i- X oo — ' o o 25 ) .c •• »■ ^ ) (/> S "o -c fo *+- 1- . ■<- O OJ <>1 o o -o -— t. ••- -c: >— s-O r— 3 ■♦-> OJ O ■ (U at -c o •»- en o •I- .— fO TD I * #. S CIJ - 1- ^ c »— -S3.— S I^ S- O) C i- -^ o I -Q J3 >) fO -Q 3 E ro * O .— cn c s -c c: jD ■*-» 3 O i/> TJ c t. o i- •.- -c CO f001--Q3 "*/ia> C JD O C •«- r- •^ >, ^ ^ 3 -0 -i-> E n3.cl-l-»>— OXJ-M O Oi-.- 3 >>-Q I- E o CO S- .— O -C D14- ■!-> "O OJ -C C r— >> s- t/) 3 +-> +J ■o f^ >) o c *o ■o ro ^ E 'D 3 O) s- c ^ s- > cn c ■r- U 3 3 E TJ ■o ^ o c (/) t- (3 XI fO 13 •.- JD .-» LO CO ^ O O O O o o O^ C\J LT) LO CT> CO (X) CT^ ko irj ^ 00 OJ ro •>* UD CO *:t CO C\J n KO r-l o »— 3 .^ s^ 3 ■'- .i^ > O -i^ > fo f^ d. (O > — E 1_ CL I- -^^ S- Q. Ci. CU — - QJ E OJ ' Cl.+->. — ■*-> ' — Q.+->( — ■— ' (OfC E'^'D ^fOW E 3 ••- 3 3 ••- ^- 3 •-- 3 £j^> ;- .^ > .^ > -r- . 3U3 cnUI3 EU3 V) •r-(0.— ajfO*— Srar— W w, . . U-— >— C-— r— ..-r-r- fO. O CQ •=£ . — CO 10%). The highest soil organic matter occurs on floodplains draining vast, acid bogs along rivers such as the Sopchoppy and New (FL), and on floodplains along spring-fed rivers (32%), tidal forests (40%), and on peat systems (up to 44%). Percent organic matter of Zone II soils in spring-fed river floodplains and Zone II backswamps with swamp tupelo is also quite high (about 36%). Alluvial river flood- plains have the lowest organic matter, averaging <5%, a good working figure to separate blackwater and alluvial flood- plains (Wharton et al. 1977). Soil Characteristics of NWTC Zones Overall, macronutrient concentrations on floodplains differ markedly from those of the uplands (Table 7 and Appendix). Ex- cept for phosphorus, nutrients (especially calcium and magnesium) are generally higher in floodplain soils than in up- lands. Unusually high concentrations occur in spring-fed and tidal systems in particular. 29 Table 8. Percentage of organic matter in selected wetland and upland soils. Sites Organic matter (%) References Pocosin 66.8 Swamp 31.3 Floodplain Zone II (alluvial) 4.5 Zone II (blackwater) 18.2 Zone II backswamps with swamp tupelo 35.3 Zone II springfed rivers 36.0 Zone III (alluvial) 3.4 Zone IV (alluvial) 2.8 Zone IV (blackwater) 7.9 Zone V (alluvial ) 2.8 Coastal Plain pines 0.4 Piedmont pines & hardwoods 1.4 Mountain hardwoods 1.5 Woodwell 1958 Woodwell 1958 This study Long et al. 1969 Perkins et al. 1962 Perkins et al. 1962 Some zones (see Appendix) are roughly comparable in calciuR' concentrations (660 ppm) to a tulip poplar forest (Shu- gart et al. 1976), which is considered eutrophic (Jordan and Herrera 1981). Some floodplain sites (blackwater Zone IV, alluvial Zone V) approach oligotrophy (44-75 ppm Ca). In the study area, the lowest nutrient levels occur in Zone V. There are marked differences in organic matter concentrations among the various zones. Some zones (II and black- water IV) have low nutrient availability due to "lock up" of nutrients in organic matter. This condition occurs particularly in swales and peat-forming soil types. Periodic drydown is very important to nutrient release from organic matter and litter. When drydown is rare or aperiodic (as in the specific zones above), nutri- ents tend to be bound in complexes with organic matter. On other sites nutrient concentrations are low as a result of a lack of inorganic inputs (as in remote swales adjacent to the upland). In black- water rivers, nutrients n:dy be complexed and exported as particulates or refractory humic substances. Soil micronutrient concentrations tend to be high, especially in acidic sites, although this may depend on the amount of inorganic input, or distance from the channel (Appendix). Cobalt stor- age by swamp tupelo (Eyde 1966) and the high zinc demand of cultivated pecan trees suggest that some floodplain vegetation may accumulate or need high levels of micronutrients. The precise relationship between soils and vegetational responses in flood- plains is unresolved. It is unknown, for example, why the low-nutrient, low-organic flats of the Oconee support a magnificent willow oak stand, or why the old growth diamondleaf oak on Turkey Creek has the lowest organic matter of any Zone IV floodplain. Equally challenging is why the mineral-rich Taw Caw silty clay loam of the virgin Congaree Swamp National Mon- ument supports extensive Zone IV wet flats with few or no large trees. 30 CHAPTER 4. FLORA OF BOTTOMLAND HARDWOOD COMMUNITIES INTRODUCTION Having established in the preceding chapters the geological and biochemical setting unique to river floodplains, we now turn to the component of the ecosystem that gives it its name--the bottomland hardwoods. The plant species that thrive here are well adapted to the stresses imposed by the hydroperiod; these trees and their adaptations are a fundamental and integral part of the geological and chemical functioning of the ecosystem. The plant species and communities that inhabit the floodplain can be use- fully thought of as buffers that absorb and dissipate the physical energies of the riverine system. Water movement is slowed and erosion is held in check through the anchoring of sediments by root systems, the deposition of sediments that are dropped from the slowed water column, and the reduction of the water column by the spreading out of water (Leopold and Wolman 1957). Without the stabilizing forces of the biota to reduce water velocities and inhibit subsequent meander movement and floodplain scour, these physical altera- tions would be extremely rapid. The buffering role of the plant com- munities is also evident in the biogeo- chemical cycles of the riverine-palustrine system. Essential mineral nutrients are captured from the water-soil complex and fixed in plant tissues which ultimately support the floodplain 's detritus-based trophic network (Wharton and Brinson 1979a). Remineralization by the soil microbiota and rapid uptake by plants during favorable (nonflooded) conditions partially close the nutrient cycles. But the nature of the riverine-palustrine system, in which the variable patterns of flooding and drydown continually interact with the floodplain substrate, requires active nutrient conservation by the biota (Brinson et al. 1980). Floodwaters trans- port nutrients not immobilized in organ- isms or bound to soil constituents back to the river, as particulate or dissolved organic matter, material adsorbed to sus- pended sediments, or solutes in the water column (principally dissolved organics). The relatively high levels of productivity exhibited by floodplain ecosystems (dis- cussed later) are sustained only through the water and nutrient subsidies provided by the watershed and transported by the river (Brown et al. 1979; Brinson et al . 1980). The floodplain flora, in partner- ship with the macro- and micro-fauna, merely postpones the loss of elements to the sea. The trapping, assimilation, and partial cycling of nutrients in the flood- plain, essentially a diversion in the relentless movement of water and sediments to the ocean, yield an extremely produc- tive and unique ecosystem. THE ANAEROBIC GRADIENT The distribution of flora in the bot- tomland hardwood ecosystem revolves around three aspects of anaerobic conditions: (1) the presence and intense selec- tive power of anaerobic condi- tions generated by the hydroper- iod on the floodplain; (2) the anaerobic gradient, varying in space and time across the floodplain due to microeleva- tional relief, the soil mosaic, and the hydroperiod; and (3) the tolerances of plant species to this gradient. Though factors such as light inten- sity, soil pH, and nutrient availability affect plant distributions in other forest communities, they are secondary to anaero- biosis in the floodplain community. In fact, these other factors are, except for light intensity, functions of saturated soils and thus anaerobic conditions. The anaerobic gradient in the flood- plain and its effects on plant distribu- tions have been noted, often as "moisture 31 gradient" or "moisture continuum" (Lindsey et a1. 1961; Gemborys and Hodgkins 1971; Bedinger 1978; Richardson et al. 1978; Fredrickson 1979; Huffman 1979; Whitlow and Harris 1979; Huffman and Forsythe 1981). These terms may be misleading; it is not the availability of water, but the inavai lability of oxygen due to the pres- ence of water. The emphasis on the anae- robic aspect of this gradient generates a clearer picture of the actual effects of flooding and saturated soils on plant sur- vival; hence, its use in this report. PLANT RESPONSES TO ANOXIA-RELATED STRESSES Stresses Generated by Anaerobic Conditions The effects of periodic or perma- nent flooding are the crucial selective stresses on bottomland hardwood plants and are responsible for the sorting of species into broad community types (Huffman and Forsythe 1981). The plant growing in a saturated substrate must respond to sev- eral physical and chemical changes, among them: (1) depletion of available oxygen in soil water, in a period as short as 3 days (Nuritdinov and Varta- petyan 1976; Phung and KnipTing 1976; Teskey and Hinckley 1977); (2) shifts in soil pH— variable, though in general a convergence toward neutrality, with acidic soils becoming more alkaline and calcareous soils becoming more acidic (Grable 1966; Kennedy 1970; Rahmatullah et al. 1976; Teskey and Hinckley 1977); (3) accumulation of potentially tox- ic compounds in the plant, the rhizosphere, and in the larger soil solution; examples are car- bon dioxide, ethanol, sulfides, nitrites, aluminum, iron, and manganese (Teskey and Hinckley 1977); (4) shifts in the redox states of chemical species, including es- sential nutrients, generally from more oxidized to more reduced; the reduced forms are considered generally less desir- able for plant uptake and assim- ilation (Brady 1974; Teskey and Hinckley 1977); and (5) shifts in nutrient availabil- ities, partially due to item (4) (Teskey and Hinckley 1977). The responses of plants to these and other flood stresses were reviewed by Teskey and Hinckley (1977), who emphasized that the key to plant survival in flooded condi- tions is the adaptability of the root system. The cessation of uptake and exchange functions through root dormancy or death during flooding affects plant metabolism in several ways. The immediate losses of these root processes is due to the lack of oxygen. The root system has access to free oxygen, necessary for normal respira- tion, through only two routes: (1) absorp- tion from the soil -air-water complex by the roots themselves, or (2) transport from aboveground plant tissues through the vascular system or intercellular spaces to the roots. Although all plants probably have a shoot-to-root intercellular space network through which oxygen can diffuse to the root system (Salisbury and Ross 1978), this system is well developed in only a few plants (rice, for example). Thus the depletion of soil oxygen by the roots eventually shuts down respiration in root cells. As respiration ceases, water and ion uptake is inhibited (1) by chang- ing membrane permeabilities in root cells, affecting movement of both water and ions, and (2) by reducing the amount of energy available for membrane transport, affect- ing primarily ion movement. The inability of flood-intolerant species to absorb and use water and nutri- ents leads to foliar water deficits, sto- matal closure, and reduced gas exchange. Consequently, transpiration and photosyn- thetic rates are slowed, cellular synthe- sis requiring unavailable nutrients is curtailed, and overall plant growth is impeded (Teskey and Hinckley 1977). The plants literally die of dehydration in standing water. Plant Adaptations to Flood Stresses Plant adaptations to flood stresses may be categorized as physical or meta- 32 bolic. The former includes the provision of oxygen to the roots or the restoration of proper root function, or both. Meta- bolic mechanisms adjust plant biochemistry to decrease the potentially harmful ef- fects of anaerobic respiration. The most successful species in saturated conditions are those that possess both physical and metabolic adaptations (Teskey and Hinckley 1977). The abilities of plant species to restore and maintain the stressed root system lie on a continuum (Teskey and Hinckley 1977): (1) very tolerant--primary root maintenance, secondary and ad- ventitious root growth, (2) moderately tolerant--primary root deterioration, adventitious root growth, and (3) intolerant--primary root deteri- oration, no adventitious root growth. Adventitious and secondary roots pro- duced under flooded conditions are anatom- ically different from primary roots in ways that enhance root function in satu- rated soils. They are more porous, facil- itating (1) oxygen diffusion from the aerial shoots (Luxmoore et al. 1973), (2) gaseous exchange between root cells and soil solution, and (3) perhaps better movement of water and ions into the root (Jat et al. 1975). They are also more tolerant to elevated carbon dioxide con- centrations and exhibit increased anae- robic respiration (Hook and Brown 1973). Some tree species produce special root structures other than secondary and adventitious roots. The classic example is the pneumatophores of baldcypress and pond cypress (knees) (Figure 18) and water tupelo and swamp tupelo (arched roots). Aerial roots may supply additional oxygen to the root system (Teskey and Hinckley 1977). Buttress formation (Figure 19) and "stooling" not only provide stabler anchoring in the less firm floodplain soils but also may help aerate the root system. Similar functions are provided by the characteristically wide, shallow, matted root systems (Figure 20) of bottomland trees which (1) provide support, (2) increase oxygen use efficiency in satu- rated conditions by their proximity to more highly oxygenated surfact sediments, (3) reduce losses of nutrients from the system through rapid uptake, and (4) pro- tect the floodplain from erosion. The primary metabolic mechanism in flood-tolerant species is a shift in the end-products of glycolysis. Normal glu- cose metabolism and energy (ATP) produc- tion in the cell proceeds via three steps: (1) glycolysis (anaerobic), (2) Kreb's citric acid cycle (aerobic), and (3) oxi- dative phosphorylation (aerobic). In the absence of free oxygen, only glycolysis is completed, and ethanol normally accumu- lates as an undesirable end product. Flood-tolerant species can generate organic acids instead of ethanol as pro- ducts of glycolysis (Crawford and Tyler 1969) and thus avoid ethanol toxicity. Furthermore, the organic acids may be transported to the stem and leaves (Chirkova and Gutman 1972; Vester 1972) and used in cellular synthesis (Crawford 1976). A second metabolic adaptation has been described for some tolerant trees by Hook et al. (1970). The roots of these species oxidize the rhizosphere, prevent- ing root deterioration and enhancing nu- trient uptake. Finally, there is some evidence that flood-tolerant species can substitute ni- trate for free oxygen as a terminal elec- tron acceptor in cellular reactions (Crawford 1976). The reduction of nitrate to ammonium (denitrif ication) then would help maintain cellular energy production and biosynthesis in roots. This benefit could occur only if excess nitrate were available in an environment where denitri- fication is the prevalent process. Factors Affecting Plant Response to Flooding Of the many factors that influence plant survival during flooded conditions, the timing, depth, and duration of flood- waters are the most critical (Teskey and Hinckley 1977; Huffman and Forsythe 1981). These characteristics are themselves func- tions of regional precipitation and local 33 Figure 18. A remarkable example of multiple-trunked stooling of Ogeechee tupelo at Sutton's Lake (Apalachicola River, FL). This slough floods to depths of 4.2 m (14 ft); cypress knees may exceed 3.7 m (12 ft) in height. weather patterns, watershed size and mor- phology, floodplain size and topographic variation, and drainage rates of flood- plain soils. The effects of flooding are most critical during the growing season, particularly during the period of leaf- out. Floods during the dormant season have relatively little effect on the physiology and survival of bottomland spe- cies (Hall and Smith 1955), other than possible damage due to mechanical abrasion or breakage. Flood depth is critical in at least three ways. First, stem lenticels (pores) may be blocked. These structures are important in some species in both root aeration (Armstrong 1968; Chirkova 1968) and the release of volatile end-products of anaerobic respiration, such as ethanol , ethylene, and acetaldehyde (Chirkova and Gutman 1972). Floodwaters deep enough to inundate major portions of the stem lenti- cels thus cause reduced oxygen supply to the roots and toxic accumulation of the anaerobic respiratory products. The second effect of flood depth is the reduced rate of oxygen diffusion through the water column to the roots with increasing flood depth. Finally, seedlings submerged by the water column may undergo severe mor- tality through anoxia, mechanical damage, and siltation. 34 Figure 19. Oak displaying buttressing, common among bottomland hardwoods. 35 Figure 20. A windthrown diamondleaf oak on a small blackwater creek floodplain (Creep- ing Swamp, NC) illustrates the large diameter of the root crown of bottomland hard- woods. The thickness ranged from 30 to 46 cm (12 to 18 inches). Such width is probably an adaptation to the high water table, but it also increases contact with the surface water during inundation. Root mats are so wide that few areas of floodplain surface are unprotected from floodplain scour. The importance of flood duration should be obvious. With the exception of species of tupelo and cypress, stresses associated with saturated soils and stand- ing water cannot be handled by plants after varying amounts of time that depend on the range of tolerance mechanisms of the individual species. Broadfoot and Williston (1973) stated that the majority 'of the bottomland species will not survive 2 years of continuous flooding. Factors that increase the dissolved oxygen concentrations in floodwaters are rainfall (Broadfoot 1967), moving water (Hook et al. 1970; Harms 1973), and lower water temperatures (Broadfoot and Willis- ton 1973). In contrast, oxygen concentra- tions may be reduced through microorgan- ismal respiration, especially in waters with high concentrations of organic matter or nutrients or both. 36 In addition to the above factors that directly affect plant survival, the activ- ities of soil microbiota are modified by flooded conditions. Decomposition and conversion processes mediated by these organisms, such as mineralization and nitrification, are affected. Wharton and Brinson (1979a) proposed a nitrogen circu- lation model for forested wetlands that summarizes nitrogen flows and the effects of floods. Extended anaerobic conditions and shutdowns in organic matter decomposi- tion may lead to the immobilization of nitrogen and other nutrients in microor- ganismal tissues. PLANT COMMUNITY PATTERNS IN THE FLOODPLAIN The wide variations in factors that influence southeastern bottomland hardwood ecosystem structure and dynamics make a comprehensive treatment of plant distribu- tions in these ecosystems a difficult task, one more detailed than is appropri- ate for this community profile. Although the selective power of the hydrological ly generated anaerobic gradient is sufficient to separate broad community types based on dominant woody species (Figure 21), asso- ciated factors blur the distinctions be- tween categories. These factors include soil characteristics, detrital decomposi- tion rates, soil and water pH, nutrient availability and turnover rates, flood depth and water velocity, light intensity, and disturbance (natural and man-caused). Differences in community structure and composition among otherwise similar sites sometimes occur. The mere presence of a species may not be related to present local topography. For example, apparently dislocated cypress may indicate the exis- tence of an old buried waterway (A.L. Radford, University of North Carolina at Chapel Hill; personal com.munication). The reasons for such complexity in floodplain floral distributions are the individual responses of plant species to the highly variable and dynamic floodplain environment. This section on plant com- munity distributions emphasizes the domi- nant types of forest cover, and notes associated understory, shrub, and herbace- ous components where field observations a 1 1 ow . The National Wetlands Technical Council Zonal Classification The zonal classification of flood- plain forest sites proposed by Huffman and Forsythe (1981) and implemented by the National Wetlands Technical Council (NWTC) was introduced in Chapter 3. Six zones based on soil moisture and hydrology are defined, ranging from aquatic (Zone I) to upland (Zone VI) ecosystems; Zones II through V represent the floodplain. The mosaic distribution of floodplain microtopography (Figure 22), soil types, and plant communities makes the use of the term zone somewhat misleading. While many examples of southeastern bottomlands exist where the plant dominance types are arranged in discrete bands, many others are arranged in a mosaic pattern. The zonal classification is a practi- cal system, but like all man-devised classification, it is flawed. Its use in the analysis of floodplain vegetation is complicated by several problems, among which are (1) the recognition of zones in the field, (2) common species whose adap- tations permit them to occur in several zones and (3) the system's exclusion of natural levees. In spite of these draw- backs, the zonal system is a useful frame- work for the understanding of broad flood- plain community patterns, and hence is used here. Woody Species Attributes A familiarity with the structural and functional characteristics of the woody species of the southeastern floodplains prepares the reader for a better under- standing of community distributions. The extant data support the concept of indivi- dual species adaptations to the selective forces of the floodplain environment. The distribution of bottomland tree, shrub, vine, and herb species over the floodplain zones is shown in Table 9. Structural and functional attributes of m,ost of the important woody bottomland species rriay be found in Putnam (1951), Putnam et al. (1960), and Eyre (1980). The survival of bottomland hardwood species under different hydroperiods pro- vides a validation of the gradient concept 37 Upland forest I White Oak. Blackgum, White Ash, Hickories, Winged Elm, Loblolly Pine) Sycamore • Sweetgum - American Elm Upland Forest River Channel First Bottom (Terrace) Second Bottom (Terrace) Upland Figure 21. The correspoiKJence between alluvial floodplain microtopography and forest cover types. (A) = river channel; (B) = natural levee (front); (C) = backswamp or first terrace flat; (D) = low first terrace ridge; (E) = high first terrace ridge; (F) = oxbow; (G) = second terrace flats; (H) = low second terrace ridge; (I) = high second terrace ridge; (J) = upland. The vertical scale is exaggerated. Figure 22. Microtopographic relief on a snail blackwater creek floodplain (Lower Three Runs Creek, Barnwell County, SC). Areas of similar elevation are similarly marked. Arrows indicate channels which are always filled with water. Quadrat is 100 m on a side. (After Hay 1977.) and the zonal classification system (Table 10). Trees in the almost constantly inun- dated Zone II may survive with roots partially inundated as much as 90% of the time and die only when inundation is permanent. On the other hand, upland (Zone VI) trees not so adapted to maintain themselves during flooding may begin to show signs of stress if constantly inunda- ted as little as 2% of the time (dogwood and black cherry) and die as the flooding interval increases to 12% to 17% of the time. Dominance Types and Their Distribution Based upon field observation and studies in the four-state study area, we have classified bottomland hardwoods on floodplains into 75 dominance types organ- ized by zones (Tables 11-14). Although Zones I (open water) and Zone VI (uplands) are relevant, they are not presented other than to introduce the nature of Zone I species. Each table organizes the dominance types for each zone by topographic setting or uses other features to aid field iden- tification. The reader should review 38 Table 9. Trees, shrubs, vines (V), and herbs (H) characteristic of south- eastern bottomlands and the floodplain zones in which they most frequently occur (A = abundant, C = common, U = uncommon or localized, R = rare). Species (except some herbs and vines) are in approximate order of their position on the moisture gradient from wettest to driest. Species largely restricted to eco- tones (E), levees (L), and peat soils (P) are also distinguished. Nomenclature generally follows Kurz and Godfrey (1962) and Little (1979). Species Ecological zones II III IV Taxodium distichum (baldcypress) A X Taxodivm asaendens (pond cypress) C X Proserpinaaa sp. (proserpinaca) C X Nyssa aquatica (water tupelo) A X Nyssa bi flora (swamp tupelo) A X Nyssa ogedhe (Ogeechee tupelo) A X Crinvm ameriaanum (strap lily) A X Leitneria floridana (corkwood) U-R X Tillandsia setaaea (needleleaf wi Id pine) (H) C X Planera aquatica (water elm) A X Orontium aquaticum (goldenclub) (H) C X Fraxinus aaroliniana (water ash) A X Fraxinus profunda (pumpkin ash) C X Iris virginioa (blue flag) (H) C X Chamaeoyparis thyoides (Atlantic white cedar) U X Pinus serotina (pond pine) C X Magnolia virginiana (sweet bay) C, P X Persea borbonia (red bay) C, P X Sabal palmetto (cabbage palm) C X Ilex myrtifolia (myrtle-leaf holly) C, P X Ilex oassine (dahoon) C, P X Lyonia luoida (fetterbush) A, P X Viburnum nudum (southern withered) C, P X Leucothoe raoemosa (swamp leucothoe) C, P X Clethra alnifolia (sweet pepperbush) C, P X Lyonia ligustrina (male-berry) C, P X Ilex coriaaea (large gall berry) C, P X Cyrilla raoemosa (titi) A, P X Alnus serrulata (alder) A X Myrioa cerifera (wax myrtle) A X Crataegus aestivalis (may hav;) U X Forestiera acuminata (swamp privet) C-U X X Hymenocallis crassifolia (spiderlily) (H) C XX (continued) 39 Table 9. (Continued), Species Ecological Zones II III IV Hymenocallis oooidentalis (spiderl i ly) (H) C, P X Impatiens aapensis (jewel weed) (H) C XX Triadnum tubulosum (St. Johns wort) (H) U X X Veimonia gigantea (iron weed) (H) U XX Senecio glabellus (butterweed) (H) C XX Woodwardia caceolata (snail chain fern) (H) A XX Onoolea sensihilis (bead fern) (H) A XX Osmunda regalis (royal fern) (H) C X X Thelypteris palustris (marsh fern) (H) C XX SmilcLx laurifolia (laurel leaf greenbrier (V) A XX Salix nigra (black willow) A X Cephalanthus oaaidentalis (buttonbush) A X Salix aaroliniana (Ward willow) C X Ilex verticillata (winterberry) U X Gleditsia aquatiaa (water locust) U X Itea virginica (Virginia willow) A X Carya aquatiaa (water hickory) C X Queraus lyrata (overcup oak) A X Junous effusus (rush) (H) A X Acer rubrwn var. drummondii (red maple) A X Saururus oernuus (lizardtail) (H) A X Diospyros virginiana (persimmon) R X Illicium floridanum (star anise) U, E X Styrax amerioanum (American snowbell) U X Amorpha frutioosa (lead plant) C X Comus (striata) foemina (stiff dogwood) C, E XX Viburnum dentatwn (arrowwood) C, E X Fraxinus pennsylvaniaa (green ash) A X Queraus laurifolia (diamondleaf oak) A X Queraus phellos (willow oak) U X Ulnus ameriaana (American elm) C X Liquidambar styraaiflua (sweetgum) A X Leuaothoe axillaris (coastal doghobble) C X Betula nigra (river birch) A, L, E X Crataegus viridis (green haw) C X Ilex deaidua (possum haw) C X Carpinus caroliniana (ironwood) A X Viburnum obovatum (Walter's viburnum) C X Gleditsia triaanthos (honey locust) R X Sabal minor (swamp palm) A X Populus heterophylla (swamp Cottonwood) U, L XX (continued) 40 Table 9. (Continued). Species Ecological zones II III IV Platanus oaaidentalis (sycamore) U, L XX Ehapidophyllwn hystvix (needle palm) U X Populus deltoides (cottonwood) L), L XX Crataegus marshalHi (parsley haw) C X Celtis laevigata (sugarberry) U X Rhododendron visoosum (svjamp azalea) C X Rhododendron oanesoens (hoary azalea) C X Sebastiana ligustrina (Sebastian bush) C XX Smilax walteri (coral greenbrier) (V) C XX Smilax smallei (Jackson greenbrier) (V) C X Berohemia soandens (supplejack) (V) A X Wisteria frutesoens (wisteria) (V) C X Rhus radioans (poison ivy) (V) A XX Traahelospermwn difforme (trachelospermum) (V) U X Brunniohia eirrhosa (ladies eardrops) (V) U X Bignonia aapreolata (cross vine) (V) A XX Commelina virginiana (spiderwort) (H) C X Ampelopsis arborea (peppervine) (V) A, L X Tovara virginiana (jump seed) (H) C X Elephantopus oaroliniana (elephants foot) (H) C X Justicia ovata (justicia) (H) C X Carex intwnesoens (sedge) (H) C X Carex typhina (sedge) (H) C X Carex lurida (sedge) (H) C X Carex louisianioa (sedge) (H) C (CP) X Coj-ex greyii (sedge) (H) C X Leersia lenticularis (cutgrass) (H) C X Leersia virginica (cutgrass) (H) C X Oplisemenus setarius (H) C X Erianthus strictus (plume grass) (H) C X Panioum agrostoides (panic grass) (H) C X Panioum rigidulim (panic grass ) (H) U X Morus rubra (red mulberry) U XX Aaer negundo (boxelder) L XX Pinus glabra (spruce pine) C X Arundinaria gigantea (river cane) (H) A X Vaociniwn elliottii (Elliott's blueberry) C X Quercus miehauxii (swamp chestnut oak) C X Ilex opaca (American holly) A X Quercus nigra (water oak) U-R X (continued) 41 Table 9. (Concluded). Species Ecological zones II III IV Carya oovdiformis (bitternut hickory) U Carya glabra (pignut hickory) U Catalpa hignonioides (catalpa) U, L Queraus pagoda (cherrybark oak) C Asimina triloba (paw paw) C Pinus taeda (loblolly pine) C Queraus shumardii (Shumard's oak) U-R, L Quercus virginiana (live oak) U Serenoa repens (saw palmetto) U Lindera benzoin (spicebush) U Fagus grandifolia (beech) C, E Aristolochia serpentaria (Virginia snakeroot) (H) C Podophyllum peltatwn (mayapple) (H) U Chasmanthium laxa (river oats) (H) C X X X X X X X X X X X X X X floodplain features (Chapter 1) and refer to Figure 40, which illustrates the micro- topography of nine selected floodplains. Table 15 is cross-referenced to Figure 40, thereby providing precise locations of many dominance types. The best examples of each dominance type have been documented by locality and are listed in Tables 11-14. These domi- nance types are intended to prepare the reader for the incredible variety of bot- tomland forest communities and associa- tions which, as yet, have been little studied. Occurrence is also indicated in each table as common, ecologically or geo- graphically localized, or rare. Where possible, reference is made to Society of American Foresters' (SAF) forest cover types (Eyre 1980); though general and not always applicable in this study area, this publication is useful. Huffman and Forsythe (1981) classed a nur„- ber of SAF types in their zonal descrip- tions, including Zone VI, and related them to soil moisture regimes for a broad regional spectrum of floodplain types. Plant Communities in Zone I Submerged vascular aquatic plants are confined to Zone I: rivers, guts, sloughs, pools, and other permanently inundated areas. The dominant aquatic plant in the Santee River floodplain swamp was an introduced species, alligator weed (Alter- nanthera philoxeroides) (Dennis 1973). Other species noted in this floodplain which are characteristic of the region in general are water weed (Eqeria densa), hornwort (Ceratophyl lum), water milfoil (Myriophyllum) , Brazilian elodea, duckweed (Lemna perpusilla), Spirodela polyrrhiza, water or mosquito fern (Azol la carol ini- ana), Proserpinaca, and frog's-bit (Limn3 bium form '-oserpine Jngia). spongia). The submerged, thin-leafed of spatterdock (Nuphar luteum) is common in many spring-fed rivers. On floodplains with tidal flushing, an inter- tidal zone vegetated by quillwort (Isoetes f laccida), eel grass (Sagittaria kurzi- ana) , water milfoil, and Ludwigia may occur (Figure 40, St. Marks River). 42 en C (/) •1- QJ >)■■- i~ O c^ la Ol CO > Q. ^ . oo u r^ (O t/i •!— r^ '^ ^ Ol -C cr. o) o •r- s --I 1^^ u O) O) -D >-, c~ O. C 01 f— 00 O ( — 3 >1^ o Ol •!- U3 0) ■<-> 3: Ol c ^ "i T3 •r— Q) -r- C s- E .— (TJ 3 o "O „ tn O) >, o ^ , E.C 0) >,4J EOO 1— o — o E O) S_ o c u +-> -1- ■M S_ 4-> O 13 O JD TD O s- oi e S- O 4- 3 -r- o 4-> ■!-> (O (O c £■0 0 c •.- 14- 3 +-> O C 03 •I- "O 0) c OO 4- 13 E O C o •■- Q. 01 oo E 0) 0) -1- > q: 4-> •.- > 4- S_ . O 3 O 00 ■— 1 l/l .C +J QJ +J O .— Ol c JD C C (O OJ (O 1— r— U II • • • • o o f~~ r— ■^ OJ OJ (O a. Ql o 3 3 1— 1— CL 3 Ol S- (J E QJ t. (O -l-J ►— < QJ —■ 3 (O t—i > > .-iOO 3 '—' O •—I o> QJ (U c c c o o o 1^ r-j tM m 3 .i<: 0(0 C QJ ■o o o c o .— O _i S_ E CL O 3 O E (O 3 >, o .c -r- s: c QJ .— +-> oo o c ■— 3 S- "CI 1_ O •!— (O O) 0) 1— I 1— I 3: s ni D. cr: .^ lO >iO S- >, O 4-> S- .^ 3 o O C >> s- >, .i^ •!— -*-> i. (O U IE oo S- 1— 1 — .^ •.- OJ OJ CL O (O :r .^ .c E JZ O E 31 C3 s- <_> (O -O <_> D. 3 +J (O QJ o Ol .C . S- 3 J3 CL JD O .i<: O- .^ u S- Ol C Ol E C 3 O •.- • CQ ■a: S a. I/O oo > 3: Q CQ 1— CQ (U (U c e: o o M M u i. oo •.- OJ QJ CQ (O i- •.- O. O (J S- r— E QJ OJ (O (O CL > ■!-> O I/O •.- (O >) cc: <_) OO OJ Ol > OJ 43 Dominance Types of Zone II The dominance types of Zone II (Table 11) occur in the wettest parts of the floodplain: very wet flats, swales, sloughs, and backswamps. Soils are satu- rated throughout the growing season (100% of the tine; Leitman et al. 1981) although fall drydown of water occurs in a number of types. Saturation in some types is maintained by seepage or by tidal fluctua- tion. The liverwort (Porella pinnata) growing on trunks of trees in this and other zones in an indicator of flooding depths and duration (Figure 23). Gum-cypress dominance types (1-10) (Figures 24-27). Subtle factors determine the relative dominance of baldcypress (Taxodium distichum), water tupelo (Nyssa aquatica), swamp tupelo (N^. biflora), and Ogeechee tupelo {H. ogeche) in the tupelo gum-cypress types. Although water tupelo occurs on disjunct Piedmont sites, it is restricted primarily to alluvial flood- plains of the Coastal Plains. Swamp tupelo is prominent in floodplains of the Coastal Plain, but it is also common in upland swamps and ponds and in the brack- ish waters fringing estuaries (Penfound 1952). Water tupelo tolerates deeper and longer flooding than does swamp tupelo and dominates on sites characterized by this hydroperiod. Ogeechee tupelo is limited to the Coastal Plain and occurs in two distinct growth formations (see types 4, 7, 8, 9) on both alluvial and blackwater floodplains. Baldcypress is replaced by the tupelos on many sites because of its erratic reproduction, slower growth rates, and insignificant stump and root sprout- ing. These factors are intensified by frequent disturbance, such as periodic logging, further favoring tupelo dominance (Putnam et al. 1960; Eyre 1980). Pond cypress (Taxodium ascendens) is the co- dominant with tupelo gums on some black- water floodplains. Tree and shrub subcanopies occur in many gum-cypress types (2, 3, 4, 9, 10, 15, 17-21) and may be extremely dense in some (types 15, 17, 20, 21). Subcanopy vegetation in other types may be limited (type 1) because of low light intensities and extended flooding. The herbaceous layer is insignificant in most types but is surprisingly dense in others (types 4, 12, 16). Swamp tupelo dominance types (11-16, 18, 19) (Figure 28). These types occur on organic black mucks or peats (the latter if bays are present). The deeper the peat, the denser the shrub understory (see type 15, which may be characteristic of blackwater river floodplains at elevations approaching sea level). These types on alluvial floodplains often occupy the swales and filled-in oxbows that flank the upland. Swamp tupelo types dominate stag- nant, non-flowing, oxygen-poor sites and can tolerate saturated soils for long periods. Bay swamps and shrub bogs dominance types (20, 21). These conparatively rare floodplain environments strongly resemble their upland wetland counterparts. Field observations were made at two sites (Table 11). The shrub bog on deep peat (type 20) had an unclosed pond pine canopy and appeared somewhat raised above the flood- plain surface. The bay swamp (type 21) was moist from constant seepage. Tidal forest dominance types (22-27) (Figures 29-31). Tidal forest types occupy the floodplains of all rivers within the zone of tidal influence, as far as 32 km (20 mi) inland along larger rivers. Soils are peats, tightly bound by interwoven root mats (Figure 29). The water table is continually high because of lunar or "wind" tides. Herbaceous layers are remarkably diverse and little studied. These flat floodplains include higher "islands" or hummocks whose tops are all at the same level (about that of storm tides) and supporting species that occur on alkaline floodplains (type 26). South- ern red cedar (Juniperus si licicola) occupies the banks and higher elevations of the tidal forest floodplains along spring-fed (alkaline) rivers (Figure 30). It prefers a basic or high-calcium sub- strate. Stands of southern red cedar have been severely reduced in Florida by exten- sive logging by pencil companies (Wharton et al. 1977). The southern red cedar is an important component of the "hydric ham- mock," a seepage wetland vegetated by live oak (Quercus virginiana) and cabbage palm (Sabal palmetto), along Florida's gulf coast (Wharton et al. 1977). Atlantic white cedar dominance types (28-30). Atlantic white cedar (Chamaecypa- ris thyoides) is believed to be a distur- 44 .r. o (U LO LO -Q ^ E ■-« fO 3 ro r—* 3 Ol ■r- -^ 1/1 OJ O-J 4-> ■.- OJ 1 -Q -C 1 a> •<- ^ ^ (_) Li_ j=i - 1 3 :n O +J t/i Li_ KO Ll- fO a> oi S ■*-> ^ a> O) c «m O fC iT3 5 CJ> c: ■-- fO r-l ft3 ■o c s- 'oj OJ o O +-> (O CL O 0) -Q OJ OO O J>£ w 5 to u -— +J u -Q 3 OJ c . -C o (/I O) -c= :£ -^ fO OJ u J= o n3 S_ c X) O) fD ■— OJ fD ." na >^ 2 c OJ > - — < O (J £- Q 1 3 O fU Q..«— LT) a: z JU » CO .^ 1 ■5 u u. ^ ■M O (T3 < — - Qi J= OJ > t3 O -i^ O >— O 13 j:^ •r- _] 01 fO •• cu o cn i— C +J 0) "O 03 2 S- QJ -^ CL O O cn s- ■< -M Q_ O XI a; >!'— >)'- >>■- >)•— >>"- "O 1 — 03 i— fD .— »o f— m •— 03 .— (_) 1 — (_) 1 — u i— u .— o 5 (T3 -f- "O - "O m -r- -a ro •■— "D O -C O) O ^ U .£1 QJ O JZ cu •r- Q. M •r- CL M ■r- CL M •-- CL rsi •r- CL hj c C7) — o s- r— 1 — cn OJ 1 — cn OJ r- cn OJ i — cn rn <— cn 03 £ o o u o o u o o u o o u o o u O U CD O u a> o (_) a> o O HI o U O) o (_) UJ cni— LiJ Ol-— L4J CTli — UJ cn»— Lu cn.— ^ OJ 3 Ol o Q- +-> ■M S- 3 03 U 03 CU +-» ■*-> 3 •<- S -t-* J= 1 -4-> 1 OJ cn l_ I/) l/l f/y S •r- OJ -C t/1 -i— I/) 1 >> 03 -t-J LO OJ S- q; E i- S- 03 fO s- at u .— o ■M 2 Cl+J CL CD ■t-j i/> 1 C >> u >> U1 O -r- U fO u s- s- "I — Ji£ ■o i- -D 0) (U .— OJ Q. r— OJ r- 4-' -o .- Q. E OJ -C 03 03 c m 3 3 CO U 03 5 3 h— +J Q. 1 S- S >> OJ 1 03 JD +-> 03 .— 1 Q. (/) Q) tn 3 OJ ni 1/1 -M ■^ CD i- 1 S^ CD CU CL (D +-» o >i-£I 03 O O 3 (D >> -D CD 1 CL ^ OJ OJ •<- U CD S O^ t/l O "CO •<-> CD -D C I- CD OJ CL -— C >> O •<- U O E ■4-> O "O i/l -O (D s- •^ O (D OJ ' — -M CD CU +-> U CL 03 OJ 3 O _-l -M (/) r- O 03 •— (/I ■*-> O) S- CD CL. ,-c: o u TD CD r- CD 03 Ol CQ O •i— O Ul 3 E.— O CU LO +-> cz +J t/i C i— C -C Lu m u CO s- CD S- 3 O) O +J .— (O 3 CO CD , o >, OO S- S- -CO o OJ ■*-> ■o ^ E C C >, CL+J 03 > o 3 +j +J r- o 3 O c CD o 03 CD Ul J_ l/l c (D O u- ^ r- CD E u u 3 >» E O CD o O -o CD -M t/l CD (/) OJ O O •*-> 3 ■o m E c LO c: Ci CD 3 OJ O in c u i— X) o S^ to c CU c o. CD +J Q. O E -M 03 s- 3 CL o 03 Q- ■M to (J 3 M- CD -C +J c CD S^ +j 1/1 *-> O o O 03 U (/I 3 e: ■o 3—' 45 >,<\J - >*+-* ? <1> 3 TO 3: (/I -i.: ^ OJ TO X 31 1 CD 0 0 3 > •— .— 01 q; 0 <— TO >- U- t— -C Li_ O^ i- CD CD CU "O CD M -I- O OJ t. TO OJ I- 0 ^ 01.Q 5 TJ -C -J TO ■>- r^ U Li- +J S- 00 TO u ^ 0 >> TO • -cos Q. o1 e> csi n; TO 3 "O +-» +-> cu 0 1- TO<: •r- S^ -c 0 00 0 _i <: (_> J=) >, . — TO U TO "O U jC >•— >>— > — TO 1 — TO 1 — U .— 0 TO •.- -D 03 -r- 73 0 -C OJ U -c - 0 S- >— 0 s- <— .— cn TO 1 — 01 TO 0 0 u 0 0 U (J QJ 0 0 O) 0 LU Cnr- LU CJi'— r— TO f— U TO •>- "O u -C tt> •I- Q. N cn TO ■>- O I- <— 1 — cn TO o o u u a> o LU ail— .— 1- >tn Q. 0 OJ S- (/) S- .— 0 3 0 +J CL TO OP OJ .-H ^-> ■M TO >1 -!-> i. W1 S E u s^ TO ■r- (D S- CD 3 +-> Q. CD ^ - a> -0 ■!-> c: >> +J t/1 +j 0 O) V) E t/) Ll_ CL-M O) X) -0 <: 0 ^^ lO 3 TO S- 3 4-> 1 — OS-— ' CD S a; ■*-> 0 S- +J Q..— L. CD "O 3 0 OJ to (D 3 TO 0. CD C JZl ■*-> ■t-> +-> 4-J Q. >>-c: TO t/1 TO 0 TO u u 01 -C CD TO E E -— OJ JC I- +J X3 CL Q. TO TO TO cn wi 03 ■.- C 3 CD 3 3 CO 0 TO -1 3 TO -♦-> t/l CO i/l CL KTt 3 VI +-> > TO u 3 -O x: JD OJ -c in a. t/i i~. +-> 3 C TO TO -t-J Ol ■*-> .— -o +-> -C cn 3 -M -4-> to S- TO CD -C CD 3 S. LO 0. 0. in >, C Q. lA (J » 0 S- 1- u Q. > 0 4-> 0 > Q. Q. Q. to 1- 3 3 C 0 -t-j (D -t-J U "O to Q. TO CL i- E <+- E ^ E 01 TO c TD cn 3 <1> O O "O -Q I XJ C -O 3 I XJ U. TO CI, TO CD >>>— ' — •■- 3 -I- .— Q.«=C TO Q. CD +-> 3 (_) TO TO CD 3 1- 0 3 3 C CD TO *— 10 i/i 0 i- c (D 4- Q. CD >» "O - E ■•-» 3 4- c in 0 TO 0 ■!-> 0 TO -M CJ 3 x: TO 4- CD i/l 0 <— .— x: 3 CD r— 14- •^ +-> c 0 r- (D 0 -r- TO I- Q. CXT3 m 3 -C T3 E Ca. C C TO (D £ ■r- , 0) E TO E cn c ■♦-> s- TO 3 0 U C TO c TO 3 CO ■0 0 TO S^ 0 U in 46 (0 ro c sz o > •a: S C7> 3 t/i e3 >s •—« c V fO m tn c •.- E OJ ? ?J i^-— >,■- 1 — fO .— fD .— u .— U ro ■-- -D m •<- "o U ^ — ■ — C7) ra 1 — en (T3 Cn fO •r- 0 S_ <— UJ C7)>— LU CT)i — ^^ >> 3 1- JD •♦-> > •M E O" CD ..- o -.- o 4- ■U 2 2 TO c E > (U >, 3 3 -M J- >, I- >, > >, O -M O +J s- O o <0 C (/) C fO I- at ^ s- J3 m 1- ro QJ S CL •o u CP U ■•-> >i >, *> +j -o •— ' o ^ DC fO QJ in c o at i/i 3 .— -o Q) #. (TJ 5 O (/) -o >>-Q S- -O 13 Q. E •<- 3 fO S- -Q s- 03 C S-. 0) ro i- 1- 3 OJ Z3 i -C t/1 -C -t-J > Q Q. CO "O O) LJ in >1 O 3 O) c +-> >> C J- OJ c OJ ro m OJ OJ S- E a> oj X) QJ 4- CL 3 Q. S_ Q. s- >, cn en ^J cni— u +-> -a S- Q. at S- ro -o at c a; Oi a) OJ >> .— QJ o > at 3 > o 13 3 a. OJ T3 oo OJ t. CQ in Q. 3 E (O o 3 -M .— JD 00 <— >j cn C fO i- .— at ••- c (D QJ S- ro =) E at E dJ o o -M at O I— C O fO M- ro =J U e — * t/1 c 5-^T- to ai «a: .Q o J3 S- CO ro CL • -O s_ >, LO QJ "O (J u >— U OJ +J ro , c o S- 5 to en :c CO Q) S- -— ^ -CTi OJ ■o r«^ O) CTt +-> ra s- 0) -.- OD *~> CM i 3 .— -M m ^ 5 CTl s- >> 5 O -:><: 3 3:.- tJ -D S ^ s- a; .^ u. E c IE U 3 OJ —I (C 3 — - cu •»- m en s- u_ 3 O -Q S 3 ^ T3 (/) •♦-> 1 — -f- (_) O > s- (U fO fO 5 5 s- H- Q.__l 3 T3 O m o <: u. S- CO 4- _l CM L. -o >- X) S-cC 0) CO -«-> (O ■^ (U O) cn -4~> -o 5 OJ O C71 o >, >>■- >>'— s^ — >>■— >— fO .— fD .— *T3 1 — m <— o .— u .— (_J >— o (TJ ■<- XJ rO -^ "O ra -r- "o fX3 ••- TD O ^ 0) O -C OJ U -C QJ (J -C Qj •.- CL (SJ •n- CL M ■.- CL N •.- CL Nl CD 03 -r- CT> - cn ro •>- o s» >— O I- <— O I- .— a i^ f— -— cn 05 . — cn fD . — C71 (D 1 — CD TO o o o o o u O O O o o u U 0) o u — u 03 ■.- -O TO -r- -O u -c: Ol 0.^01 •.- CL N •>- CL rsj CD m -r- CD TO •<- O i- .— O S_ ^ r— LJ) fO .— cn TO O O O O O U o ^ -— TO f— U TO -.- -O U -C OJ •.- Q. JM cn TO -r- o i- >— . — CD TO O O U U OJ o LU CD.— , 2 2 S- TO CL-^ . — < QJ tn >iXJ >> t/1 QJ O -M TO Q. 1 TO TO TO "O 1 > >, C S- U OJ 3 O -Q U .O C to O TO TO > c +J TO CO -QUE S- 'oj +-) S_ +-) U at CD J=t >i CL i~ Cl. Ol O) 4-> J= 3 V- 2 TO QJ >» O) c TO W W -M (/) o. u UJH 3 C ".TO o ? Q. TO t •.- 1 3 CO ■>— QJ CL 1 E O -M O QJ "O o i- 1 C I — 1 Irt TO C 1— CD OJ .— .c S^ CO — s- lA 2 -- a> > QJ TO S- QJ to TO -O to s_ TO Qi irt -:■; Q. CO s- CLJH CL "O 3 to >1 ■o U 1 CL 3 c: o 3 jZ) C 3 -C > (O E 4-> a» ■M 4J TO I- +-> +J U .C u u >> TO C/l 3 ■o (/I O QJ to u\q O) CL a. s- a. -C Q- 2 0) X TO QJ ■o 1- 1 E -c QJ E -C -M E •t-' +J 1 — OJ -r- ■M ■D CL >, TO ••-> -o TO -M 3 TO >^ TO TO OJ >, TO 2 — c 2 •.- O 2 TO OJ -C TO •— ' O -C CD S. <_) J=t t/1 2 3 cn 3 CO to -Q Q. 3 +J^' **- 3 3 +-> O Q. 4- O E >, O J^ TO .— 3 QJ c/1 •• CO to QJ -C O -C cn >».— u -^ .a u TO -c QJ -D ♦J t- -M QJ (/) C ■M O I- OJ TO E QJ C C 3 O •r- CO O Q. E -o ^ E O C O ■o O C U Q.-i- >, CO .— .— QJ 00 •l- .— i- QJ O TO i- OJ CO 3 O S- CO U t— U OJ 4-) X: QJ -r- ■M E C C •r- «T ■.- S- CD 4- >> TOO) -4-1— QJ S- to TO I — C C TO •M O QJ 3 t- 3 •-- C t/1 O JD -M ■.- 3 TO S- 4-> +-> 3 TO "— • TO C -l-J E fl QJ U r— C to 3 **- -f- QJ Qj ■ — O TO > S- 4- C O CL TO CO 3 .— C 1- CD TO 3 TO QJ C "O TO C ^ 3 : u 2 E O ■M C O E •.- O s- re 2 •.- 4- 3 +-> to CD TO QJ _^ C S^ CD U •>- 3 TO TO E ■*-> CL.Q i~ TO QJ O CO Qi QJ • *«- to C'—^ •r- O ■i-J +-» >v i- CO TO C J3 QJ QJ TO > QJ Q. C to •.— Q. E S- >> "i TO H-^ oo o <3J C- — +-> -o U •»- {A O +-» XI QJ U to s- c -O -Q CD CD *3" Ol 3 C . V) >, 4-> TO - to JD O _ QJ .-« O ■ 1- ■♦-> - O OJ QJ • 4- Qj Q.-0 3 >> QJ »— CO *-> M- TO 1- "O "U Li- TO « ■- C«t 2 I— TO CO "O OJ TO • 4-» -d i- . J C •.- QJ +J ) TO +-" >— C J C i- TD QJ - •»- QJ "O to . E -M •.- QJ J O C 4- S- 1 T3 ■«- "■— ' Cl u 0) o n ■f- QJ ■»-> >^ .c > TO (— .c o I. 3 -o ■*-) QJ to (J C ■M c T) -M TO OJ r ■♦-> .c t J ■*-> TO C Ti C c- o ■*-» TO •a SI ^ E to TO 48 m^ .ife*». Figure 23. The dark area below the pencil is a liverwort (Porella pinnata) zone, the upper boundary of which indicates that high water has stayed at that level for at least 16% of the year (58 days, not necessarily consecutive). The zone above it is a green moss. The upper boundary of Porella growth is helpful for quickly determining depth and duration of flooding. 1^ y4t^ - .^■ Figure 24. The outermost swale (next to the upland) along the lower Roanoke River (NC) supports almost pure stands of either swamp or water tupelo on black, muck soils. Here, in a water tupelo stand, in response to the anaerobic muck, roots form blocky aerial knees which resemble black rocks. 49 Figure 25. These Ogeechee tupelos on the Apalachicola River floodplain (River Mile 15.6) are tall, straight, and large, measuring 1 m DBH (diameter at breast height) although along many Coastal Plain rivers they form grotesque, many-trunked stools. In the forest pictured, water tupelo is a co-dominant, and strap lily (Crinum americanum) is a characteristic herb. Figure 26. Scenic Ebenezer Creek (Effingham County, GA) is a unique variant of Zone with a deep lake-like channel. Cypress buttresses are abnormally enlarged. 50 11, i *r> Figure 27. Drydown in water tupelo (Zone 1 1 ) on Ebenezer Creek (GA). These backwater environments are dominated exclusively by tupelo and cypress. During drydown, nutri- ents concentrate, and duckweeds form dense surface layers. Note: (1) the precise height of the high water line (approximately 2 m or 7 ft), (2) relation of buttress swell to high water mark, and (3) counter-clockwise "swirl" of buttresses. Figure 28. The outermost backswamp (Zone II) on many alluvial floodplains is dominated by a swamp tupelo on acid, highly organic muck soils. Here on the Escambia (Hwy. 184 bridoe, FL) the shrub zone (Cyril la, Itea) is sparse but prominent. There is a clay aquiclude several feet below the surface mucks. Seepage moisture from the adjacent upland, as well as rainfall, may be important to some of these associations. 51 Figure 29. This tidal forest (Zone II) of sweet bay, pumpkin ash, swamp tupelo, and cypress at West Pass (Suwannee River, FL) has a characteristic interwoven mat of large roots close to the surface. This extremely tough layer protects the forest and the shore from the destructive erosion of constant wave action and storm tides. There is no natural levee. High tide comes nearly to the top of the root crowns. Fiddler crabs and olive nerite snails are abundant on the forest floor. Showy herbaceous plants such as iris, butterwort (Senecio glabellus) and aster (Aster vimineus) are surprisingly common. 52 Figure 30. Three shoreline dominants, sweet bay (B), red cedar (C), and cabbage palm (P), are characteristic of the tidal zone of an alkaline blackwater river (St. Marks, FL). An intertidal zone (T) can be seen between the root zone (H) at the high tide line (partly in shade) and the dark water (W), which has a band of submerged plants (S), here partly exposed. 53 Figure 31. A small grove of virgin cypress is preserved on Lewis Island (Altamaha River, Mcintosh County, GA) where they grow in unique association with both sweetgum and water tupelo. Such giant cypress were characteristic of the upper tidal zone of the great alluvial rivers of the Southeast. Apart from the grove has a moderately dense herb layer including Both it and sweetgum are typically found in Zone IV. large swollen base. 54 a few shallow sloughs, most of the swamp palm (Sabal minor). Cypress here do not have the bance-adapted successional species. Fire is the most common precursor to white cedar development, though flooding, wind- throws, or logging yield the same effects. Atlantic white cedar is usually found in bog stream swamps on peat overlying sandy soils that are characteristically poor in nutrients, in a unique tidal forest type, or in acid backswamps of certain Florida rivers. Dominance Types of Zone III Zone III includes the wet flats, bank-edge strips, low levees, and depres- sions in Zones IV and V. Dominance types (Table 12) in this zone are semi-perma- nently inundated for a major part of the growing season, as well as in winter and spring. Although the hydroperiod is long (about 6 months). Zone III areas are sub- ject to annual drydown. Soils are satu- rated 40% of the year (Leitman et al. 1981). Pioneer dominance type (1). The banks and point bars of the southeastern rivers often are occupied by the black willow (Sal ix nigra) and other species such as silver maple (Acer saccharinum), and some- times Cottonwood (Populus ~ deltoides). These early serai stages are succeeded by Zone IV types as elevation increases from soil accumulation. The successional se- quence is a function of meander movement rates and point bar formation. Rivers with intact forests on fine cohesive sedi- ments migrate so slowly that mature forest establishment keeps pace with the river channel, and pioneer stages never develop. Swift meander movements over unconsolida- ted sands produce tapered slopes on point bars, and several serai stages may be found. Shrub, small tree, and herb dominance types (2-4). Semi -permanent pools occur in depressions, old oxbows, and scour channels. They are dominated by several species of willows, shrubs (e.g., may haw (Crataegus aestivalis)), and small trees (e.g., water eTrii [Planera aquaticum)). Overcup oak-water hickory dominance WW. types (5-10). The most poorly drained flats of the floodplain, in which water stands well into the growing season, are characteristically dominated by the over- cup oak-water hickory (Quercus lyrata- Carya aquatica) type (Figure 32) and its variants. These flats are relatively small (about 2 ha or 5 acres) in the Southeast, and seldom are dominated exclusively by these two species. The wet flats of the Congaree River (SC) are dotted with numer- ous depressions, so small as to be occup- ied by a single overcup oak. Overcup oak, undesirable for lumber, often is left by loggers. A near-virgin stand of overcup oak on the Santee River floodplain (SC) contains trees approaching 1.2 m (4 ft) in diameter. Additional sites occupied by this type include small shallow depres- sions in Zones IV and V, and narrow bands bordering deeper depressions that contain cypress-tupelo or water elm. Both overcup oak and water hickory avoid seedling and sprout mortality from inundation by leaf- ing out late in the spring. Both species reproduce well; overcup oak through con- sistently good acorn crops, and water hickory through good mast and prolific sprouting (Eyre 1980). Water locust (Gleditsia agjjatica) -water hickory stands are rare variants of this type. The ex- tended hydroperiod in the sites occupied by overcup oak-water hickory and water locust-water hickory forests inhibits herb growth, and thus the understory is re- stricted to small trees and shrubs (Eyre 1980). Dominance Types of Zone IV Zone IV (Table 13) forms the bulk of the floodplain on Coastal Plain alluvial rivers above tidal influence, chiefly on flats or terraces of low relief. Two irregularities are common: "washboard" terrain caused by parallel scour channels (often sandfloored) and "hummocky" terrain where trees either stand above the general floodplain level on hummocks or have tor- tuous scour channels around and through them. Zone IV is seasonally inundated or saturated for 1 to 2 months of the growing season, and more or less continuously inundated during winter and early spring. Soils are saturated about 22% of the year (Leitman et al. 1981). Shrub and herb layers are scanty. Stiff clay soils or subsoils act as aquicludes which pond rainwater on alluvial floodplains, while the more porous sands dominating black- water floodplains preclude this ponding. The diamondleaf oak (Quercus lauri folia) appears to dominate both the alluvial and blackwater floodplains. It is remarkably 55 1/1 >, >> >> U- :s ^ U TD o 4-J -^ - o 'O S_ QJ to CD -C JD OJ -C s -C rO * rO O u E -M +J C\J o na c U o O >^ c -C 2^ ro <: <: u_ CJ ^ CJ O X> I — iTJ 5" O) ^ -r- 3 > fO u_ <— CM TO -M OJ U- rts 2 LD 1 — (T3 +J Q.^-^ 1/1 > o fO ra t3 {_> ij 03 -C O) 2 "- -— c/? 3 t/i to o -o +-> n: -c O t_J t>o o ■*-> -r- ^ O _J O -"OO U _J 'O s- 1 +-» TO U_ ■•- t Q -M •<- U- ro -c j3 (U Li_ TO •— -C Li_ a> c -C -c: 2 OJ 03 •- (J OJ U OJ u - fO TO O C CM C *4- O «=C 4- .— C TD C7) 3 <— U fD U TJ 4- ■»-> O rO 5 -,- C C fD (U -l-J O >, 2 3 C •* -M Q. O i- O O CL J3 <-J s: < Oi <: o C_> DI t>o CQ t/1 CO C_J >-)^ 1 — O >— u TO ■r~ "D U -C CD U ^ OJ Q. fsj •.- Q. Nl Q- TO •■— cn TO 'I— O L. .— O s- »— C7) TO .— en na O O U o o u LU CTi.— ;- Q. NJ u: TO -r- O S- r— C7^ TO o O O o > 1 — TO TO •^ -o U -C Ol Q. N cr TO •>— o t- r— C71 TO O o u w > >, J +J -C ■M u O Ll_ c <: -C LO 2 in ■o o to C -M O o Q-4-) o (/) CL i- a> O) to c > o o TO r- QJ (-J tn JD -— • to OJ OJ— N u E >> CL O f— 0) Ul . — 56 Figure 32. Large overcup oaks occupy depressions (Zone III) in the dominantly Zone IV floodplain of the Congaree Swamp National Monument. Photo by George Taylor. 57 > C •<- o f^ ■o .«> S- (/I to s- <: jD a> ir> c >> 13 C3 CO s- CM »x> 03 OJ >, CM 5 C +-> QJ CT 3 c a: — ■ m o^ ■O IE ^■5, +J .* CL +-* S-. -M =n s- fo cC E <: TO ja ra cc O *T3 o o +-> > OJ 3 , ^ CD OJ fO o sz cu OJ .— u cn-o en E s- u en OJ s- OJ "O >— ■o ro .— -.- > en ■*-> •- 3 fD TD ■4-J o «c n: C CO C3 o S- Q- CM OJ 4-J -Q CZ -^ OJ E >> oo CD S- OJ -J (D C_> 3 ■4-> CJ o xa -c ll 3 i/^ 3: c: C7> 3 t/^ 3 o t>o o cz <-J fO ■M \^ +J - ■M E ^ E ^ 3 -O S- > OJ u TO CD 3 TO ■■- rO E > Q_ U d TO CD n:: +-> i^ CJl 3 O (-J (_) Q. C (J JD C C U o -o O O O ct U C ^ O t-> s: — 03 I— U rry 'r- TD (J -C O" •I— Q. IM cn TO ■'— O S^ .— .— C71 TO O O U U CD O uj cn»~ s_ 0) o >^ t- >— TO >> r- ct: TO ' U TO .,- -Q (_J ^ OJ CL N cn TO ■<- o 1_ 1 — cn TO 'o o u u CD O UJ CT^ >> .— TO U TO "CJ o -C (D CL rsi cn TO O S_ cn TO >> TO . — >,^^ >^'— ■— TO -— TO ,— u TO U X) TO •<— "O U •<- CD U -C CD ■1- CL tsj ■■- CL M cn TO -^ cn TO ■■— O S- .— O S- ^ .— cn TO ( — cn TO o o u o o o UJ oil — 3 >> >> CL E TO O 3 -c: Ul OI 1 1 — cz .^ O. TO TO TO O O E -.- s- 4- "O CD TO (D E CD i- •=£ TD JZ. .c C +-> +-> o E 3 3 3 58 i/i I — cr^ O — ' >i Ll- 3 o cu - fO -C +-> (J s^ OJ QJ O LO -C s^ (_) t/1 ro o ^ JD _1 +-» _l -— OJ -— Ol -C i- 3 1- I— u a fO fD -t-> o. a; 3 <: cr: (_J "D +-> i_ o fO u o Ul XI fO O) OJ CO cn >•— >!'— >>■- 1 — fO <— m .— TO ■ — u r— U .— u fO ■■— "O fO -I— "O «T3 ■<- "O <-i jc: OJ U -C OJ <_> -C OJ ■-- Q. fsl •r- Q. fsJ ■.- Q. ISI £= c cn (T3 ■>- cn fD ■'— O) t3 ■<- o o s_ •— O 5- r— O S- f— £E f— cn (D ■ — cn (i: (U 1 — cn fD £: c O O (J O O (-) 1- O o u Ol •— « 13 -D <— _:«: .^ eI ro C O u TO 4- u JLJ tT3 TO TO (U >• Ul 4- QJ JO 4- M- •<- TO ■— OJ TO TO U QJ .— . CO OJ OJ OJ -— Q. -Q -c j:i ■o TD i/l C +-> -M c O C o o > i 'i 5 E OJ e: TO TO CL TO "O 3 •-— C O TO -O O -t-J Q TO OJ TO ■<- U O -— •I- a. 5 59 wet-tolerant, occasionally found as a co- dominant with swamp tupelo (type 10), and rarely can be found mixed with a few cypress. Floodplain flats dominance types (1- 10). The diamondleaf oak dominates the Zone IV flats of all the major river types. Even so, these forests are more diverse than the wetter overcup oak-water hickory types in Zone III. Frequent associates in this zone are green ash (Fraxinus pennsyl vanica), American elm (Ulnus americana), sweetgum (Liquidamber styraciflua) (TTgure 33) and, less com- monly, sugarberry (Celtis laevigata). The swamp palm (Sabal minor) (Figure 34) is a good general indicator species for this zone, as is possum haw (Ilex decidua), Walter's viburnum (Viburnum obovatum) and various hawthorns (Crataegus spp. ). Occa- sionally, the spruce pine (Pinus glabra) occurs although it is considered to be associated with lowest elevations of Zone V. It has a wide moisture tolerance and may occur with diamond! eaf oak on both alluvial (type 9) and blackwater (type 10) floodplains. Dominance types on other sites (11- 17). Zone IV oaks (diamondleaf , willow) occur in a variety of other situations: scour channels (type 13), ridges of ridge and swale topography (type 14), and the lower elevations of relict dune ridges (types 15, 16) where they occasionally mix with wet variants of live oak. Narrow sandy ridges (type 15) may bear shrubs that are not usually considered wetland species: wild olive (Osmanthus), yaupon (Ilex yomitoria) and saw palmetto (Serenoa repensT The reader should be cautioned that delineating Zones IV and V on some black- water and Piedmont rivers can be confus- ing. Due to the sandy soils of some blackwater floodplains, microedaphic and microtopographic mosaics become even more divided. On some floodplains (Zone IV, type 10) an apparent mix of Zone IV and V species may occur (Figure 35). On Piedmont floodplains, owing to the numerous scour channels and fast-draining clay soils, there also may be this apparent mix of Zones IV and V to the casual observer. Dominance Types of Zone V Zone V comprises the highest eleva- tion floodplain associations occurring on old natural levees, flats, higher ter- races, and Pleistocene ridges and dunes. Inundation averages once yearly (Congaree, SC). See Figure 11 for graphic example. Duration of flooding ranges from 2% (5.3 days) to 12.5% (33 days) of a 265-day growing season. Soils are usually sandier and less fertile than those of lower zones. Zone V dominance types observed in the study area are listed in Table 14. Zone V associations appear to dominate many Piedmont floodplains; however, in the Coastal Plain these associations may be restricted to 5% to 10* of the floodplain surface. As in Zone IV, the plant asso- ciations grow on both Pleistocene and Holocene floodplain surfaces. Understory species are more conspicuous in this zone. In fact, two understory species, the paw paw (Asimina triloba), a subcanopy tree, and river cane (Arundinaria gigantea) are generally good indicator species. River cane is most luxurious in this zone al- though dwarfed stands grow in Zone IV. The diversity of both herbs and shrubs is maximal in this zone. Zone V flats and old levee ridge dom- inance types (1-11) (Figures 36 and 37). Two hardwood species are characteristic and widely distributed: swamp chestnut or cow oak (Quercus michauxii ) and cherrybark oak (Q. pagoda). Water oak (^. nigra) occasionally occurs as a co-dominant spe- cies in these associations. Two pines are present: spruce pine at the wetter end of the spectrum and loblolly (Pinus taeda) at the drier end. In the Congaree, record loblolly pines grow on old levee ridges slightly elevated above Zone IV surfaces (Figure 36). Spruce pine seems to require a more continuous water supply and even occurs on upland slopes under seepage conditions. Some species that are wide- spread on the uplands apparently can adapt to the floodplain conditions of Zone V. Some hickories are common in Zone V over clay-rich subsoil sites (types 6-8) and, rarely, form hickory flats (type 6). 60 Figure 33. The sweetgum is a long-lived component of virgin alluvial bottomland forests throughout the Southeast. These giants in the Congaree Swamp National Monument (CSNM) occupy the higher portions of Zone IV floodplain. Lower Zone IV areas may lack the larger trees. The CSNM contains a number of national and state record trees. Figure 34. Many Zone IV bottomland hardwoods on Coastal Plain alluvial river flood- plains (Ocumulgee River, GA) have an understory of Sabal minor, a dwarf palm confined to Zone IV floodplains. The similar but rarer needle palm (Rhapidophyllum hystrix) also occurs here. 61 1 Figure 35. The floodplain of the blackwater Canoochee (Fort Stewart, GA) is somewhat anomalous in the co-dominance of diamondleaf oak (Zone IV species) with either spruce pine or loblolly pine (Zone V species). The herbaceous ground cover is dominated by river cane and grasses such as Panicum rigidulum and Erianthus strictus. American holly is present. The shallow flooding and permeable soils are probable factors allow- ing for a mix of Zone IV and V species. Diamondleaf oaks here grow more slowly than those on the alluvial Oconee floodplain. 62 en Qj +-> ' — O fC O S- 00 >— CJ (13 00 E eg oj >>^~ ID t 3 -M C to o :£ c: o 3 i/l o -^ uo 4-J E -t-J (TJ "O ID CU 4-J 0) ^ 01 c >^-'- S- 0) > a, Q. ■D E o— E Dl 3 (J (Tj C C -— <: 3 i-J o o CC CD (/I oo 0 s; >) *•.-■.- >>cu ■,— 3 (D 'O C7> c 3^^ -^ IE QJ ^ S- o ^ f^ cn.— -r- Lrt — cC ■4-> .— OJ > CL+J +J C3 3 t3 3 1 — ID C fO E >1 O *» •« QJ (_) -M — •> E >> QJ 0) QJ O +-' t- ■M -D > Dl OJ C U (B C 0) o -o d O ■" tS QJ (_> --- o U ct S- C .— Q_ >— s_ U ct O O t— -^ Li_ ' «£ -O O CD ro CO c , -M CTS 03 •f- n: fO oo o S +J CJ> 00 - U1 l/l +j -^ (U C "O OJ — CT) fD O O O E >>■— >)■— >— "3 ■ — fD >— U >— U ra •!— "D fO -r- "O U -C CD u j=: OJ -.- CL M •r- CL Nl CT - ro OJ >, CL a. +-> r— 1 "3 i— >i Q- 5 O C E 63 t— t «T3 ■*-> tJ CDCO 1 .— 3 ^ I- 1^ I— I «D t_J LJ_ C\J 3 — :c •- o 2 i>o ^ 2 >i O - O CD -. fo E > 03 x: <— -O 03 a> 3 o >— o ecu "3 fO CU S. O o o «— - JD U <-J ^ ^ C3 u I- ci QJ t>0 W fO (— e) (/> j:3 3 <— - » s: o s^ * n Q -U (U -C 3 t-J OJ <— t_) o Q O) OP i^ O Ll_ >> s. OJ JD u -o +? " 3 a. ■i- c - s- Ol CJi CO JZ 'D OJ Ol -ii •.- Qj t^ O »— U O 3 fO +-' >» fo m •r- o O t_) — V Q. dJ ^ Q o: z -J 31 < oc >>■— >.•— S*!-— >>•— >>-— >>'- 1 — fD >— (O r- to •— fO 1 — fO ( — ra 1 — O •— u >— u •— o -— u i— u m •<- "O to ••- -XD 03-^-0 fO -1— "O fa •.- -o TO ■»- TD u ^ QJ u -c a> U JZ ■-- Cl N •^ CL IM ■^ a. M ■.- Q. N -.- CL rsi •■- Q. N c CD fO ■>- cn 03 •<- cn fO •■— cn (C ■«- tD^ m ■■- t^ < — LjJ oil — LU oil — UJ cn.— UJ Oli— Lu cn»— OJ o u ■!-> 3 V) L. s- CX 0) XI o o 4-> TO 3 CL C +-> CL 1/) E o •<— 3 OJ c ■o 01 TO TO a. "O O o c o Ol TO Li_ E 64 Figure 36. Although normally an upland species, the loblolly pine also grows in Zone V areas of many floodplains. These large examples occur on old levee ridges and low ter- races in the Congaree Swamp National Monument (SC) only a foot or so in elevation above the lower Zone IV floodplain. Photo by George Taylor. 65 Figure 37. An old levee ridge near Cedar Creek (Congaree Swai ,p ".ational Monument, SC) supporting Zone V vegetation. The large tree to the left of the figure is a cherrybark oak. Paw paw and spicebush are common in the understory. On slightly higher parts of this same ridge, beech occurs. Some upland herbs (may apple, broad beech fern, wild yam, poison oak) may occur. Photo by George Taylor. 66 Low ridges with certain edaphic con- ditions sometimes support a few white oak (Quercus alba) in Zone V associations. Although tulip poplar (Liriodendron tul ip- i fera ) is extremely rare in the study area, it can occur in Zone V associations. Dominance types of other sites (12- 16). Beech (Fagus grandifolia) or beech- magnolia hammock is frequently the first association encountered at the ecotone of floodplain and upland. A beech "fringe" is characteristic of many Piedmont allu- vial floodplains. Since beech-southern magnolia hammock also exists under seepage conditions on the uplands (usually adja- cent to the floodplains. Zone VI), the high water tables of Zone V often enable these species (primarily beech) to grow on "islands" (types 12, 13) and old natural levee ridges (types 10,11).. Southern mag- nolia (Magnolia grandif lora), however, is rare on floodplains in the study arpi. In areas of ridge and swale topog- raphy (type 14) (Figure 38), Zone V asso- ciations occupy the higher elevations, succeeding the associations of lower ele- vations. Scour channels, because they allow rapid drainage, frequently contain Zone V species mixed with those of Zones III and IV (type 16). This is similar to the occurrence of live oak (and saw pal- metto) at the "lip" or edge of banks and scour channels. Figure 38. Narrow, long ridges between swales on the lower Roanoke floodplain (NC) of probable late Pleistocene age have an almost diagrammatic zonation of Zone V hardwoods beginning with diamondleaf oak at the edge of Zone II and progressing up through swamp chestnut oak to cherrybark oak. 67 Plant Communities on Natural Levees, Floating Logs, and Stumps The zonal classification scheme does not make specific provision for the plant communities that occur on natural levees and on floating logs and stumps, and thus they will be discussed here. Natural levees. The height, width, soil texture, and drainage characteristics of natural levees vary considerably, often fostering the highest plant species diver- sity on the floodplain. Species character- istic of all floodplain zones (II-V) com- monly occur on levees, not only because of the differences among levees, but also be- cause of variations on individual levees, which are often a mosaic of microenviron- ments (Radford et al. 1980). Recently formed levees on nontidal reaches of alluvial rivers support pioneer tree species, particularly on front sides (cottonwood, black willow, river birch (Betula nigra), silver maple), while mid-seral species (sycamore (Platanus occidentalis), sugarberry, American elm, green ash, and sweetgum) occupy stabilized levee ridges and backslopes (see Figure 21 and the discussion of Zone IV dominance types). Boxelder (Acer negundo) and catalpa (Catalpa bignonioides) are pioneer species that seem to prefer levees to any other floodplain sites. River birch is often the dominant on sandy Piedmont levees as well as on disturbed flood- plains. Baldcypress (Zone II), and overcup oak, water hickory, and water locust (Zone III) are found on low stable levees. Higher, well-drained broad levee ridges, such as those on the Roanoke River in North Carolina, may host Zone V species, including swamp chestnut oak, cherrybark oak, Shumard's oak (Quercus shumardii ), paw paw, and spicebusFi (Lindera benzoin) (J.M. Lynch, Department of Community Development and Natural Resources, North Carolina Heritage Program, Raleigh; per- sonal communication). Live oak frequently occupies the high river front edges be- cause of the "dry lip" effect discussed earlier (Zone V discussion). Tidally influenced forests, like those found on the St. Marks River (FL), show zonation on the present levee (Table 15). The drier river front is dominated by live oak and saw palmetto (Zone IV); the levee top supports southern red cedar, cabbage palmetto, and sweetbay (Zones II and III); and the backside contains inner swamp species such as swamp tupelo in ad- dition to dahoon ( 1 1 ex , cassine) groundsel tree (Baccharis glomeruliflora), cabbage palmetto, southern red cedar, sweet bay (Magnol ia virginiana), and wax myrtle. There are distinct differences between the communities that occupy old levees and the present developing levee, primarily because of the changing hydro- period. The trend is for older levees to become dominated by species characteristic of drier sites as the floodplain geomor- phology changes. Good examples are found in the ridge and swale topography of sections of the Roanoke River (NC). The ridges (old levees) show distinct zonation from Zone IV species (primarily diamond- leaf oak) near the swale edge, through wet-site Zone V species (swamp chestnut oak and cherrybark oak), and finally to dry-site Zone V loblolly pine. Floating logs and stumps. In addi- tion to the communities of Zones II-V and natural levees, a unique flora sometimes occurs on floating logs (Figure 39) and stump remnants. Dennis (1973) described such communities in the Santee Swamp (SC). Twenty-four species were noted, eleven of which did not occur in the larger survey of the swamp. The community samples were homogeneous, dominated by Boehmeria cylindrica and Hypericum walteria. The selective forces acting on fallen logs and stumps are uniform and severe, efficiently eliminating species that cannot tolerate shifting conditions of Inundation, expo- sure, and possible substrate instability. In addition to Dennis (1973), Conner and Day (1976) noted several plants that grew on rotting logs and stumps, including ferns, strap lily (Crinum americanum), Hymenocallis eulae, spiderlily (H. occi- dentalis), Hydrocotyle spp. , southern wild rice (Zizaniopsis miliacea), and Panicum spp. Understory Species A structured understory community exists beneath the floodplain forest canopy that may rival it in species diver- sity. Of 110 species found on the Santee 68 tu OJ -o OJ JZ ns c ^- E o U) • •r~ SZ >— 1 o U >-' •* QJ ID 1— 1 U OJ O) C 0) v a; t|- C 3 s- O O cn Q) r~j •■- ^- ly) U. O) C 01 s- O ID C 1 •1- ^ •1- m W) ■O JD +J 0) ID to <— 0) (— ZS JD x: >— IB •M 1— o • •r— •!— C ^ 1 — ■r— V) Q. OJ +-> Q. >=1- Q. U ID -H >, I/) OJ f-H C S_ ■-H 0) ID O) U S- ^ to c ■•-> 3 OJ ID e J3 •■- •r— • ID E (O O 1— o r- "It -a o. c TO OJ •1- o O 1- f— O 3 S_ Ol .— Ol ID Q. >+- •.- 0) 3 u. a.-t-> OJ a. 1 c c ID to •1— 'r— to e >> a; (/) >i +J +-> (-> +-> to "O (/) (U ID 1 — • CO >— ID =» O to -O S- 0) ■t-J QJ c: - C (O -r- ID'— o lO •1- I— 1 r>l 1/1 s_ ^ c q; ID *• O-c > O) •* ■I- ■)-> 1 — 1 +j l/l JD 3 O OJ ID O) J3 +-> Q.I— r~- >>— XI S- S_ ■»-> ID 4-> OJ t— » 1— 00 4- 0) — . ■.- ID E t— 1 OJ c C O Q. O •.- I->J O) >>■•- E c +-> ■!-> o o o Q.-0 -M Pvl QJ -r- U S- O to •s C U •1- S- CO ID (/) 4- Q. XI -o c to — ~ ID o t— ) (— +-> N O 1 to ■(-> >— 1 • «i , O) QJ ■(-> c JZ 1— +J Q. QJ Ln c E (J t-H to Oi ID C s- U X ID Ol OJ ID 0) C r- +J a. XI ■•-' s- i ID 0) c: o o (— r— •r- Ll. -o QJ Ol > to QJ --^ .^ .— > s- 1 iD I. >■ E QJ .— Q» QJ > 1 Q. V. •r— l~i o i- '—• r— oO ■t-J to 1>1 C l/l XJ QJ QJ QJ c 4-J to c: ID • r~ QJ O p— U) S- M d. Q. ' => S- OJ ■ — N '■"^ o IvJ to 00 ID 1 — (-> s- C OJ o > •r— •1— -t-J s_ ID U »\ O s- 1 — QJ > ■o •1 — c Qi ID ID u > S_ QJ OJ QJ r^ -t-J QJ cr> ID ID E ' — ^ > TT >, S_ S 3 CO QJ ,—. 3 • 1 — 1 1— o S- .— ■t-> -C c > f— t C— ID -t-J -I- >, QJ 1 QJ .— 00 c: •1- o C E > QJ XI 1 S u •1— ID O ■— > O > -tJ ID • r- -t-J to QJ .— ^^ c -> -t-J 3 JC QJ QJ 0| J3 -C (J x: QJ to QJ «-r- •1- -M QJ -O -M C Ol > ^ ■— S_ -1- O) S_ QJ ID 1 — 1- S XI O S Q. Q. .— O QJ to to -t-J I ID > QJ U E QJ 3 -o 1— > i. C ID •1- Q. ID Q.'— «t 1 — (/) r-~ OJ ». Q.1— t QJ .C - QJ E 1 > -tJ ^ C ID 5» QJ •r- ID •■- S — • S O Q. t/) — ' O r- O 4- CL .^ 1= lO O — o •.- to 00 1 Q.'— to ■— 1 3 tn QJ —1 (_) 1 S_ 1—1 S_ 1—1 Q. — - QJ 1—1 QJ > 1—1 O O— ' LlJ -t-> QJ QJ 3 e ID Q. cT 1 CL 3 s- s_ 1 o o E QJ QJ to ■— ID O -t-> -t-> to QJ QJ 3 -t-J -C ID ID '— QJ CL CL to -t-J 3 3i~- I- 3 3 .2^ •t-J Q.-t-J *-> U e -■ — ^ 3 -O -O 1—1 >, ^-^ ID QJ CTv QJ C 1-1 tj 1. f— 1 CL-O (J 1— 1 »« 1 — ID — - -O QJ 1 E ID 1 ■a o r— +J 1 — 1 ID >, ••"D 1 — 1 c o -£= E ID ID 1—1 s > 1 s_ Ji£ o un ID ^ 1—1 O (J 1 • 1 — 1— 1 Q.-C 1—1 3 ^ — ' (J s_ U QJ -t-> QJ -t-> ID > ID O 3 4- .p_ •!-> QJ 3 E •1- -C C 3 00 •1— l/l QJ ID r— to '-^ x: - — > QJ QJ S- CM Ol s. en i. S. QJ 1 3 QJ 1 O Q. -tJ 1— 1 O -t-J 1—1 -C >, ID 1— 1 1 — ID 1—1 to U 3-~- to 3 — u (_) to «l QJ r^ QJ ID S_ •r— ID > Ol 3 e 1 — o r— <_> ID #> QJ _I •■ -^ U. i — ID ID 1— •1 -r— OJ > 1/1 QJ 3 - C 1— S- o 1— QJ .^ ID -*-> u 1 S_ o ■r- O r— to Q- x: ID o 3 -tJ o O" ID 69 i- n B 01 i- 3 O o 0) J3 03 (J (U O > •>- •r- -t-) 1- IT3 U » O $- I— > "O in 0) 4- 1 a> 'd- > O O 1 .— x: > 0) r~- X — ■ Ol to OJ s_ ■1- 0) Q. (O n— . — ID i. E QJ 0) 3 OJ OJ c x: ■!-> c c 3 c/) +-> (D O (J 0) c o J^ ro tvj (u i. QJ Ol S- S- X) cn-t-> en u +-> E c o c o m 1 --^ E E ro O. fO w- i Ol >, i- OJ 3 O) Q. 1 — ro a> (/I 1— 1 QJ O ro r- s_ > • 0) •.- .- E -M Qi Q. 0) > O Ol O o l/l C O) Q.-0 (U — ^ . — ^ O XI i — O Q. 3 Ol S =c CQ ^ o u , E ^ — x OJ -C 3 1 cn x: a. +J D1 o f— I +-> 3 ■r- .i^ r^ O 1 •.- ■(-> 3 <-) 1 OJ 1 +j 1— . s 03 >, aj 1^ >> Q. 1—1 OJ en.— 03 E u 3 ra E E +J~— . a; O JD XI 3 c +-> X3 3 03 C x: x: ^— ^ JO c to C 3 Ol "O en o r-- 0--t-> en £z Q.-t-> en c/1 u c 3 0) 1 -O E OJ 03 -1- E CU 03 .:^ 03 03 O 0) ►— 1 •1- 03 CU -c E 03 O) x: U ■'-J^— r— en t— 1 u 3 S Q. o 3 3 Q. 03 "U Q. 00 o *■ — — 1 s- -C CLXJ +-> 3 Ol •I- ■(-> r- S O Ol o .^ a> 4-> ui ■ •-- 3 3 csj cn u 3 O) O — < .— en e ■— < 00 o —- — ' ,-!-> >, CL 03 Q. to O 3 O - to C L C to <1J 03 c= ro +-> S_ (_) I — O (DO. > O 3 '— M- U I — i- to CNJ •O — ■ OJ (O 3 03 to >— I 3 XI +-> 3 OJ ^ O) T3 C 0) O 3 (J -f- O +J >— c x: o <-> o - 03 -t-" C fO 'f— 03 03 O 3 -C to (J >—■»-> >-l •1-1—3 x: 03 o to ^ s_ 10 • r- cc S_ QJ ^ — ^ -C cc +J o 0) o C_3 c o 01 o > •■- •1- +-> S- n3 » o S- 1— Ol > -o •1- c en n .^ E OJ (D +J f— s = 1/) c ■o 4-) ID JC Ifl O 4-> QJ s_ •r- .C CQ 3 i. Ol OJ •r- ^ OJ 1 — 0) QJ "O 0) '~ OJ 4-> J3 .c +J .c: cn O 4J • r— T3 -r- ^ ^^ — s -^ ■(-> rtJ •!— +-> r— +J O >i3 S3 C >,! — O ■(-> +-> S> (/>» ■— U 1- V) l« •!- S_ >4- S_OlS_OEQ.O Q. 30010.0) .^ SZ £ OQ.>E-l->T3-OOOrD.C :ei/) Ol OJ i- t-H ■O I C 5> ID (/) O (D O) S- O) ID > J3 OJ >, ■— S- s- •*-> OJ ID .C ■— U Ol c '^? o c JD >D O U T3 ■*-> r— •r— O 3 1 S- LO .^ OJ I (O E S» O ct ^^ •t-> I •r- 01 .C S .— W -)-> Q. 03 Ol " *D ' — ■ S u) E c r~- +-> OJ I 3 (O -o OJ >■ O •— -• _l 4- S- CD OJ 3 o u ■r- O 3 -- OJ 3 CL O 3 J3 +-> X ■ — • O S- i-H Ol I TD +-> ►-< n— (D I—* O 3~-^ (D •1 — > 3 •\ r— CO , ^ — % >> o 3 Ol > E nz CD o T3 ■a o Ol +J •r— t— •1 — ai •f— > tn •f— Qi i- 0) 3 • »> o C >, 0) I/) Jli c -a E O O) -C c (U rT3 (U O CD CL O Ol o n 4J l_ => "O (O hJ ■D OJ lO M- -.- S_ •.- <+- ^_ OJ (J > OJ s_ en c S_ fO OJ Ol s_ ■»-> O O C Q. O) -o >— c l •.- O 3 C T3 «3 ai 3 +J ■*->■!-> n3 c JZ +-> O LU a> -Q 1/1 E rtJ -a >, O — '■— OJ ID i_ s- c l-sj >. (O O) 1 — -r— +-> u in QJ I/) O 2 (O 0) O Qj Qj en •*-> l/l .c cc z oo E o o •■- c o ^ — . -^ — . ^— V z r-j i_ -r- •.- ■!-> S- -o ^ o 0) nj +-> > o •<-> .— •!- Ol Ol ■— S- E C 0) ■— c .c +-> lo (T3 4-> 03 O. O 3 3--- t» -^ O tn »— ' ^ (U I O cn.:^ "O >— ' (J T3 "O C >• OO Ol o > lO O) 0) CL o > I O O -r- T- T3 E (04-3CS-t-C3 •I— rc i- ^ OJ <0 •»— -TSajQ-OE ' C 1/1 I — I/) «a; >i ' •!- +J .C T3 I Ol ' — tl) U fO+JCJi^C"Oi — c u r— -I— 0'0'»~CO(0'0 Li_ 3 E OQ-iO-c: (_)>■ I >> ca E-— CO .C I 3^ T3 c 3 o -O .C OJ 1 03 ,— +J (U O 3 (O •»— CTli — I ■ ■ XI 3 O O) -C CA Q. 1/1 I •» V) "O 3 lO >— ' -C (/) O) -t-" I— I Ol (U 1— S- — ' 3 S- O Ol (U O Q. O E I— >,+-> -C 1T3 ■— t/) U W) O 3 (U 1 CL O c Q. -r— E 4-> J^ •>- Q. (/I lO rrs Q. 3 C 3 ^ O >+- l— T3 o s- o j«: t— '^ C ■!-> E S_ r— ^ ID O 1 o S- 3 JD 1— 5» 0) 1/1 4- C >,Xi — +J QJ 4J i- o •o .,— cn C CO S_ .— O) T3 (/) ■o O OJ OJ c '^- o •r- •1- ^ ^ o •>- e: Q. S- ■t-> (J (J +-> Q. T3 O) =3 C O u .c 1 1 --^ QJ Q. ■4-> 1 •!- T3 <-! 1 r^ QJ O •- E >— 1 ^ o ID OJ 3 1 — to O ID •— < -(-) 1— 1/1 3 "O - — OJ CO "O -Q •— < •r- OJ 3 S- >, >> 3 S- »^ ID O ►-1 (D +-> Q..— J3 CI . +-> Q. ID 3 '-' >, S- CO ,-— C CO O M— i- U QJ "O +J 1 — 1- ■!-> Q. ro +J ■— — • i- +-> Q. a 01 ro (O O (O >^ 3 ID "O >— • QJ •»— 3 3 3 3 J2 4- C O 1/13-0"— h- 3 ■*-> E ID +-> I- ID QJ QJ 3 QJ +-> QJ .C ID +-> (J 3 C ID 1- ID .— O cC <_> JD Li_ O s_ QJ ^ » x: QJ +J ■♦-> > 3 s- •.- CD o s_ z cn r— r— OJ ID •■- .:.£ •.- > o > QJ c 3a ID r— o .— ■!-> a: ID ID 72 B Ol Ol Ol •a Ol o o rvi u s_ c aj o > •■- •I- +-> u - o s- ■— > -o c 0 i- O) r— > Oi cz s- 10 01 JC 1= f^ CD c in r^ 13 (O fO ro 0 n— c i: f— Q. =3: oo t/l Z3 c Ol > 03 m E Ol Ol ■(-> s. o +-> o i- .^ XI >o o X •>- LlJ s- in ■" u >i in TO O) -o c -o c CL 3 •■- c n3 ui +-> Ol -:>: u 1. u ■1- > o OJ ro Q. to s- (— S- JZ 3 OJ • 10 4- S_ > O) 0 03 t-H .. j3 -I- I— Ol S_ -r- >, OJ +J -— ~T3 +J O) C O Q >, C .C O ^ — x: rj ■!-> ivi ro Ol - .,— to Q. 03 S s I/) • r* 03 0 0 1 — 1 -0 0 -0 Q. s_ = 01 Q. dJ ^ ■^ E r— in +-> 03 +-> -0 0) 3 5 s_ c en 0 10 >) 03 03 I/) 1 E <— J3 1 1. V) ja 0 fO X •1- 03 ■!-> -0 03 = 0 +J 0) s x: E E T3 f^ -r- 10 0) 03 U. S Q. s- CL ■o OJ 3 o u 03 >■ 1 1 S >> E ja'-- 1 Q.I — 3 LT) CO 0 O) m 1 .•^ CO C '—1 10 S_ U O) to — - 03 Q. ■(-> 03 >> 0 03 >> H- (J <— S S- Q. T3 o; (DO +-> r— Q.-0 +-> C (U 03 3 C 03 03 3 JD •♦-> fO S 0 e I Q. CL O CL E o •■- 3 03 c m s -c 03 CO o 10 +-> o O) +-> •■- J3 S- >1 3 3 CL4-> fC CO 03 "a CL o I 03 ■'-3 c: -I- >— I 0) "O 03 C S- 1— 1 Q. 03 ■— O , 1 ■0 03 03 -^ i- 1 -r- 3 JD «* 3 •I- +-> CO 1 CM in X JZ 1 -fJ -o 1 0 -C +J JZ (U 1 — 1— ■ c s- a>-i- CO O) (O •— I Q.-r- 3 03 3 -Q ^ 03 Q.-C 1/1 ~,'-^ •.- 0 Ol to ■0 J^ 1— >4- > E Q- 03 Q. 03 03 OJ 3 3 3 >» U- 4- E-— o.+j -0 U c -O -r- _l > a> O C I— +J 3 CT> Q.-I— 03 to 3; "O •.- S_ .^ 03 •■- .C Q.I— C -t-" S- t_3 to 03 <0 03 43 m Ol Ll- 01 C S- l. •■- O) •• OJ O >—+-'+-> O) OJ t- <0 03 OJ O; C -ii 3 C c ■— <— -i<: s- c 09 ro 03 U OJ 03 3 T3 OJ > 3 3 ••- C: r— •.- 3 LT) ■*-> O XI S- C/1 73 ■a 01 -o 3 (J LO C 0) o c <_) o — hv] ^ m E s_ to 0) i. c O) o > -r- ■>- +-> u - o S- I— > -a •r- C OJ s o r^ 1 JD ■t-> 3 c OJ 0) •— ■o i. .. QJ OJ > ■a ro ■O -r- a> s- ' — ^ JD rtJ IB ■t-> C ■o ^-' (O 01 -r- • r— cn CJ M S- (O s_ Ol oo 13 E — .^ r— S- fO J^ " O "O ■ — ■+-> 1- S- O) Ol -a ■t-> +-> ifl c 0) Q. 01 O s- o i- "O a; >— I/) o X -o OJ 1- Ol c o ■I- tM (/) 3 Li- H3 +-> O +J +j O) c Ol E O > I— 3 ro O 03 OJ OJ O "O O) -l-J -Q "O 1/1 OJ Cn +-> -r- (_) fO O) ■!-> lO J3 E O) .i<: Cl"0 J3 I— Ol O O i- •—< 2 E -Q 3 ns •— ' lO 3 JD O "O D1 ID LO Ol OJ s- -^ U 1 c lO >,— "O s- c r— (O ro QJ >, t— < JO JD Q. S- E S- -O Ol " +-) 0) > (/I to 14- -1- -^ •s OJ 1 i- L. lO i- Ol 03 ^ o C QJ s: s- 4- •r- c 03 S- ■>- • e: . — Q.<— +-> 03 (/I (O OO • ■o .i<: ;/) -t-J O 03 ro Li_ 74 Figure 39. An example of understory species taking advantage of the dry environment of a floating log in order to exist in Zone II. Photo by Gordon Fritz. 75 River, SC floodplain, for instance, 25 were canopy trees, while 28 species were shrubs or subcanopy trees, 15 woody vines, 2 woody grasses, and 40 herbaceous species (Dennis 1973). The importance of the herbaceous ground cover on a floodplain is a function of light and area! extent of zones of higher elevation (Zones IV and V) where these species are most often found (Knight 1973). In Zone IV floodplains, "sedge glades" are sometin-'es found. In the Congaree Swamp (SC), dominant sedges are Carex lurida and C^. grayi, while C. intumescens. C. atlantica, and Leers i a virginica are commorT] Other common ground cover plants are crossvine (Aniso stichus capreolata), southern vein orchid (Haban- eria flava), violet (Viola affinis)r poison oak (Rhus toxicodendron) and swamp milkweed (Asdepias incarnata). The major composite is the butterwort (Senecio glabellus). Common shrubs and vines on North Carolina floodplains are spicebush (Lindera benzoin), buckeye (Aesculus sylvatica). Viburnum spp. , Japanese honey- suckle (Lonicera japonica), greenbrier (Smilax rotundi folia), poison ivy (Rhus radicans), grapes (Vitis spp.), and black- berry (Rubus spp.) (Knight 1973). Herba- ceous ground cover is less extensive in more frequently inundated parts of the floodplain. A characteristic and wide- spread herb in Zone II of many floodplains (in the Coastal Plains) is goldenclub (Orontium aquaticum). Useful species lists for floodplain understory and ground cover plants are found in Oosting (1942), Houck (1956), Beard (1958), and Wells (1970). Floodplain Transects of Selected South- eastern Rivers To provide a sharper focus on the complexity of floodplain ecosystems. Figure 40 portrays nine southeastern allu- vial, blackwater, spring-fed, and tidally influenced floodplains via horizontal transects. The spatial relationships between ecological zones and forest domi- nance types are diagrammed and accompanied by Table 15, which depicts these relation- ships. The rivers profiled here are the same referenced in Tables 11 through 14, where specific site locations were given for dominance types and variants observed in the field. The diversity of the bottomland hard- wood canopy increases with the complexity of floodplain topography. Figure 40 and Table 15 describe examples of the varia- tions encountered on floodplain transects. Alluvial rivers, such as the Congaree, Ochlockonee, and Alcovy (Figure 40) often exhibit the greatest topographic and plant community diversity. In contrast, the floodplains of many Coastal Plain black- water and spring-fed rivers (Figure 40) generally are more uniform topographi- cally, and the plant community diversity is lower. Regular tidal flooding caused by lunar or wind energy imparts a distinct character to plant communities under its influence. These forests are the least studied of all river swamps (Beaven and Oosting 1939). They occupy the lower 16 to 32 km (10 to 20 mi) of many unaltered floodplains, primarily in Florida and Georgia (Figure 40). High tides raise the water levels in the floodplain (or "tide- plain") to the most elevated portions of the relief with fair regularity. Tidal swamp forests usually extend upstream until levees appear. The floodplains of these forests are dominated by Zone II species, except for higher islands con- taining Zone IV species such as diamond- leaf oak, swamp palm, southern red cedar, and cabbage palm. These floodplains are distinctive in harboring animal species characteristic of more brackish downstream waters, such as fiddler crabs (Oca) and square-backed crabs (Sesarma). Another interesting aspect of the tidal forest is the presence of an intertidal zone, either mud or sand, which is often extensive mats or beds of aquatic vegetation (such Ludwigia repens, Cabomba El odea spp. , or Nuphar luteum] vegetated by submerged or as Isoetes, cardiniana. DISTURBANCE AND SUCCESSION IN BOTTOMLAND HARDWOOD PLANT COMMUNITIES Several examples of distrubance and successional trends have been referenced in the preceding sections of this chapter. They will be summarized here along with others that have not yet been mentioned. 76 CONGAREE RIVER, SC «.,";SruVE'.°Tr.?-H G F B E E BD C B A ROANOKE RIVER, NC ^•^0MmmJ^M}K ~ E D E D E F G V V II IV IV V CANOOCHEE RIVER. GA A B E C E D E ^ muijMMi IV- V Figure 40. Cross-sectional transects (aspect is looking downstream) of nine southeast- ern rivers and floodplains, indicating zones (I-V) and major vegetational and natural features (A-J). This figure is cross-referenced to Table 15, which provides an expla- nation for each vegetational or natural feature category (A-J) of each floodplain. See Table 15 also for cross references to dominance types (Tables 11-14) found on these transects. Flooding The most common natural disturbances in bottomland ecosystems are associated with floods. The biota are to a variable degree adapted to flood forces. Annual inundations adapt the bottomland hardwoods for the larger and more catastrophic flood events that occur with low frequencies (100- to 1000-year floods). The wide, shallow root crowns and trunk buttresses, which are adaptations to the moist, anae- robic conditions, serve to counter exces- sive scour and toppling by flood or wind. Deleterious effects may occur, however, depending on flood timing, frequency, depth, and velocity. The categories of flood disturbances include (1) anaerobic conditions, (2) mechanical abrasion and breakage of plant tissues, (3) siltation, (4) propagule and seedling washout, and (5) erosion. Flooding may retard or speed succes- sional trends. Severe erosive flooding inhibits point bar succession (see Chapter 1), slows natural levee development and community establishment on levees, and can forestall the filling in of scour chan- nels, swales, and depressions between hum- mocks. Moderate flooding enhances mature community development through less damag- ing effects on plant survival and growth and through reduced erosion and net depo- sition of floodwater sediments. Fire Fire is not important as a natural disturbance in bottomlands because of the prevalence of water and the lack of a sub- stantial litter layer. This is especially true of the wetter portions of alluvial floodplains. However, Putnam (1951) states that a serious fire season occurs on an average of about every 5 to 8 years in bottomland hardwood forests (in the Mis- sissippi Valley). It is conjectured that the Indians maintained canebrakes in Zone V understory by deliberate fall burning. 77 perpetuating and increasing large cane stands, which are now relict. Though crown fires are rare, ground and surface fires may occur. These fires move rapidly across the floodplain floor, damaging or destroying all trees less than 10 years old, as well as shrubs and herbaceous growth. In addition, ground fires open wounds in larger trees, increasing sus- ceptibility to disease and insect attack. Fire is a major determinant of com- munity composition in selected vegetation types. Infrequent but regular fires favor Atlantic white cedar while inhibiting southern red cedar (Eyre 1980). H.S. Larsen, writing in Eyre (1980) indicates that fire may sweep into the dense shrub zone of sweet bay-swamp tupelo-red bay sites along the narrow bot- tom of perennial streams during drought years. It is possible that the narrow peaty swamps along small streams with At- lantic white cedar or pond pine canopies (type 29) may burn in drought years. The only recent instance of a widespread fire on a floodplain in the study area occurred prior to 1976 on the Oklawaha River (FL). Caddy et al. (1975) described the successional sequence in the Congaree Swamp (SC) that begins with either fire or clearcutting, and proceeds from even-aged sweetgum-intolerant hardwood stands to more mature communities dominated by dia- mondleaf oak and more tolerant hardwoods. Windthrow Windthrow is the primary disturbance to plant succession on floodplains in the study area. Gaps in the canopy resulting from fallen canopy trees (Figure 41) are common in the floodplain. Topplings due to old age, disease, soft sediments and insecurely anchored root systems, root scour, lightning strike, or fire cause openings in the canopy that temporarily stimulate understory woody and herbaceous growth. The factors that influence suc- cessional response to such gaps include gap size, existing composition of seed- lings and saplings and their relative shade tolerances, possible inhibition due to shading by extensive understory tree (paw paw, holly (Ilex opaca), ironwood (Carpinus caroliniana)) or canebrake development, site characteristics, and probability of propagule recruitment. Understory shading may limit the develop- ment of diverse, well -stocked seedling and sapling layers, retarding succession (Gaddy et al. 1975). Biotic Disturbances At least three categories of biotic disturbances exist in the floodplain: (1) propagule predation and seedling and sap- ling herbivory by browsing animals, (2) disease, and (3) insect outbreaks. The quantitative effects of these variables on plant community structure and composition have received little attention. Cattle and deer browsing can kill seedlings, particularly if floodwaters concentrate browsing on higher ground in the flood- plain. Baldcypress seedlings are even eaten, and water tupelo will survive only one cropping by deer (F. Vande Linde, forester, Brunswick Pulp Land Company, Brunswick, GA; personal communication). Conner and Day (1976) discussed the effects of grazing by the forest tent caterpillar on both baldcypress-water tupelo and bottomland hardwood forests in Louisiana. Water tupelo is most severely affected, suffering extensive defoliation. These authors suggested that the increas- ing frequency of outbreaks over wide areas is due to a corresponding increase in the areal extent of tupelo-dominated sites. These sites, in turn, occur as a result of the selective logging of baldcypress (see below). Conner and Day (1976) also specu- late that the susceptibility of water tupelo to defoliation may be one factor that formerly favored the maintenance of nearly pure stands of baldcypress. Lumbering Selective cutting and clearcutting generate some of the most noticeable changes in floodplain forests. The heavy exploitation of baldcypress is the classic example. Such logging has shifted the forest composition on countless sites. Cypress stumps endure for many years, and their presence may indicate what the orig- inal forest on a given site was like and something of the hydrology. Many areas which now support water tupelo (e.g., the Altamaha River, GA), green ash (e.g., the Great Pee Dee River, SC), or water tupelo- green ash (e.g., the Oklawaha River, FL) 78 Figure 41. This windthrown bitternut hickory on Special Management Area (Oconee National Forest, created by this natural event. In mature forests provide vegetation in all states of succession, system of uneven-aged management. i the floodplain of the Murder Creek GA) illustrates the large opening such openings are common enough to The virgin forest thus has its own formerly bore forests of very large cypress. Klawitter (1962) discussed the three periods of baldcypress logging on the Santee River (SC). Clearcutting of hardwoods other than baldcypress may also lead to entirely new forest overstories. Sweetgum and shade- intolerant hardwoods pioneer after clear- cutting in the Congaree Swamp (Gaddy etal. 1975), and mid-seral sugarberry- American elm-green ash stands follow extensive logging of Zone IV (Nuttall oak- willow oak) forests in the Mississippi Valley. Agriculture The growing of rice altered many floodplains was introduced around 1700 in swamps on smaller streams that into large navigable rivers (e.g. Creek and Santee River, SC). dams were constructed across small feeder has completely Rice culture Zone II emptied , Wambau Reserve creeks, providing a reservoir to supply water to the rice fields, even on coastal islands (e.g., Hobcaw Barony, SC). This system persisted until 1885 (Klawitter 1962). When the tidal flooding method was developed in South Carolina in 1750, large-scale rice plantations became feasi- ble, and entire floodplain forests of cypress were burned or buried by slave labor (e.g., Santee River, SC). The fields with their remnant levees from these plantations are used today as waterfowl refuges (Figure 42). Unsuccessful attempts at cotton and other agriculture have taken place on many southern floodplains, especially in the great fall line swamps of the Flint and Oconee Rivers in Georgia (Wharton 1977). These abandoned areas support a variety of forest cover. One area on the Roanoke River (NC) bears an almost pure stand of large cottonwoods. Boxelder flats can be found in such disturbed areas along the Chattahoochee and Alcovy Rivers (GA). 79 Figure 42. Aerial view of relict rice fields on former bottomland hardwood forests that are presently managed for waterfowl. Photo courtesy of the Georgia Department of Natural Resources. PRIMARY PRODUCTIVITY OF FLOODPLAIN FORESTS High productivities of the floodplain forest (Conner and Day 1976) are made possible by several subsidies offered to the floodplain by the watershed and river, including particulate and dissolved or- ganic matter, water, soil (especially clay and silt), and nutrients (inorganic, sediment-adsorbed, and organically com- plexed). These inputs support what is essentially an increased rate of ecosystem community metabolism, reflected in (1) annual litterfall and nutrient turnover rates as high or higher than most temper- ate deciduous forests; (2) relatively high detrital decomposition rates, except in systems with permanently ponded water; (3) periodic "flushing" of accumulated refrac- tory organic detritus and metabolic by- products; and (4) the operation of several microbial conversion processes character- istic of widely varying conditions, such as nitrification, denitrif ication, ammoni- fication, methanogenesis, sulfate reduc- tion, and general nutrient mineralization (Wharton and Brinson 1979a). In addition to these physical and chemical subsidies, the river contributes macro- and microfauna during flood periods that both speed detrital decomposition and 80 participate in the floodplain's food chains, nutrient cycles, and import-export pathways. The major factor contributing to the high productivity of the floodplain forest is the pulsing of the wet-dry cycle. Conner and Day (1976) made an analogy between these floodplain forests and the tidal marshes in terms of the positive effects of fluctuating water levels: "This periodic flooding acts somewhat in the same manner as tidal flooding in saline marshes, in that fluctuat- ing water levels are energy subsidies which control variations in hydric conditions, temperature, nutrient levels, and available oxygen (Hester 1973; Butler 1975)." Bottomland hardwood communities that either are permanently flooded with slow- moving to stagnant water, or are regularly damaged by unusually high and irregular destructive floods are not as productive as communities that undergo periodic mod- erate floods. This has been illustrated clearly (Figure 43) by Odum (1978), who graphically compared the productivity of stagnant, seasonally flooded, and abra- sively flooded systems with a regional average of all wetland and upland forest types . Communities in permanently ponded conditions, or on sites where poor drain- age leads to continuously high water tables and the accumulation of acidic peat soils, have lower productivities primarily because of low nutrient turnover, due to anoxia, nitrogen limitation, and low pH. Brown et al. (1979) and Conner and Day (1976) presented data that demonstrate the reduced productivities of still water systems. Productivity values gleaned from the literature for 19 upland and bottomland forest types are presented in Table 16 and generally support the concept of a flood subsidy depicted in Figure 43. A second verification of this concept is shown in Figure 44, where productivity data from sites in several zones are plotted (from Gosselink et al. 1981). Gosselink et al. (1981) stated: X^^^ o O QL / _Re_£ion_aMe_vel \ _ - / , 1 ' \ Stogn ant Seasonal flooding Slowly flowing Stress -• Subsidy- Abrasive flooding — Stress a gradient of compared with Figure 43. The effect of flooding on productivity as a regional level that might be expected in the absence of standing or flooding water. The graphic model takes the form of a stress-subsidy curve. For southern swamps Conner and Day (1976) estimated annual net production for stagnant, slowly flowing and seasonal flooding conditions as of the order of 0.2, 0.7, and 1.2 kg dry matter per square meter, respectively (Odum 1978). "Forest production appears to peak at the once-per-year flood frequency if flooding is during the winter because this regime furnishes the optimum environment for plant growth in terms of nutrient input by flood waters, summer soil moisture, and possibly aerobic conditions during the summer leading to inorganic nutrient release from organic debris." Primary productivity data in the lit- erature are much more common for the tree canopy and woody subcanopy (small trees and shrubs) components of floodplain com- munities than for the herbaceous, aquatic vascular, and nonvascular components. Aquatic plant productivity in river chan- nels and drainage tributaries, permanent ponds, and temporary sloughs and swales has received the least attention. Brinson and Wharton (1979a) suggested that the productivity of alluvial stream communi- 81 Table 16. Net primary productivity (g dry wt/m /yr) for bottomland hardwood communities (primarily Zone II), com- pared with other wetland and upland environments. Community type New primary productivity (g dry wt/m^/yr) Dwarfed cypress strand (FL) 367 Okefenokee cypress forest (GA) 595 Oak-hickory upland (MO) 600 Cypress-water tupelo (IL) 678 Drained cypress strand (FL) 681 Cypress-tupelo (Green Swamp, FL) 760 Oak-pine uplands (NY) 796 Slash pine flatwoods (FL) 830 Northern hardwood upland (NH) 898 Cypress-hardwood (Green Swamp, FL) 950 Mature cypress dome (FL) 956 Elm-ash-sweetgum (Zone IV) (IL) 967 Spruce-fir upland (Great Smokies) 980 Upland cove forest (TN) 1050 Cypress strand (FL) 1111 Riverine cypress-water tupelo (LA) 1140 Mixed bottomland hardwoods^ (LA) 1174 Cypress-water ash creek forest^ (FL) 1607 Tulip poplar upland forest (TN) 2400 References Carter et al. 1973 Schlesinger 1978 Rochow 1974 Mitsch et al. 1977 Burns 1978 Mitsch and Ewel 1979 Whittaker and Marks 1975 Golkin 1981 Whittaker and Marks 1975 Mitsch and Ewel 1979 Brown 1981 ^ S. Brown (pers. comm. ) Whittaker and Marks 1975 Whittaker and Marks 1975 Burns 1978 Conner and Day 1976 Conner and Day 1976 Brown 1981 Whittaker and Marks 1975 ^Sandra Brown, Department of Forestry, University of Illinois, Urbana. Red maple-water tupelo-box elder-cottonwood-cypress-swamp dogwood-willow (mix of Zones II and IV with pioneer species). ''Also contains diamondleaf, oak, sweetgum, red maple (mix of Zones II and IV), Flow partially regulated by low dam. ties is probably low because of heavy silt loads. Productivity of the Satilla River and Okefenokee Swamp does not seem to be limited by low nutrient availability and acidic conditions. Brinson observed ex- tensive production of filamentous algae in floodplain ponds during the winter dormant season and suggested that this component may provide a temporary sink for inorganic nutrients during winter and early spring. Aquatic vascular productivity can be high in localized areas, depending on light intensity and water velocities. Species that may contribute heavily to community productivity are Alternanthera phi loxero ides, Myriophyllum spp., Lemna spp., Spirodela spp., Egeria densa, Ceratophyllum spp., Lim^nobium spp., and Azolla spp. (Dennis 1973). The prominence of the herbaceous ground cover varies dramatically among the forest cover types, as discussed in the previous section on dominance types. In general, the highest herb densities and productivities are found on the driest floodplain sites (heavy growths of various composites follow drydown in Zone IV). This is a function of hydroperiod and light intensities. One species that can produce tremendous amounts of biomass in 82 ZONE II III IV V 2000- 2 1600- o D Q O IT > < a: a. < 2 < 1200 800- 400 Figure 44. et al. 1981) FLOODING FREQUENCY - YRS FLOOD Organic matter production in ecological zones (adapted from Gosselink Numbers represent specific floodplain sites. short periods is river cane. Wharton (1977) reported that this plant may pro- duce 4506 kg of edible leaves per hectare (4000 lb per acre) per year, and 11 to 16 tonnes of organic material per hectare (5 to 7 tons per acre) within the first 3 years of growth. A striking feature of many floodplain plant communities is the prominence of woody vines. Dennis (1973) recorded 15 species in study plots in the Santee Swamp (SC), including Smilax spp., Vitis spp., cross vine (Anisostichus capreolata), supplejack (Berchemia scandens), poison ivy (Rhus radicans), climbing hydranger (Decumeria barbara), Virginia creeper (Parthenocissus quinquefolia), and trumpet vine (Campsis radicans). TFe contribution of this component to community productivi- ties may be large locally, particularly in canopy gaps created by windthrow and along river banks. 83 CHAPTER 5. FAUNA OF BOTTOMLAND HARDWOOD ZONES Bottomland hardwood forests support a diverse fauna that matches the floristic and hydrologic complexity which is so characteristic of these communities. The moisture gradient and hydroperiods of floodplains provide a habitat continuum for a wide range of aquatic to terrestrial to aerial species. The fauna is here also treated within the zonal concept. Because of the large numbers of taxa, only abun- dant or dominant animals or groups can be mentioned. (For further information, see Wharton et al. 1981.) Some overlap among zones occurs, especially between Zones IV and V, which share many species. The mobility of many species and their over- lapping distribution in response to vary- ing environmental regimes make combining discussions of faunal assemblages in Zones II and III useful. It should be recog- nized that placing an animal in one or even two zones does not necessarily restrict it to these areas. Floodplain inhabitants are opportunists, and many move freely into irregularly flooded or dry areas over the year. FAUNA OF ZONES II AND III Invertebrates Given the diversity of vegetational dominance types in Zone II, it is not sur- prising to find that faunal components also vary. In terms of fauna, the environ- ment of a tupelo gum-cypress forest with hydroperiods approaching a year is mark- edly different fron, a similar forest in a tidal area with daily water level fluctua- tion or a forested site with permanently saturated soils. An example of this phe- nomenon is illustrated in Figure 45 for a coastal section of the blackwater Suwannee River (FL). The tree associations within Zone II of the Suwannee change with dis- tance from the coast in response to less- ening tidal influences (i.e., to the extent of daily inundation and salini- ties). The vegetative changes are changes in species morphology as well as species replacements. The coastal forest comprises dwarfed swamp tupelo, pumpkin ash, sweet bay, cabbage palm, and cypress which transform upstrean^ to an association of taller Ogeechee tupelo, water tupelo, pumpkin ash, and cypress. The faunal associations change abruptly from a brack- ish water snail-fiddler crab community (Neretina-Uca) to a freshwater snail - crayfish community (Vivipara-Cambarus ) at the point upstream where natural levees first occur. Macroinvertebrates dominate Zone II and the wetter depressions and pools of Zone III. Parsons and Wharton (1978) doc- umented a cyclic sequence of dominant macroinvertebrates in isolated pools (Fig- ure 46) (Zone III) in Piedmont flood- plains: initially stoneflies dominated, followed sequentially by the isopod Asellus sp. and amphipod Hyalel la azteca, snail oligochaete worms and midge fly lar- vae, and finally an association of sphae- riid clams (Sphaerium and Musculium). Sklar and Conner (1979) found an almost equal distribution of aniphipods, oligo- chaetes, gastropods, and turbellarians (densities of 10,700/m2) on vegetation in a tupelo gum-cypress association. Beck (1977) found that detrital substrates, such as Zone II soils, were extremely pro- ductive, averaging 2885 organisms/m' in a large alluvial system (Atchafalaya Basin, LA). The dominant macroinvertebrates were a tubificid annelid (Peloscolex multi- setosus), an isopod (Lirceus lineatus), an amphipod (Gammarus tiqrinusX a mayfly (Caenis), a phantom midge larva (Chaoborus punctipennis), a chironomid (Chironomus), a pulironate snail (Physa), and a finger- nail clam (Pisidium). Ziser (1978) found an average density of 1296 organisms per 100 g of duckweed and water hyacinths in a Louisiana swamp. The dominants were the naidid worm (Dero), three snails (Physa, Ferrissia and Promenetus), an oribatid mite (Hydrozetes), two" damsel flies (Enallagma, Ischnura), a back swimmer (NeopTea striola), a'nd midge and biting midge larvae. 84 Figure 45. Comparison of the ferent reaches of a blackwater (L) low tide. Upper figure: tidal forest of dwarfed swamp fiddler crab (Uca minax) and bottomland hardwoods and characteristic fauna at dif- river near the coast (Suwannee, FL). (H) is high tide; River Mile 4, intertidal zone, largely exposed roots; tupelo, pumpkin ash, sweet bay and cypress (B) olive nerite snail (Neretina with: (A) , , ^,..., . reclivata) . Lower figure: River RTTe 15, intertidal zone wide, with: (G) spatterdock (Nuphar luteum), fanwort (Cabomba Carolina) and Elodea sp. ; forest is tall Ogeechee tupelo-water tupelo- pumpkin ash-cypress with: (C) crayfish (Procambarus seminolae), (D) snail (Vivipara qeorgianus), (E) dwarf salamander (Eurycea quadridigitata). (F) southern leopard frog (Rana utTTcularia). 85 86 Arthropods, crustaceans, and mollusks dominate the niacroinvertebrates of the Congaree, Swamp (SC). Three dragonflies (Epiaeschna heros, Tetragoneuria cynasura, Gomphus exjjisl are abundant and assumed associated with Zone the red Procambarus II. Crayfish such as cl ar kii (west of the Mobile system) and P. troglodytes (east of Altamaha system) are (along with crayfish from Zones IV) an important food for a host of vertebrates such as the eel, cat- fish, warmouth, amphiuma, glossy water snake, ibis, otter, and raccoon. Densi- ties ranging from 21 to 46/m2 have been reported (Konikoff 1977; V. Lambou, Envi- ronmental Protection Agency, Las Vegas, NV; personal communication). Crayfish, in fact, form one-third of the faunal biomass of the Suwannee River floodplain (Wharton 1S77). Large fishing spiders (Dolomedes spp.) and Pi rata maculatus, which are found under the liverwort Porel la platy- phyl loidea, are characteristic of Zone II. Several snails (Vi vipara, Campeloma, Pomacea, and Lioplaxl live in and around Zones II and III (Figure 47). Fingernail clams of the genera Sphaerium, Eupera, Musculium, and Pisidium often dominate the benthic biomass of Zones II and III. These tiny (<10mm) clams are present in enormous numbers. Some clams (Anodonta, Ligumia, Corbicula) occur in Zone II sloughs, but the clam fauna is in general poorly known. The Altamaha River (GA) is unique in pos- sessing six endemic clam species, all in the family Unionidae (Figure 48). These clams require particular species of fish as hosts for their larvae; a diverse fish fauna may be essential to clam diversity and survival. Vertebrates The most characteristic fish fauna of inundated Zone II sloughs are top minnows (Fundulus spp., Gambusia affinis), killi- fishes (Heterandria formosa, Lucania parva), swamp darter (Etheostoma fusi- forme), pirate perch (Aphredoderus saya- nus), lake chubsucker (Irimyzon suce"tta), yellow bullhead (Ictalurus natalis), flier (Centrarchus macropterus ), warmouth (Lepo- mis gulosus), and three top predators: the bowfin (Ami a calva), redfin pickerel (Esox americanus), and chain pickerel (Esox niger). Dominant amphibia are the lesser siren (Siren intermedia) and amphiuma (Amphiuma means), which seek refuge in root holes and crayfish holes during dry- down. The amphibious salamanders include the southern dusky (Desn.ognathus fuscus auriculatus), the many-lined (Sterochilus marqinatus), and the dwarf (Eurycia quad- ridigitata). The mud salamander "(Pseudo- triton montanus) and the two-lined sala- mander TF bislineata) occur around the edge of Zones II and III. Frogs are less specific to Zone II but include the cricket frog, river frog (Rana heckscheri ), and southern leopard frog, and at breeding times several other species such as the bird-voiced tree frog (Hyla aviyoca). Some depression pools (Zone III) may support annual breeding aggregations of spotted and marbled sala- manders, as well as temporary water-breed- ing frogs and toads from Zones IV and V. Only a few reptiles are locally abun- dant in Zone II areas. The dominants appear to be the eastern mud turtle (Kino- sternon subrubrum) , glossy water snake (Natrix rigida), perhaps the mud snake (Farancia abacura), and certainly the red- bellied water snake (Natrix erythrogaster ) and cottonmouth (Agkistrodon piscivorus). In a tupelo gum-cypress association of an anastomosing blackwater creek (Zone II, Four Hole Swamp, SC) the yellow-bellied turtle (Chrysemys scripta), brown water snake (Natrix taxispilota), "greenish" rat snake (Elaphe obsoleta)~and anole (Anolis spp.) were abundant (Hall 1976). Passerine (perching) birds character- istic of Zone II are limited largely to the prothonotary warbler, tufted tit- mouse, parula warbler, and common grackle. The wood duck nests near water if possible and often in Zone II. The yellow-crowned night heron and green heron are common breeding residents, and rookeries of great blue heron, great egret, and white ibis also occur in Zone II. The red-shouldered hawk is a charac- teristic raptor of Zone II. Swallow- tailed kites feed and nest in this zone on Wambau Creek (SC) and perhaps on the Altamaha River (GA). The snail-eating limpkin is found chiefly here (and along sloughs in Zones IV and V) above tidal range where Vi vipara georgiana and other snails abound. Many wintering birds such as robins make heavy use of tupelo fruits; 87 Figure 47. Snails of the genera Vivipara (shown) and Campelop'a are often abundant in shallow aquatic zones (Zone II) above tidal influence. 88 01 m xz — o ^ to m o 01 >> i- o +-> n3 i- o .a m 10 c o o " OJ '^ U -r-, E J= s- o o "o ■'- s- o c -c .,- ^- o •*-> > (J c: ._ •I- d (D ro ,-^ in 1T3 •»— T3 t •r— s_ •0 0 C M- O) o; r— 10 . "O -a e 03 10 .. +J O) c Cl o §;= u 0) a. 00 • ^ E to QJ S- XJ •r— c U- , q: 00 «:J- fO ^ ai fO s_ E 3 ro cn+J (O Q. ■^^ 0 C JZ 0 -0 •r— 0 E •f— 10 +J IC Q. •r- 5 ^~ #1 UJ| T3 Ol u. «3: m Ojct 89 the seeds are eaten by squirrels. Although the rice rat occasionally appears In Zone II, small mammals are usu- ally absent In most of it. Mink, raccoon, beaver and otter may use tupelo gur.,- cypress forests (Zones II and III) in par- ticular. FAUNA OF ZONE IV Invertebrates The invertebrate fauna of Zone IV can be subdivided according to their dominant use of this floodplain zone: (1) inundation fauna -- inverte- brates occupying the substrate and water column during periods of flooding (2) litter fauna — invertebrates occupying the leaf litter layer during dry periods (3) persistent fauna -- inverte- brates occupying the floodplain habitats in various life history stages throughout most of the year. Sniffen (1980) characterized the inundation fauna of the Creeping Swamp (NC) floodplain (Zone IV) as a large and diverse component of this small blackwater floodplain. The most conspicuous inverte- brates were six species of "red" lumbricu- lid worms and four species of "white" enchytraeid worms, three tubellarian flat- worm species, and several roundworm spe- cies. Oligochaete worms and copepods were numerically the most abundant inverte- brates (16,470/m2). Isopods, although fewer in number than the worms, were the dominant biomass component (1114 mg dry wt/m2). Ostracods were numerous (829/m^, as were nematodes (4348/m2), Kidge fly larvae, amphipods, water mites and ccllem- bola were also abundant. There are relatively few definitive studies of the litter fauna of floodplain communities in Zone IV. Grey (1973) con- ducted the most thorough faunal survey of this particular habitat. His study of the Santee River (SC) floodplain determined that mites (Acari) and springtails (Col- lembola) were by far the dominant litter organisms, accounting for about 92% to 95% of the organisms during any season. The mites comprised 48.1% and 77.4% of the total population count in the summer and fall, respectively; the springtails, 47.6% and 13.1%. Both groups are detrital "shredders" important to the decomposition processes in the upper litter and humus layers. Earthworms, important food sources for salamanders and shrews, are also an abundant and important component of the litter fauna. Parsons and Wharton (1978) found three genera of earthworms (Eisenia, Al lolobophora, and Sparganophi lus) in the floodplain of the Alchovy River (GA). Harper (1938) noted that earthworms (principally the genera Diplocardia and Helodrilus) preferred dense, packed flood- plain soils with water tables below 23 cm (9 inches). Other invertebrate fauna using Zone IV throughout the year and throughout their entire life cycles are principally crayfish and insects. Some 23 chimney- building floodplain crayfish species occur east of the Mississippi River and 19 east of the Escambia River (FL) (Wharton et al. 1981). Several, such as Procambarus pubischelae and £. seminolae, seem to favor blackwater floodplains; others on almost all floodplains are Cambarus dioqenes and Procambarus acutus. Many insect species whose larvae inhabited sloughs and pools within Zones II and III may be found in Zone IV as flying adults during drydown. Mobile species, such as dragonflies and butterflies, span many zones. Some, like the abundant snout butterfly (Libytheana bachmanii ) and the hackberry butterfly (Asterocampa celtis), can be categorized by their larval prefer- ence for sugarberry, a Zone IV tree. Vertebrates Amphibians, especially salamanders, are abundant in Zone IV. The marbled sal- amander (Amby stoma opacum) is generally restricted to this zone; others such as the mud salamander (Pseudotriton montanus) seldom occur elsewhere. The dominant plethodontid salamanders include the two- and three-lined salamanders (Eurycia). The four-toed salamander (Hemidactylum) occurs here and in Zone II. The green and 90 leopard frogs (Rana) and cricket frogs (Acris) are dominant anurans; the upland chorus frog (Pseudacris niqrita) and grey tree frog (Hyla versicolor) are common locally. The bird-voiced tree frog (Hyla avivoca) occurs here and in other zones, especially at breeding time. Reptiles in Zone IV are represented by the abundant box turtle (Terrepene Carolina) ; the giant gulf coast form (T. c. major) also occurs on floodplains. There are few snakes in Zone IV other than the rat snake (Elaphae obsoleta) and sub- species. Boyd (1976) encountered copper- heads and rattlesnakes in Zone IV study areas, but these snakes (more characteris- tic of Zone V) may have come from a nearby hillside. Tinkle (1959) reported the black racer (Coluber constrictor), kingsnake (Lampropeltis getulus), and ribbon snake (Thamnophis sauritus) on a narrow levee ridge, assumed to be Zone IV from the site description (or in succession to Zone V), although these snakes are not frequently encountered in Zone IV. Many bird species are found in Zone IV. In a study in the Congaree Swamp, numbers of species were similar among floodplain Zones II, IV and V; however, population densities were almost always highest in Zone IV (Hamel 1979i Hamel and Brunswig 1980). Characteristic birds in the Congaree Swamp are the barred owl, downy and red-bellied woodpeckers, and cardinal (Hamel 1979). The wild turkey is known to nest and feed in Zone IV (Kennedy 1977). In fact, bottomland hardwoods sup- port the highest population densities (1 per 10 acres vs 1 per 25 acres of upland) of eastern wild turkey (Florida Game and Fresh Water Fish Commission 1978). Dominant Zone IV floodplain mammals are the deer mouse in the Piedmont, the cotton mouse in the Coastal Plain, and the golden mouse in creek swamps and areas of dense shrub and vine growth. Short-tailed and southeastern shrews are abundant in this zone but may retreat to higher zones during innundation. Most of the larger mammals in Zone IV are also common to Zone V. The woodrat (Neotoma floridana), which nests in the ecotone adjacent to the up- lands, forages in Zones IV and V. It nests in Zone IV along spring-fed rivers. Two of the few vertebrates that are confined almost exclusively to Zones IV (and V) are the semiaquatic swamp and marsh rabbits (Syvilagus aquaticus and S. palustris). Swamp rabbits are found more often in Piedmont floodplains while marsh rabbits are confined mainly within the Coastal Plain. The swamp rabbit is adapted with large feet and slightly splayed, strong-nailed toes for swimming and tra- versing unconsolidated terrain (Lowe 1958). Herbivorous swamp rabbits reached a density of 5.6 individuals per 100 acres in the Lowe study on the Oconee River, GA. FAUNA OF ZONE V Invertebrates Many invertebrate species are common both to Zones IV and V as well as to levees (Wharton et al. 1981). The detri- tivore community of the predominantly Zone V Alcovy River (GA) floodplain is charac- terized by abundant millipedes (Cherokia qeorgiana, Narceus americana) and camel crickets (Ceuthophilus qracilipes). Also abundant are a scarab (Onthophaqus) and three carabid beetle genera (Carabus, Abacidus, and Chlaenius). The grazer com- munity includes two katydids (Pterophylla camel lifol ia, Scudderia rhombifolium). Other grazers common to Zones IV and V and abundant in the Congaree Swamp are the zebra swallowtail (Graphium marcel! us) (whose larvae feed on the paw paw), the Carolina satyr (Euptychia hermes sosybia), the red spotted purple (Limenitis archip- )us astanax) and the pearl crescent pus (Is Physoides tharos) butterflies. Of the spiders shared by Zones IV and V, the most abundant ground dwellers on a Piedmont floodplain are the wolf spiders (Schizocosa ocreata, Lycosa helluo), and in the Congaree, Schizocosa crassipes. In the Congaree the dominant aerial spiders are the orb weaver (Neoscona arabesca), the spinyback (Micrathena gracilis), and Frontinela spp. Most of the 11 species of snails recorded from Congaree probably inhabit Zone V. The dominant ones are the great zonite (Mesomphix vulgatus), the white- lipped forest snail (Mesodon thyroidus) 91 and the cannibal snail (Haplotrema con- cavum). Vertebrates Zone V shares vertebrate species coiri- mon to uplands as well as Zone IV, The large, spotted salamander (Ambystona macu- latum) and mole salamander (A. talpoideum) seem confined to Zone V. The red sala- mander (Pseudotriton ruber) is shared with Zone IV. Two common upland species are the ubiquitous sMmy salamander (Plethodon qlutinosus) and the red-backed salamander (Plethodon cinereus). Two toads, the nar- rowmouth (Gastrophryne carol inensis) and spadefoot (Scaphiopus holbrooki), Tnhabit sandier portions of Zone V. In Zone V are skinks of upland mesic slope forests, such as the ground skink (Leiolopisma) and Eumeces inexpectatus, in addition to JE. fasciatus. Among the snakes recorded are the copperhead, canebrake rattlesnake (Cratalus horridus atricaudatus), northern brown fStoreria dekayi ), garter (Thamno- phis sirtal is), rough green (Opheodrys aestivus), and ribbon snakes. Occasion- ally, even upland species such as the black racer and coachwhip (Masticophis flaqellum), are found. We do not know how many species migrate annually from upland areas into Zone V when high water recedes, or conversely, from Zone V to the uplands during short periods of high water. In the Congaree Swamp the common yellowthroat, pine warbler, wood thrush, and eastern wood peewee seem to prefer Zone V habitats. Zone V is perhaps the preferred nesting and feeding ground of the wild turkey (Figure 49). North Caro- lina's only breeding colonies of cerulean warblers (outside the Blue Ridge Moun- tains) and Mississippi kites occur in a 60-km (37-mi) section of old growth timber along the levees of the Roanoke River, two-thirds of which (a 20C-m or 656-ft wide strip) is dominantly Zone V vegeta- tion. A number of birds that are commoner in Zones IV and V than in other zones include the white-breasted nuthatch, Swainson's warbler, Carolina wren, and yellow-throated vireo. Breeding bird densities are generally higher in the floodplain than in adjacent upland forests (Dickson 1978). Kennedy (1977) noted that more birds preferred Zone IV and V hard- woods than other dominance types (e.g., cottonwood-wil low-sycamore or cypress- tupelo). Zones IV and V are the principcl environments of the rare and endangered ivory-billed woodpecker, Bachman's warbler and probably the cougar (Wharton et al, 1981). Black bears (on Bear Island) con- gregate on the higher, unlogged, acorn- rich Zone IV and V bottomlands. Upland forest forms sometimes occurring in Zone V are the least shrew (Cryptotis), pine vole (Pitimys). and, rarely, the common mole. Other mammals are the same as reported for Zone IV. Although Zone V environments may com- prise a relatively small part of the total floodplain acreage (for example, only 5% in Congaree Swamp), these old levee ridges are extremely important in the life histories of many floodplain species. They provide food, winter hibernacula, and for the more terrestrial forms, high water refuge and migration and dispersal routes. In a number of southern swamps lacking a Zone V, mounds of earth ("cattle mounts") often were constructed by early human residents to provide refuge for livestock during high water. Tinkle (1959) found narrow, long levees indispensible for the egg-laying activities of many amphibious snakes and turtles; he also discovered that the swamp palm (Sabal minor) growing there provided a major hibernaculum for small vertebrates. THE USE FISH OF BOTTOMLAND HARDWOOD ZONES BY Many fish species use Zones II through V during inundation. At least 20 families and up to 53 species of fish spawn and/or feed on the floodplain (Lambou 1963; Holder et al. 1970, 1971; Bryan et al. 1975, 1576; Huish and Pardue 1978; Walker 1980; Wharton et al. 1981; and others). The catfish, sunfish, gar, perch, and sucker families are particu- larly well represented. Fish depend on an annual water level fluctuation to limit intra- and interspe- cific competition for food, space, and spawning grounds (Lambou 1959). Fish dis- tribution and abundance are thus keyed to this cyclic phenomenon (Lambou 195S, 1962; Bryan and Sabins 1975; Hern et al. 1980). As most swamp-wise fishermen know, the time and extent of overflow control the size of the year classes of black bass and sunfish (Lambou 1962). On the Danube 92 Figure 49. bottomland leaves, an been found The crop contents of Food iteins include d fruits of hackberry, supplejack, and poison in turkey crops. Photo by Brooke Meanley. a wild turkey killed in April in the Arkansas River snails, scarabeid beetles, pecans, jack-in-the-pulpit ivy. In Florida, crayfish have floodplain (Germany) fish yield was 14.6 kg/ha (13 lb/acre) with a 20-day inunda- tion, increasing to 49.2 kg/ha (44 lb/ acre) with a 198-day inundation, with the delayed effects recognizable a year later (Stankovic and Jankovic 1971). The use of the floodplain by fishes in a blackwater creek (Creeping Swamp, Pitt County, NC) was studied by Walker (1980) by use of two-way weir traps in shallow drainways on the floodplain (Zone II) (Figure 50). With the exception of the redfin pickerel, fish moved on the floodplain only at night. Most fish were caught in January through March, the time of maximum inundation, although large fluctuations occurred at other times in these small streams (watershed approxi- mately 80 km2 or 31 mi 2). Common species (in order of abundance on the floodplain) were pirate perch, redfin pickerel, flier, mud sunfish, eastern mud minnow, American eel, banded sunfish, creek chubsucker, blue-spotted sunfish, redear sunfish (shellcracker), bowfin, shiner, brown bullhead, pumpkinseed, bluegill, golden shiner, warmouth, redbreast sunfish, swamp darter, and green sunfish. Included in the catch of the floodplain weirs were 928 adult crayfish (Procambarus acutus and Fallicambarus uhleri ), both floodplain varieties. Fish trapped on the Creeping Swamp floodplain feed on floodplain inverte- brates, principally copepods, ostracods, amphipods, isopods and midge fly larvae (Chironomidae) (Robert Sniffen, Institute of Marine and Coastal Research, East Caro- lina University, Greenville, NC; personal communication). Delicate forms such as oligochaete worms and flatworms (Planaria) disintegrate rapidly and leave few or no fragments; hence their con- fish diet may be underesti- Woodall et al. (1975) and (1976) on the Luxapalila River (MS, AL) found a preponderance of "terrestrial" invertebrates in stomachs of fish collected on the floodplain. Holder et al. (1970) compared the fish populations of inundated floodplains (Zone II) and sloughs of the Suwannee River. While the standing crop over the floodplain averaged much less (11-17 kg/ha or 10-15 lb/acre during the 3- to 10- month inundation period) compared to that of the sloughs (262 kg/ha or 234 lb/acre), 93 identifiable tribution to mated. Both Arner et al . Two-way traps with wire mesh wings, (Creeping Swamp, NC) and in shallow that 21 fish species used the floodplain extensively. With the exception of four species, more individuals were taken on the floodplain than in the channel. Although most fish utilized the floodplain from January through March, flooding occurred fre- quently at other times in this small watershed (80 km^ or 31 mi^) (Walker 1980). Figure 50. water creek set in this small Coastal Plain black- drainways on the floodplain, revealed 94 the surface area of the floodplain was much larger. Sloughs were satrpled after the water had ceased flowing off the floodplain, and fish were concentrated by falling water levels. Holder et al. (1970) stated that "high water over the floodplain provided space, food, and increased habitat for the reproduction and growth of fish. " Movement of fish on floodplains often is keyed to temperature. Holder et al. (1970) found ripe males and females of several species trying to cross the sill between the Okefenokee Swamp and the Suwannee River coincident with high water at the following time and water tempera- tures: fliers, bowfin (February, March, 11°-13°C or 52°-56°F); yellow and brown bullheads (March, 11°C or 52°F); warmouth (March-April, 16°-19°C or 60°-67°F); chain pickerel (March-April); lake chubsucker (April, 21°-24°C or 70°-76°F). Floodplains are important spawning areas for several species of herring (Clupeidae). Hickory shad (Alosa m.edio- is) ^spawn in oxbow lakes, sloughs, and cr tributary streams of the Altamaha River (GA) (between River Mile 20 and 137). Blueback herring (Alosa aestivalis) spawn in the same areas of the bottomland hard- woods; they have remarkably adhesive eggs which adhere to twigs and objects on the floodplain floor and resist being swept away by sheet flow. Ripe bluebacks were taken in an over 161-km (100-mi) long sec- tion of the Altamaha in backwater lakes and flooded low areas "that are accessible to these fish only during spring flood stages" (Adams and Street 1969). Studies of larval fish on the flood- plain or in sloughs and waterways deep within the floodplain suggest that the immature stages of roughly one half of the fishes of the lower Mississippi River used the floodplain as a nursery (Gallagher 1979). Analysis of the temporal, spatial, and size distribution of larval fishes supported this contention; spawning of 7 out of 10 of the most common taxa took place in backwater habitats (Atchafalaya Basin, LA) (Hall 1979). Temporal and spatial delineation of niches of larval fish on the floodplain have been summar- ized further by Larson et al. (1981) and Wharton et al. (1981). TROPHIC RELATIONSHIPS Energy flow in riverine systems involves both detritus and grazing path- ways. Although rivers appear to shift from autotrophy (predominantly grazing pathway) in mid-sections to heterotrophy (predominantly detritus pathway) in lower sections (Vannote et al. 1980), many lower river "detrital" food chains may still involve zooplankton "crazing" on phyto- plankton. For example, Wallace et al. (1977) found over 300,000 diatoms/liter in the lower Altamaha River (GA). The graz- ing pathway is important even in Coastal Plain blackwater streams; Patrick (1972) characterized these streams as being domi- nated by the diatom genera Eunotia and Actinel la. For clarity, trophic pathways on the floodplain have been divided into two systems (dry system pathways and wet system pathways). These two "systems" are not always clear cut. For example, mal- lards prefer to feed on acorns when they float during inundation. The dry system (Figure 51) is functional during drydown when the floodplain is not inundated. While the system is largely detrital, the grazing pathway of the terrestrial faunal assemblage is also pronounced, with appre- ciable consumption of the products (nuts, berries, leaves, bark) of the bottomland hardwoods and other primary producers. Trophic pathways in the litter layer of the dry system are similar to those in the uplands. Gist and Crossley (1975) in trophic studies of upland forest found that millipedes consume up to 120 g/m^/yr of deciduous litter detritus. Fungal hyphae were the principal food of snails and collembolans. The predatory mites consumed primarily collembolans. The second trophic system, (Figure 52) is a wet system functioning in pools and during inundation. Primarily detri- tal, it involves the bulk export of detri- tus into sloughs, oxbows and rivers, thus feeding a largely aquatic fauna. Aerial swarms of midge flies, mosquitoes, and mayflies emerge periodically, however, to regale legions of swifts, tree frogs, bats, and dragonflies. Since much of the energy of the wet system is exported, the process needs to be summarized in more detail. Detritus 95 ■M • "O Qi " •> (/I C -C (U 1- -C 2 iTJ fO I/) S_ /D (J fD C **- fD i/l (U ' ^ •.-I— 1_ D.. 3 — O •^ -M +J ; c ez fO C c o O -^ i- ^ - U O if O J3 i . O o . — irt CD i/J S- . OJ 3 C t. 3 E 01 ■-- -0)0; i TJ i- M t_ M jk: _ (_) -C fO fO (TJ .r- ^/l 1- OJ S- .— cnj3 cj> (/) 5 O) C C -o -M <— TS o c iJ O) u^ T3 fC OJ .^ >i O in OJ ■"" ^ 1- *>1 '^ '*" "^ J=i E TJ c JU ^ 3 x: 3 -o ^ -q s_ oj .ecu , >, u; 0> 3 fO J- 3 Q. 3 O • — O O E Of 1_ , — (O Q-*'- rtS O) i*_ ^ to E I- 3 jc O ■!— (O W1 u .i+_ - > m *J t/) (U I cx-o 3 C71 3 ■,- o ! t- t/>. > O) fD fO o dj . — to i/i di >. > to ■*-' I- C -o QJ " ». fO •■- C +-> IJ tj^ X Ol 13 '^ -^ -^ ^ -o 'i- 1- J3 CX *^ >> ^ ^^ h (/) Qj CD -C jii — 3 S- ^ 13 "O JT i/> U-) x: »- Qj Q-^ ■ ^ 13 - to _ - o J- i. — - c--^ >^ o a> ■»- 0* • .— L> > at f— 4-» u »— o c .c 3 u ■*-• o n ■r- h- o > QJ Cn e «TJ (U O '^-■ i- r- i— <_> >> Q.'*- C m O 4- • J^ i. o- O) •> >, fO 1- Jm£ t/> "D > o -o t/i fO (O " ■»-> C ^ E >» •.- (O u 3 Q) ^ -M- cr wi ■M X O) Ul wi - ? 4- -o -o £: ■- ^ -Q QJ ^ O Q. OJ -n 4- O; .. c cn TO QJ TO TO ,^ > 1 — cn TO TO t- CO 13 > E TO E (/I .— U (U <— TO •— QJ •M C t- QJ (/) TO ^ QJ -»-> U TO JD 4- 1/1 CU ^ O a> to CJl trt s- o c o *a -.- -a ■M TO o i_ -O >— to ■- TO S_ m (V ■o .— -* +-> (/) OJ t- « TO r- - S "H o <" Q. S_ _ tors ande eron TO E ^ TO XJ s- 5 m 4J TO c TO *^ Q. QJ ■♦-J E "3 . — o •D a, OJ >, 4- 2$ +-> ^ +J i;-*^ == (D -o 3 <" >] CJ QJ "O •^^ S- T3 3: S- TO c cn QJ OJ > TO - OJ S- in TO - u .^ to ^ O TO 3; -^ C Q. "^ +-> J=i 01 Q. •r- E -t- •- E TO ^ 3 . -C S. O- wi ^ •> O) to +J -C s +-> O M_ t/l ■o 3 -M E -n- I^ f^ o,— 5 3 '^ -a o ,— E to trt (- 4- (u TO >>•.- to u >* c= XI u ,^'-M TO i/l «TJ lU 1 — TO E QJ >^ QJ lyi "— !- 4-1 cn c 3 c >— ITJ 5- rtJ (J TO QJ -f— TO 4~ -1 o -^ ^ Ss TO S- Ql Ol 4-> TO ^ TO M- . t/> Q- ■*-> s- (U s- CNj :3 o "o c » *- ■K 0) ••- . ^ u o I- >.'*- to 3 Ul E c (b - Ol C X) TO O E !- •^ TO C .— a. O O Li_ f~ TO U 3 tn -*-> 97 (leaves, twigs) is in the form of coarse particulate organic matter (CPOM, particle size >1 mm), fine particulate organic matter (FPO^:, particle size <1 mm), or dissolved organic matter (DOM, parti- cles 10 mg/1 ) since signif- icant quantities are complexed (Reuter and Perdue 1977). Excess humic acids tend to immobilize mercury (Miller et al. 1975) and other heavy metals (Giesy and Briese 1978). The fate of these substances after complexation is important. Humic acids, being insoluble, tend to be immobilized as bottom sediments. Soluble lighter fulvic acids, constituting the bulk of DOC in southeastern rivers, sometimes complex with water contaminant metals; largely unavailable as microbial foods, they do not enter the downstream food chain (Reuter and Perdue 1977). Some of these flocculate with electrolytes at the inter- face of fresh- and saltwater. Another important consideration is the residence time required for processes involved in water quality improvements. Residence time is the time that water remains over the floodplain floor and in the sloughs and depressions. Examples of processes are conversion of DOM to POM by microbes (Slater 1954; Brinson et al. 1980) or by freezing (Giesy and Briese 1978), the complexing of metals with organic matter, the adsorption of ions by clay particles, and the establishment of reducing conditions essential to operate the metabolic pathways of sulfate reduc- tion, denitrification and methanogenesis. Obviously, any human activities which decrease residence time, such as channeli- zation, interfere with or eliminate these vital floodplain functions. MODIFICATIONS OF RIVER AND FLOODPLAIN Construction or other modifications on floodplains may cause profound changes in their ecology. Man's direct or indirect 105 manipulation of the hydrologic regime may result in a complete transformation of river and floodplain morphology (Schumn; 1969). Clearly, anything that man does to the natural, orderly river channel will induce changes as the river attempts to regain its original efficient configura- tion. Sediment inputs resulting from clearcutting, or sedin^ent starvation from reservoir construction, dredging, shorten- ing and even snagging and dragging are among potential impacts. The construction of reservoirs can have both direct and indirect effects, the result of coupling the natural energy of moving water with man's complex of activi- ties on the uplands, l^hile the dissolved solute chemistry of the water is not mark- edly changed, the sediment load settles and is reduced as a result of the stilling effect of the reservoir. Release of water from the reservoir results in scour and resuspension of sediments that move stead- ily downstream as the upper segment of the river lowers its bed (Schumm 1971). In the Red River below Denison Reservoir (OK), scour widened the channel from 1.5 to 3.0 m (5 to 10 ft) per year but the river did not regain its pre-reservoir sediment load for 322 km (200 mi) (Ein- stein 1972). The large dams above Augusta, GA, virtually have stopped sediment move- ment from the Piedmont (Meade 1976). Res- ervoirs on the Santee and Savannah trap from 85% to over 90% of incoming sediment. Since as much sediment still is carried in the lower reaches as in pre-dam days, Meade concluded that the immediate source must be the river bed, banks, and flood- plain. Other factors may include extensive shortening and dredging of the Savannah by the Corps of Engineers with corresponding gradient increase and much abnormal bank destruction. Meade stated that since 1910 reservoirs on the Savannah River have re- duced the sediment load delivered to the ocean by 50%, thus depriving the Continen- tal Shelf of its former river influx of inorganic nutrients. Flood peaks higher than the 1- to 2- year inundations are also impaired by dam- ming. Much riverborne sediment deposits in estuaries are in the "sediment-trap" at the freshwater-saltwater interface. In pre-dam days the locus of these deposits moved back and forth seasonally in the estuary, and floods flushed out accumu- lated sediments with a periodicity of 3 to 5 years (Meade 1976). This flushing action no longer occurs, and sediments accumulate in the estuary and must be removed by costly dredging. Another example of direct downstream effects is flow regulation by huge reser- voirs on the Roanoke River (NC). The Roanoke River (NC) has a "dead zone" 6- 12 m (20-40 ft) wide from near the levee top down to the river devoid of all bot- tomland hardwoods except a few willows at the upper edge. This zone, with numerous fallen dead trees, resulted from water levels held artificially high by upstream discharges, well into leafout time. In swales along the Roanoke the lack of former flushing action may have caused several feet of silt to accumulate (Pat White, Williams Lumber Company, Mackeys, NC ; personal communication). Indirectly, flow regulation is having an even more pronounced effect on bottom- land hardwood communities. Damming may severely modify or eliminate the seasonal hydroperiod, allowing upland row-crop agriculture and forestry on the flood- plain. On the Savannah River (GA) flood- plain, below a series of large reservoirs in the Piedmont, hundreds of acres of Zone IV bottomland hardwoods are being sheared off, and the floodplain is being planted in soybeans. Even without flow regula- tion, the floodplains of many southern rivers are being cleared of bottomland hardwoods and prepared for conversion to pine plantations, although, ironically, hardwoods often outgrow pines on these sites. Planted loblolly stands on the Altamaha, Oconee, and Ocmulgee Rivers in Georgia have developed a thick understory of wax myrtle and sweetgum, typical Zone IV species, which may indicate the land is still most suitable to the natural commun- ity. Such system-wide changes threaten or eliminate the life support functions of floodplain zones. The most serious impacts of reser- voirs on bottomland hardwoods downstream arise from regulating the normal annual rise and fall of the river to which the whole system is keyed. The effects of 106 reservoirs can be summarized: (1) reduc- tion of silt and associated nutrient inputs for some distances below dams, (2) excessive bed and bank scour below dams with accompanying modification or exter- mination of benthic and epibenthic fauna, (3) loss of bank-stabilizing vegetation by frequent flow changes, (4) disruption of normal fish breeding and feeding on the floodplain, (5) elimination of sufficient high v/ater for the annual flushing of detritus from the floodplain, and (6) encouragement of clearcutting, site con- version to tree plantations and row crop agriculture on formerly saturated flood- plains. CHEMICAL COUPLING WITH THE WATER TABLE AND ATMOSPHERE It is seldom acknowledged that Pied- mont alluvial rivers crossing the Coastal Plain may have numerous blackwater tribu- taries, the DOM of which is subsequently camouflaged by suspended inorganics. The lower sections of alluvial rivers are, in effect, chemically mixed rivers, in pro- portion to the respective discharges that each type stream contributes. Likewise, there is lateral coupling with the under- ground aquifer in limesink zones. At times of high water, acid blackwater may enter and corrode underground corridors; con- versely, aquifer flow at times raises the pH and the nitrate level of the river (Kaufman and Dysart 1978). Wetlands, including bottomland hard- woods, modify temperature and moisture content of the lower atmosphere. They ameliorate freeze conditions and provide a more equithermal refuge for many animals which could not otherwise exist at that latitude. Wetlands modify lake and sea breezes, the urban boundary layer, and even the behavior of tropical cyclones. In Florida, relatively minor changes in land-water coverage and soil moisture result in surprisingly large changes in sea breezes, cumulus cloud formations, and precipitation (Gannon et al. 1978). COUPLING VIA FAUNAL MOVEMENTS Rivers and their floodplains are also coupled with marine systems through anadromous, catadromous and other marine species. Blue crabs occur to RM (river mile) 50 in the Altamaha. The southern flounder (Paralichthys lethostigma) and striped mullet (Mugil cephalus) migrate and feed as far as 193 km (120 mi ) up the Altamaha, while two common coastal fishes, the hogchoker (Trinectes maculatus) and the needlefish (Strongylura marina), actu- ally spawn in the mid-reaches of the river (John Adams, Georgia Power Co., Environ- mental Laboratory, Atlanta; personal com- munication). Various shad species parti- tion the river into spawning and nursery sections (Adams and Street 1969; Adams 1970). American shad (Alosa sapidissima) spawn in the Altamaha itself between RM 60 and 120, with primary nursery centers at RM 21-30 and RM 100-110. Hickory shad (A. miediocris) spawn in floodplain oxbows, sloughs, and tributaries between RM 20 and 137. Blueback herring (A. aestivalis) spawn on the floodplain floor between RM 5 and 137, with primary nurseries between RM 10 and 30. Examples of catadromous spe- cies are American eel and mountain mullet (Aqonostomus monticola), the latter in gulf coast rivers. Striped bass formerly traveled up the Savannah as far as Tallulah Gorge and still ascend many Coastal Plain streams. Two species of sturgeon (Atlantic and shortnosed) ascend rivers entering the Atlantic slope. Other animals such as manatees use the Altamaha at least to the limit of tidal range and go up the Suwan- nee to Manatee Springs. The glochidian larvae of many clams can travel up- and downstream attached to fishes, often providing a mechanism for repopulating depleted areas. Terrestrial fauna may be coupled to the uplands, as when deer who base their home range on floodplains graze in up- lands. Conversely, upland forms such as the black racer and pine vole may use the floodplain at drydown. The narrow green- belts of bottomland hardwoods also provide routes for migration and restocking. SUMMARY In summary, bottomland hardwood swamps are integrally coupled to the sur- rounding uplands, downstream estuarine 107 systems, and the atmosphere. By virtue of of an apparent resilience to specific these couplings, the swamps provide inval- disturbance, the ecological values of the uable services and resources to the envi- bottomland hardwoods can be impaired by ronment. These "ecological values" are a development or alterations which do not function of the interaction of the bottom- take into account the openness of these land hardwood ecosystem and its primary ecosystems to the riverine and upland driving force, the fluctuating water ecosystems to which they are hydrologi- levels of the riverine systems. In spite cally coupled. 108 REFERENCES Adams, J.G. 1970. Clupeids in the Alta- piaha River, Georgia. Contrib. Ser. No. 28. Coastal Fisheries Div., Game and Fish Comm., Ga. Dep. of Nat. Resour. , Brunswick. Adams, J.G., and M.W. Street. 1969. Notes on the spawning and embryological development of blue-back herring (Alosa aestivalus Mitchell) in the Altamaha River, Georgia. Contrib. Ser. No. 16, Marine Div., Coastal Fisheries Div., Game and Fish Comm., Ga. Dep. of Nat. Resour., Brunswick. Allen, J.R.L. 1965. A review of the ori- gin and characteristics of recent alluvial sediments. Sedimentology 5(2): 89-191. AUman, P.L., and D.S. Dittmer. 1966. Environmental biology. Comm. on Biol. Handbooks, Fed. of Am. Soc. for Exp. Biol . 694 pp. Applequist, M.B. 1959. A study of soil and site factors affecting the growth and development of swamp blackgum and tupelo gum stands in southeastern Georgia. Ph.D. dissertation. Duke Univ., Durham, N.C. 181 pp. Armstrong, W. from the Physicol . 1968. Oxygen diffusion roots of woody species. Plant. 21: 539-593. Arner, D.H., H.R. Robinette, J.E. Fraiser, and M.H. Gray. 1976. Effects of channelization of the Lutapalila River on fish, aquatic invertebrates, water quality and fur bearers. U.S. Fish Wildl. Serv. Off. Biol. Serv. FWS/OBS-76-08. Bartram, W. 1971. Travels through North and South Carolina, Georgia, east and west Florida. Q.M. van Doren, ed. Dover publications (1928 reprint). Beard, L.S. 1958. A floralistic study along the Lower Deep River of North Carolina. Master's Thesis. Univ. of N.C. at Chapel Hill. Beaven, G.F., and H.J. Costing. 1939. Pocomoke Swamp: a study of a cypress swamp on the Eastern Shore of Mary- land. Bull. Torrey Bot. Club 66: 367-389. Beck, K.C., J.H. Rentes, and E.M. Perdue. 1974. Organic and inorganic geo- chemistry of some coastal plain rivers of the Southeastern United States. Geochim. Cosmochim. Acta 38: 341-364. Beck, L.T. 1977. Distribution and rela- tive abundance of freshwater macro- invertebrates of the lower Atchafa- laya River Basin, Louisiana. M.S. Thesis. La. State Univ., Baton Rouge. Beck, W.M. 1965. The streams of Florida. Bull. Fla. State Mus. 10(3): 91-126. Bedinger, M.S. 1978. Relation between forest species and flooding. Pages 427-435 in P.E. Greeson, J. P. Clark and J.E. Clark. Wetland functions and values: the state of our under- standing. Am. Water Resour. Assoc, Minneapolis, Minn. Bedinger, M.S. 1980. Hydrology of bottom- land hardwood forests of the Missis- sippi Embayment. (Unpubl. ms. draft.) ■nke, A.C., D.M. Gillespie, and F.K. Par- rish. 1979. Biological basis for assessing impacts of channel modifi- cation: invertebrate production, drift, and fish feeding in a south- eastern blackwater river. School of Biol, and Environ. Resour. Center, Ga. Inst. Tech., Atlanta. ERC 06-79. 187 DD. 109 Binford, M.W. 1977. Crustacean zooplank- ton ecology of the Atchafalaya River Basin, Louisiana. M.S. Thesis. La, State Univ., Baton Rouge. Blench, T. 1972. Morphometric changes 1. Pages 287-308 jn R.T. Glesby, C.A. Carlson and J. A. McCann, eds. River ecology and man. Academic Press, New York. 465 pp. Boto, K.G., and W.H. Patrick, Jr. 1978. Role of wetlands in the removal of suspended sediments. Pages 479-489 in P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland functions and values: the state of our understand- ing. Am. Water Resour. Assoc, Minn- eapolis, Minn. Boyce, S.G., and N.D. Cost. 1974. Timber potentials in the wetland hardwoods. Pages 131-151 in N.C. Blount, ed. Water resources, utilization and con- servation in the environment. Taylor Printing Co., Reynolds, Ga. Boyd, H.E. 1976. Biological productivity in two Georgia Swamps. Ph.D. Disser- tation. Univ. of Tenn., Knoxville. 98 pp. Bozeman, J.R. 19G4. Floristic and edaphic studies of the Altamaha River sand ridge, Georgia. M.S. Thesis. Univ. of N.C. at Chapel Hill. Brady, N.C. 1974. The nature and proper- ties of soils. 8th ed. MacMillan Publ. Co., New York. 639 pp. Braun, E.L. 1950. Deciduous forests of eastern North America. Blakeston Co., Philadelphia, Pa. 596 pp. Brinson, M.M. 1977. Decomposition and nutrient exchange of litter in an alluvial swamp forest. Ecology 58: 601-609. Brinson, M.M., H.D. Bradshaw, R.N. Holmes, and J.B. Elkins, Jr. 1980. Litter- fall, stemflow, and throughfall nutrient fluxes in an alluvial swamp forest. Ecology 61(14): 827-835. Brinson, M.M., R. Plantico, B.L. Swift, and J.S. Barclay. 1981. Functions, values and management of riparian ecosystems. U.S. Fish Wildl. Serv. Biol. Serv. Program FWS/OBS-draft. 5roadbent, F.E. fractions. 1953. The soil organic Adv. Agron. 5: 153-184. Broadfoot, W.M. 1967. Shallow water impoundment increases soil moisture and growth of hardwoods. Soil Sci. Soc. Am. Proc. No. 34: 562-664. Broadfoot, W.M., and H.L. Williston. 1973. Flooding effects on southern forests. J. For. 71(9): 584-587. Brown, S. 1978. A comparison of cypress ecosystem and their role in the Flor- ida landscape. Ph.D. Dissertation. Univ. of Fla., Gainesville. 569 pp. Brown, S. 1981. A comparison of the structure, primary productivity, and transpiration of cypress ecosystems in Florida. Ecol . Monogr. 51(4): 403-427. Brown, S. , S. Bayley, and J. Zoltek. 1974. Preliminary results of longterm ef- fects of sewage effluent on water quality and tree growth in swamp- lands. Dep. of Engineering Sci., Univ. of Fla., Gainesville. Brown, S., M.M. Brinson, and A.E. Lugo. 1979. Structure and function of riparian wetlands. Pages 17-31 in R.R. Johnson and J.F. McCormick, tech. coords. Strategies for protec- tion and management of floodplain wetlands and other riparian ecosys- tems. U.S. For. Serv., Washington, D.C. Brupbacher, R.H., J.E. Sedberry, Jr., W.P. Bonner, W.J. Peavy, and W.H. Willis. 1970. Fertility levels and lime sta- tus of soils in Louisiana. Bull. 644. Agric. Exp. Stn., Dep. of Agronomy, La. State Univ., Baton Rouge. 120 pp. Bryan, C.F., and D.S. Sabins. 1979. Man- agement implications in water quality and fish standing stock information in the Atchafalaya River Basin, Louisiana. Pages 193-316 in J.W. Day, Jr., D.D. Culley, Jr., R.E. Turner and A.J. Humphrey, Jr., eds. 110 Proceedings Third Coastal Marsh and Estuary Symposium. La. State Univ. Div. of Continuing Ed., Baton Rouge. Bryan, C.F., P.M. Truesdale, and D.S. Sabins. 1975. Annual report, a limnological survey of the Atchafa- laya Basin. La. Coop. Fish. Unit, La. State Univ., Baton Rouge. 203 pp. Bryan, C.F., D.J. DeMont, D.S. Sabins, and J. P. Nevflnan, Jr. 1976. A limnologi- cal survey of the Atchafalaya Basin. Annu. Rep. La. Coop. Fish. Res. Unit, La. State Univ., Baton Rouge. 285 pp. Bryan, K. 1940. phenomena in 91: 523-524. Soils and periglacial the Carol inas. Science Buell, M.F. 1946. Jerome Bog, a peat filled "Carolina bay." Bull. Torrey Bot. Club 73: 24-33. Burkholder, P.R., and L.M. Burkholder. 1956. Vitamin B12 in suspended solids and marsh muds collected along the coast of Georgia. Limnol. Oceanogr. 1(3): 202-208. Burns, L.A. 1978. Productivity, biomass and water relations in a Florida cy- press forest. Ph.D. Dissertation. Univ. of N.C. at Chapel Hill. Burton, T.M. , and G.E. Likens. 1975. Energy flow and nutrient cycling in salamander populations in the Hubbard Brook Experimental Forest, New Hamp- shire. Ecology 56: 1068-1080. Butler, T.J. 1975. Aquatic metabolism and nutrient flux in a south Louisi- ana swamp and lake ecosystem. M.S. Thesis. La. State Univ., Baton Rouge. 58 pp. Callahan, J.T. 1964. The yield of sedi- mentary aquifers of the Coastal Plain, southeast river basins. U.S. Geol. Surv. Water Supply Pap. 1669-W. 56 pp. Carlston, C.W. 1965. The relation of free meander geometry to stream dis- charge and its aeomorphic implica- tions. Am. J. Sci. 263: 864-885. Carter, M.R., L.A. Burns, T.R. Cavinder, K.R. Dugger, P.L. Fore, D.B. Hicks, H.L. Revello, and T.W. Schmidt. 1973. Ecosystem analysis at the Big Cypress Swamp and estuaries. U.S. Environ. Protect. Agency Region IV, Atlanta, Ga. EPA 904/9-74-002. Carter, V., M.S. Bedinger, R.P. Novitzki, and W.O. Wilen. 1978. Water re- sources and wetlands. Pages 344-376 in P.E. Greeson, J.R. Clark, J.E. Clark, eds. Wetland functions and values: the state of our understand- ing. Am. Water Resour. Assoc, Minn- eapolis, Minn. 674 pp. Carver, R.E. 1971. Holocene and late Pleistocene sediment sources. Con- tinental Shelf off Brunswick, Geor- gia. J. Sediment. Petrol. 41(2): 517-525. Cauthron, F.F. 1961. A survey of inverte- brate forms of the Mississippi River in the vicinity of Baton Rouge, Loui- siana. M.S. Thesis. La. State Univ., Baton Rouge. 41 pp. Chabreck, R.H. 1978. Wildlife harvest in wetlands of the United States. Pages 618-631 _iji P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland func- tions and values: the state of our understanding. Am. Water Resour. Assoc, Minneapolis, Minn. Chirkova, T.V. 1968. Oxygen supply to roots of certain woody plants kept under anaerobic conditions. Fizol. Past. 15: 565-568. Chirkova, T.V., and T.S. Gutnan. 1972. Pnysiological role of branch lenti- cels in willow and poplar under conditions of root anaerobiosis. Fizol. Rast. 19: 362-385. Clark, J.R., and J. Benforado, eds. 1981. Wetlands of bottomland hardwood forests. Proceedings of a workshop on bottomland hardwood forest wet- lands of the Southeastern United States held at Lake Lanier, Georgia, June 1-5, 1980. Developments in Agricultural and Managed-forest Eco- logy, vol. 11. Elsevier Scientific Publ. Co., New York. 402 pp. Ill Conner, W.H., and J.W. Day. 1976. Pro- ductivity and composition of a bald cypress-water tupelo site and a bottomland hardwood site in a Loui- siana swamp. Am. J. Bot. 63: 1354- 1364. Copeland, B.J. 1966. Effects of decreased river flow on estuarine ecology. J. Water Pollut. Control Fed. 38(11): 1831-1839. Cosner, O.T. 1962. Geology of part of the Ocmulgee River flood plain at Macon, Georgia. (Unpubl. ms.) U.S. Geol. Surv., Richmond, Va. Costa, T.E. 1974. Response and recovery of a Piedmont watershed from tropical storm Agnes, June 1972. Water Resour. Res. 10(1): 106-112. Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of wetlands and deepwater habitats of the United States. U.S. Fish Wildl. Serv. Biol. Serv. Program FWS/OBS- 79/31. Supt. of Documents, Washing- ton, D.C. 103 pp. Crawford, R.M.M. 1976. Tolerance of an- oxia and the regulation of glycolysis in tree roots. Pages 388-401 jji M.G.R. Cannel and F.T. Last, eds. Tree physiology and yield improve- ment. Academic Press, New York. Crawford, R.K.M., and P.D. Tyler. 1969. Organic acid metabolism in relation to flooding tolerance in roots. J. Ecol. 57: 237-246. Crow, J.H., and K.B. MacDonald. 1978. Wetland values: secondary production. Pages 146-161 jm P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland functions and values: the state of our understanding. Am. Water Resour. Assoc, Minneapolis, Minn. Cummins, K.W. 1973. Trophic relations of aquatic insects. Annu. Rev. Entomol. 180: 183-206. Cummins, K.W., and G.L. Stream ecosystems. (Fall). 1-9 pp. Spengler. 1978. Water Spectrum Curry, R.R. 1972. Rivers - a geomorphic and chemical overview. Pages 9-31 in R.T. Oglesby, C.A. Carlson and J.TTT McCann, eds. River ecology and man. Academic Press, New York. 465 pp. Day, J.W., Jr., T.J. Butler, and U.H. Conner. 1977. Production and nutri- ent export studies in a cypress swamp and lake system in Louisiana. Pages 255-269 in M. Wiley, ed. Estu- arine processes. Vol. 2. Academic Press, New York. Deevey, E.S. 1979. In defense of mud. Bull. Ecol. Sec. Am. 51(1): 5-8. Delaune, R.D., W.H. Patrick, and J.N. Brannon. 1976. Nutrient transforma- tions in Louisiana salt marsh soils, Publ. LSU-T-76-C09. Center for wet- land resources, La. State Univ., Baton Rouge. 38 pp. Delcourt, P. A. 1980. Late Quaternary environmental history of the Little Tennessee River Valley, southeastern Tennessee. Bull. Ecol. Soc. Am. 61(2): 120. Dennis, W.M. 1973. A synecological study of the Santee Swamp, Sumter County, South Carolina. M.S. Thesis. Univ. of S.C. , Columbia. 61 pp. Dickson, J.C. 1978. Seasonal bird popu- lations in a south central Louisiana bottomland hardwood forest. J. Wildl. Manage. 42(4): 875-883. Drobney, R.D. 1977. The feeding ecology, nutrition and reproductive bioener- getics of wood ducks. Ph.D. Disserta- tion. Univ. of Mo., Columbia. 170 pp. Dury, G.H. 1977. Underfit streams: ret- rospect, perspect and prospect. Pages 281-293 vn K.J. Gregory, ed. River channel changes. John Wiley and Sons, New York. 448 pp. Eargle, D.H. 1940. The relation of soils and surface in the South Carolina Piedmont. Science 91: 337-338. Einstein, H.A. 1972. Sedimentation (sus- pended solids). Pages 309-318 in_ 112 R.T. Oglesby, C.A. Carlson and J. A. McCann, eds. River ecology and nan. Academic Press, New York. 465 pp. Ellis, M.N. 1936. Croseon silt as a fac- tor in aquatic environments. Ecology 17(1): 29-42. Eyde, R.H. 1966. The Nyassaceae in the Southeastern United States. J. Arnold Arbor. Harv. Univ. 47: 117-125. Eyre, F.H., ed. 1980. Forest cover types of the United States and Canada. Soc. of Am. For., Washington, D.C. 148 pp. Fisher, S.G. 1977. Organic matter pro- cessing by a stream-segment ecosys- tem. Fort River Massachusetts, U.S.A. Int. Rev. Gesamten. Hydro- biol. 62(6): 701-727. Fisk, H.N. 1944. Geological investiga- tion of the alluvial valley of the lower Mississippi River. U.S. Army Corps of Engineers, Miss. River Comm. , Vicksburg, Miss. 78 pp. Fisk, H.N. 1947. Fine grained alluvial deposits and their effects on Missis- sippi River activity. Miss. River Comm. Waterways Exp. Stn. , Vicksburg, Miss. 82 pp. Fisk, H.N. 1951. Mississippi River valley geology relation to river regime. Proceed. Am. Soc. Civil Engineers 77(80): 1-16. Florida Game and Fresh Water Fish Commis- sion. 1978. Conceptual fish and wildlife management plan for lower Apalacicola environmentally endan- gered lands. Office of Environ. Serv., State of Fla., Tallahassee. (Unpubl. draft. ) Forest Service Florida's For. Serv. 1971. Forest Service Georgia 's For. Serv. 1974. Resource Bulletin. 1970. timber. U.S. Dep. Agric. Resour. Bull. SE-20, Feb. Resource timber. Resour. Bulletin. 1972. U.S. Dep. Agric. Bull. SE-27, May Forest Service Resource Bulletin. 1974. North Carolina's timber. U.S. Dep. Agric. For. Serv. Resour. Bull. SE-33, Dec. 1975. Forest Service Resource Bulletin. 1978. South Carolina's forests. U.S. Dep. Agric. For. Serv. Resour. Bull. SE-51, Fredrickson, L.H. 1978. Lowland hardwood wetlands: current status and habitat values for wildlife. Pages 296-306 vn P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland functions and values: the state of our understand- ing. Am. Water Resour. Assoc, Minneapolis, Minn. Fredrickson, L.H. 1979. Floral and faunal changes in lowland hardwood forests in Missouri resulting from channeli- zation, drainage and impoundment. U.S. Fish Wildl. Serv. Biolog. Serv. Program FWS/OBS-78/91. 130 pp. Fredrickson, L.H. 1981. Management of lowland hardwood wetlands for wild- life - problems and potential. (Un- publ. paper.) Univ. of Mo., Gaylord Memorial Lab., Fuxico, Mo. Froehlich, W.,L. Kaszowski, and L.Starkel. 1977. Studies of present-day and past river activity in the Polish Carpath- ians. Pages 411-428 jm K.J. Gregory, ed. River channel changes. John Wiley and Sons, New York. 448 pp. Gaddy, L.L., T.H. Kohlsaat, E.A. Laurent, and K.B. Stansell. 1975. A vegeta- tion analysis of preserve alterna- tives involving the Beidler Tract of the Congaree Swamp. Div. Nat. Area Acquisition and Resour. Planning, S.C. Wildl. and Mar. Resour. Dep. 109 pp. Gagliano, S.M., and B.G. Thorn. 1967. Deweyville terrace, gulf and Atlantic coasts. Pages 23-41 j_n Technical Report 39. Bull. No. 1, La. State Univ. Studies, Baton Rouge. Gallagher, R.P. 1979. Local distribu- tion of ichthyoplankton in the lower 113 Mississippi River, Louisiana. Thesis. La. State Univ., Baton Rouge. 52 pp. Gannon, P.T., Sr., J.F. Bartholic, and R.G. Bill, Jr. 1978. Climatic and meteorological effects in wetlands. Pages 576-588 in P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland functions and values: the state of our understanding. Am. Water Resour. Assoc, Minneapolis, Minn. Gardner, J. A., Jr., W.R. Woodall, Jr., and J.G. Adams. 1975. A preliminary study of the fauna of the Altamaha River. Paper, 2nd Thermal Ecol. Symp., Augusta, Ga., April 2-5. Garten, C.T., Jr., L.A. Briese, R.A. Geiger, R.R. Sharitz, and M.H. Smith. 1975. Radiocesium levels in vegeta- tion colonizing a contaminated flood- plain. Pages 489-497 in F.G. Howell, J.B, Gentry and M.H. Smith, eds. Mineral cycling in southeastern eco- systems. ERDA Symposium Ser. Conf. 740513. Gasaway, R.D. 1973. Study of fish move- ments from tributary streams into the Suwannee River. Annu. Prog. Rep., Statewide Fisheries Investigation. F-21-5, Study VI, job 2. Game and Fish Comm. , Ga. Dep. of Nat. Resour., Atlanta. Gemborys, S.R., and E.J. Hodgkins. 1971. Forests of small stream bottoms in the coastal plain of southwestern Alabama. Ecology 52(1): 70-84. Georgia Power Company. 1971. (Feb.) Envi- ronmental report. Edwin I. Hatch Nuclear Plant, Ga. Power Co., At- lanta. Section II, 25 pp.; Section III, 7 pp. Gibbs, R.J. 1970. Mechanisms controlling world water chemistry. Science 170: 1088-1090. Giesy, J. P., and L.A. Briese. 1978. Par- ticulate formation due to freezing humic waters. Water Resour. Res. 14(3): 542-544. Gist, C.S., and D.A. Crossley, Jr. 1975. A model of mineral -element cycling for an invertebrate food web in a southeastern hardwood forest litter community. Pages 84-106 i_n F.G. Howell, J.B. Gentry and M.H. Smith, eds. Mineral cycling in southeastern ecosystems. ERDA Symposium Ser. Conf. 740513. Golkin, K.R. 1981. A computer simulation of the carbon, phosphorus, and hydro- logic cycles of a pine flatwoods eco- system. M.S. Thesis. Univ. of Fla., Gainesville. 225 pp. Gosselink, J.G., and R.E. Turner. 1978. The role of hydrology in freshwater wetland ecosystems. Pages 63-78 J_n R.E. Good, D.F. Wingham and R.L. Simpson, eds. Freshwater wetlands: ecological process and management potential. Academic Press, New York. Gosselink, J.G., S.E. Bayley, W.H. Conner, and R.E. Turner. 1981. Ecological factors in the determination of riparian wetland boundaries. Pages 199-219 in J.R. Clark and R. Benfo- rado, eds. Wetlands of bottomland hardwood forests. Proceedings of a workshop on bottomland hardwood forests of the Southeastern United States held at Lake Lanier, Georgia June 1-5, 1980. Development in Agri- cultural and Managed-forest Ecology, vol 11. Elsevier Scientific Publ. Co. , New York. Grable, A.R. 1966. Soil plant growth. in A.G Academic Press, Inc. aeration and Norman, ed. , New York. Grant, R.R., Jr., and R. Patrick. 1979. Tinicum Marsh as a water purifier. Pages 105-123 in Two studies of Tini- cum Marsh. Conserv. Found., Washing- ton, D.C. Gregory, K.J. 1977. The context of river channel changes. Pages 1-12 in K.J. Gregory, ed. River channel changes. John Wiley and Sons, New York. Grey, W.F. 1973. An analysis of forest invertebrate populations of the Santee-Cooper Swamp, a floodplain habitat. M.S. Thesis. Univ. of S.C. , Columbia. 114 Hack, J.T., and J.C. Goodlett. 1960. Geo- morphology and forest ecology of the mountain region of central Appalach- ians. U.S. Geol. Surv. Prof. Pap. 347. 66 pp. Haines, B.L. 1S81. Predicted nutrient losses from pine and hardwood forest soils at Coweeta under nine regimes of artificial acid rain. Dep. of Bot- any, Univ. of Ga. , Athens. (Unpubl.) Haines, E.B. 1975. Nutrient inputs to the coastal zone: the Georgia and South Carolina shelf. Pages 303-324 in L.E. Cronin, ed. Estuarine re- search. Vol. 1. Academic press. New York. Hall, C.A.S. 1971. Migration and metabo- lism in a stream ecosystem. Rep. 49. Water Resour. Res. Inst., N.C. State Univ., Raleigh. Hall, H.D. 1979. The spatial and temporal distribution of ichthyoplankton of the upper Atchafalaya Basin. Thesis. La. State Univ., Baton Rouge. 60 pp. Hall, R.J. 1976. A preliminary report on the herpetological survey conducted in Four Hole Swamp, 25 March-18 Oct., 1976. Natl. Audubon Soc. (Unpubl.) Hall, T.F., and G.E. Smith. 1955. Effects of flooding on woody plants. West Sandy Dewatering project, Kentucky Reservoir. J. For. 53: 281-285. Hamel, P.B. 1979. A study of the breed- ing birds of the Beidler Tract in the Congaree Swamp, Richland County, South Carolina. Unpubl. progress rep. to Columbia, S.C. Audubon Soc. 20 pp. Hamel, P.B., and N.L. Brunswig. 1980. Birds of the Francis Beidler Forest in Four Hole Swamp, South Carolina. Chat (in prep. ). Happ, S.C. 1945. Sedimentation in South Carolina Piedmont Valleys. Am. J. Sci. 243(3): 113-126. Happ, S.C, G. Rittenhouse, and G.C. Dobson. 1940. Some principles of accelerated stream and valley sedi- mentation. U.S. Dep. Agric. Tech. Bull. 695. U.S. Supt. of Documents, Washington, D.C. Harms, W.R. 1973. Some effects of soil type and water regime on growth of tupelo seedling. Ecology 54: 188- 193. Harper, M. 1938. The ecological distri- bution of earthworms as found in the developmental stages of the flood- plain. M.S. Thesis. Univ. of 111., Urbana. 25 pp. Harvey, A.M., D.H. Hitchcock, and D.J. Hughes. 1979. Event frequency and morphological adjustment of fluvial systems in upland Britain. Pages 139-167 j_n D.D. Rhodes and G.P. Williams, eds. Adjustments of the fluvial system. Kendall-Hunt Publ. Co., Dubuque, Iowa. 372 pp. Hay, J. 1977. A comparative analysis of ^^^Cs dynamics in two floodplain forests along a southeastern Coastal Plain stream. Dissertation. Emory Univ. , Atlanta, Ga. Hebert, C.E. 1977. A population study of small mammals in the Atchafalaya River Basin, Louisiana. M.S. Thesis. La. State Univ., Baton Rouge. 96 pp. Hendrickson, 0. 1981. Outputs of gaseous nitrogen and carbon from riparian strips on an agricultural watershed [tentative title]. Ph.D. Disserta- tion. Univ. of Ca., Athens. Hern, S.C, V.W. Lambou, and J.R. Butch. 1980. Descriptive water quality for the Atchafalaya Basin, Louisiana. EPA-600/4-80-04. Environmental moni- toring ser. 168 pp. Hester, J.M., Jr. 1973. Productivity, en- ergetics, and food webs. Pages 59-67 j[n North Carolina swamps. Unpubl. ms. Dep. of Environ. Sci. and Engineer- ing, Univ. of N.C. at Chapel Hill. Heuer, E.T., Jr. 1976. Relative abun- dances of squirrel and rabbits in three forest types of the Atchafalaya 115 River Basin, Louisiana. M.S. Thesis. La. State Univ., Baton Rouge. Hodges, J.D., and G.L. Switzer. 1979. Some aspects of the ecology of south- ern bottomland hardwoods. Journal pap. 4087. Miss. Agric. and Forestry Exper. Stn. Pages 360-365. Holder, D.R. 1970. A study of fish move- ments from the Okefenokee Swamp into the Suwannee River. Sports Fisheries Div., Ga. Game and Fish Comm. , Ga. Dep. Nat. Resour., Atlanta. Holder, D.R., L. McSwain, W.D. Hill, Jr., W. King, and C. Sweet. 1970. Popu- lation studies of streams. Statewide Fisheries Investigation, Annu. Prog. Rep. F-21-2, Study XVI. Game and Fish Comm. , Ga. Dep. of Nat. Resour., Atlanta. Holder, D.R., J. Sandow, L. McSwain, W.D. Hill, Jr., W. King, and C. Sweet. 1971. Population studies of streams. Statewide Fisheries Investigation Annu. Prog. Rep. F-21-3, Study XVI, job 2. Game and Fish Comm., Ga. Dep. of Nat. Resour., Atlanta. Holland, L.E., C.F. Bryan, and J. P. New- man, Jr. 1980. Some relationships between water quality and rotifer plankton in the Atchafalaya River Basin, Louisiana. Contrib. 32, La. Coop. Fishery Res. Unit., La. State Univ. , Baton Rouge. Hook, D.D., and C.L. Brown. 1973. Root adaptations and relative flood toler- ance of fine hardwood species. For. Sci. 19(3): 225-229. Hook, D.D., C.L. Brown, and P.P. Kormanik. 1970. Lenticel and water root devel- opment of swamp tupelo under various flooding conditions. Bot. Gaz. 131: 217-224. Houck, D.F. 1956. Flood plain flora of the Deep River Triassic Basin. M.A. Thesis. Univ. of N.C. at Chapel Hill. 52 pp. Huffman, R.T. 1979. The relation of flood timing and duration to varia- tions in bottomland hardwood forest community structure. Page 106 j_n Strategies for protection and manage- ment of floodplain wetlands and other riparian ecosystems. Publ. GTR-WO-12, U.S. For. Serv. , Washington, D.C. Huffman, R.T., and S.W. Forsythe. 1981. Bottomland hardwood forest communi- ties and their relation to anerobic soil conditions. Pages 187-196 vn J.R. Clark and J. Benforado, eds. Wetlands of bottomland hardwood forests. Proceedings of a workshop on bottomland hardwood forest wet- lands of the Southeastern United States held at Lake Lanier, Georgia June 1-5, 1980. Developments in Agricultural and Managed-forest Ecol- ogy, vol. 11. Elsevier Scientific Publ . Co. , New York. Huish, M.T., and G.B. Pardue. 1978. Eco- logical studies of one channelized and two unchannelized wooded coastal swamp streams in North Carolina. U.S. Fish Wildl. Serv. Biol. Serv. Program FWS/OBS-78/85. 72 pp. Hunt, L. 1977. "Turtling" in southern Georgia. Bull. Ga. Herp. Soc. 3(4): 3-4. Jahn, L.R. 1979. Values of riparian hab- itats to natural ecosystems. Pages 157-160 ijT^ Strategies for protection ana management of floodplain wetlands and other riparian ecosystems. Publ. GTR-WO-12, U.S. For. Serv., Washing- ton, D.C. Janzen, D.H., and P.S. Martin. 1980. Neo- tropical anachronisms: fruits the mastodons left behind. Bull. Ecol. Soc. Am. 61(2): 120. Jat, R.L., M.S. Dravid, O.K. Das, and N.N. Goswami. 1975. Effect of flooding and high soil water conditions on root porosity and growth of maize. J. Indian Soc. Soil Sci. 23 (3): 291-297. Johnson, R.I. 1970. The systematics and zoogeography of the unionidae (Mol- lusca: Bivalvia) of the southern Atlantic slope region. Bull. Mus. Comp. Zool. Harv. Univ. 140(6): 263-449. 116 Jones, R.H. 1981. A lowland forest com- munity classification in the northern coastal plain of South Carolina. Thesis. Clemson Univ., Clemson, S.C. Jordan, C.F., and R. Herrera. 1981. Trop- ical rain forest: are nutrients really critical? Am. Nat. 117(2): 167-180. Kaufman, M.I. 1969. Color of water in Florida streams and canals. Bur. of Geol. Map Ser. 35. Fla. Dep. of Nat. Resour. , Tallahassee. Kaufman, M.I. 1970. The pH of water in Florida streams and canals. Bur. of Geol. Map Ser. 37. Fla. Dep. of Nat. Resour., Tallahassee. Kaufman, M.I., and J.E. Dysart. 1978. Nitrogen, phosphorus, organic carbon, and biochemical oxygen demand in Florida surface waters, 1972. U.S. Geol. Survey. Water Resour. Invest. 78-43. Kaushik, N.K. 1975. Decomposition of allochthonous organic matter and sec- ondary production in stream ecosys- tems. Pages 90-95 vn Productivity of world ecosystems. Proceedings Symposium, Special Commission, Inter- national Biological Program 1972. National Academy of Sciences. 166 pp. Keller, E.A. 1977. The fluvial system: selected observations. Pages 36-46 in A. Sands, ed. Riparian forests in California: their ecology and conser- vation. Inst, of Ecol., Univ. of Calif., Davis. Publ. 15. Kennedy, H.E. 1970. Growth of newly planted water tupelo seedlings after flooding and siltation. For. Sci. 16: 250-256. Kennedy, R.S. 1977. Ecological analysis and population estimates of the birds of the Atchafalaya River Basin in Louisiana. Ph.D. Dissertation. La. State Univ., Baton Rouge. 200 pp. Kibby, H.V. 1979. Effects of wetlands on water quality. Pages 289-298 in Strategies for protection and manage- ment of floodplain wetlands and other riparian ecosystems. Publ. GTR- WO-12, U.S. For. Serv., Washington, D.C. Kilpatrick, R.A., and H.H. Barnes. 1964. Channel geometry of Piedmont streams as related to frequency of floods. U.S. Geol. Surv. Pap. 422-E. 10 pp. Kitchens, W.M., Jr., J.M. Dean, L.H. Stevenson, and J.H. Cooper. 1975. The Santee Swamp as a nutrient sink. Pages 349-366 in E.G. Howell, J.B. Gentry and M.H. Smith, eds. Mineral cycling in southeastern ecosystems. U.S. Energy Res. and Devel . Adm. Tech. Inf. Center. Kjerfve, Bjorn. 1976. The Santee-Cooper; a study of estuarine manipulations. Contrib. 127, Belle Baruch Inst, for Marine Biol, and Coastal Res., Univ. S.C, Columbia. Klawitter, R.A. 1962. Sweetgum, swamp tupelo and water tupelo sites in a South Carolina bottomland forest. Dissertation. Duke Univ., Durham, N.C. 176 pp. Knight, R. 1973. Plant communites. Pages 3-7 in North Carolina swamps. Unpubl. ms. Dep. of Environ. Sci. and Engi- neering, Univ. of N.C. at Chapel Hill. Konikoff, M. 1977. Studies of the life history and ecology of the red swamp crawfish, Procambarus clarkii, in the lower Atchafalaya Basin Floodway. Final rep. to U.S. Fish and Wildl. Serv. by Dep. Biol., Univ. Southwest. La. , Lafayette. 81 pp. Kuenzler, E.J., P.J. Mulholland, L.A. Ruley, and R.P. Sniffen. 1977. Water quality in North Carolina coastal plain streams and effects of channel- ization. Rep. 127. Water Resour. Res. Inst, of the Univ. of N.C, N.C State Univ., Raleigh. 160 pp. Kurz, H., and R.K. Godfrey. 1962. Trees of southern Florida. Omni Press, Sarasota, Fla. 311 pp. 117 Lambou, V.W. 1959. Fish populations of backwater lakes in Louisiana. Trans. Am. Fish. Soc. 88(1): 7-15. Lambou, V.W. 1962. Comments on proposed dam on Old River, Batchelor, La. La. Wild!, and Fish. Comm. 37 pp. Lambou, V.W. 1963. The commercial and sport fisheries of the Atchafalaya Basin Floodway. Proc. (17th) Annu. Conf. Southeast. Assoc. Game Fish Comm. 17: 256-381. Langdon, O.G., J. P. McClure, D.D. Hook, J.K. Crockett, and R. Hunt. 1S81. Extent, condition, management, and research needs of bottomland hard- wood-cypress forests in the South- east. Pages 71-85 _in J.K. Clark and J. Benforado, eds. Wetlands of bot- tomland hardwood forests. Proceedings of a workshop on bottomland hardwood forest wetlands of the Southeastern United States held at Lake Lanier, Georgia, June 1-5, 1980. Developments in Agricultural and Managed-forest Ecology, vol. 11. Elsevier Scientific Publ. Co., New York. Larson, J.S., M.S. Bedinger, C.F. Bryan, S. Brown, R.T. Huffman, E.L. Miller, D.G. Rhodes, and B.A. Touchet. 1981. Transition from wetlands to uplands in southeastern bottomland hardwood forests. Pages 225-273 j_n J.R. Clark and J. Benforado, eds. Wetlands of bottomland hardwood forests. Proceed- ings of a workshop on bottomland hardwood forest wetlands of the Southeastern United States held at Lake Lanier, Georgia, June 1-5, 1980. Developments in Agricultural and Man- aged-forest Ecology, vol. 11. Else- vier Scientific Publ. Co., New York. Leitman, H.M. 1978. Correlation of Apalachicola River floodplain tree communities with water levels, eleva- tion and soils. M.S. Thesis. Fla. State Univ., Tallahassee. 57 pp. Leitman, H.M., J.E. Sohm, and M.A. Frank- lin. 1981. Wetland hydrology and tree distribution of the Apalachi- cola River floodplain, Florida. U.S. Geol . Surv. 204 pp. Water Supply Pap. 21S6-A. Leopold, L.B., and W.B. Langbein. 1966. River meanders. Sci. Am. 214(16): 60-70. Leopold, L.B., and M.G. Wolman. 1957. River channel patterns: braided, meandering and straight. Pages 39-85 in Geological Survey Professional Paper 282-B. U.S. Govt. Printing Office, Washington, D.C. Leopold, L., M.G. Wolman, and J. Miller. 1964. Alluvial processes in geomor- phology. W.H. Greeman and Co., San Francisco, Calif. 522 pp. Lindsey, A. A., R.O. Petty, D.V. Sterling, and W. Van Asdall. 1961. Vegetation and environnient along the Wabash and Tippecanoe Rivers. Ecol. Monogr. 31(2): 105-156. Little, E.J., and J. A. Quick, Jr. 1976. Ecology, resource rehabilitation, and fungal parasitology of commercial oysters, Crassostrea virginica (Gme- lin), in Pensacola Estuary, Florida. Fla. Mar. Res. Publ. 21, Contrib. 280, Fla. Dep. Nat. Resour. Mar. Res. Lab., St. Petersburg. Little, E.L., Jr. 1979. Checklist of United States trees (native and natu- ralized). Agric. Handbook 541, Dep. of Agric. For. Serv. 375 pp. Livingston, R.J., G.J. Kobylinski, F.G. Lewis III, and P.F. Sheridan. 1975. Long-term fluctuations of the epiben- thic fish and invertebrate popula- tions in Apalachicola Bay, Florida. (Unpubl. ms.) Dep. Biol. Sci., Fla. State Univ., Tallahassee. Livingston, R.J., R.L. Iverson, and D.C. White. 1976. Energy relationships and the productivity of Apalachicola Bay. Final res. rep. to Fla. Sea Grant College. 437 pp. Lock, M.A., and H.B.N. Hynes. 1976. The fate of "dissolved" organic carbon derived from autumn-shed leaves (Acer 118 saccharum) in a temperate hardwater stream. Limnol. Oceanogr. 21(3): 436-443. Long, F.L., H.F. Perkins, J.R. Carreker, and J.M. Daniels. 1969. Morpho- logical, chemical, and physical characteristics of eighteen repre- sentative soils of the Atlantic coast flatwoods. Bull. 39. Agric. Res. Stn., USDA and Ga. Agric. Exp. Stns. 74 pp. Lowe, C.E. 1958. Ecology of the swamp rabbit in Georgia. J. Mammal. 39: 116-127. Lowrance, R.R. 1981. Nutrient cycling in an agricultural watershed: waterborne nutrient input/output budget for the riparian zone. Ph.D. Dissertation. Univ. of Ga. , Athens. Lunz, G.R., Jr. 1938. The effects of the flooding of the Santee River in April, 1936 on oysters in the Cape Roman area of South Carolina. Part II. U.S. Corps of Engineers, Charles- ton, S.C. 33 pp. Lush, D.L., and H.B.N. Hynes. 1973. The formation of particles in freshwater leachates of dead leaves. Limnol. Oceanogr. 18(6): 968-977. Lush, D.L., and H.B.N. Hynes. 1978. The uptake of dissolved organic matter by a small spring stream. Hydrobiologia 60(3): 271-275. Luxmoore, R.J., R.A. Fisher, and L.H. Stolzy. 1973. Flooding and soil temperature effects on wheat during grain filling. Agron. J. 65: 361-365. Lyell, Sr. Charles. 1849. A second visit to the United States of North Amer- ica. Vol. I. Harper and Bros., New York. 273 pp. Lynch, J.M., and J. Crawford. 1980. Re- connaissance survey of the lower Roa- noke River floodplain, N.C. Submitted to The Nature Conservancy and N.C. Natural Heritage Program. 4 June. 58 pp. MacDonald, P.O., W.E. Frayer, and J.K. Clauser. 1979. Documentation, chro- nology, and future projections of bottomland hardwood habitat loss in the lower Mississippi alluvial plain. Vol. I: basic report. U.S. Fish Wildl. Serv., Div. Ecol. Serv. , Washington, D.C. Maddox, G.E. 1972. Ecologic-economic values in phreatophyte control. Pages 257-259 in S.C. Csallany, T.G. McLaughlin and W.D. Striffler, eds. Watersheds in transition. Symp. Proc. Am. Water Resour. Assoc, Urbana, 111. Malcolm, R.L., and W.H. Durum. 1976. Organic carbon and nitrogen concen- trations and annual organic load of six selected rivers of the United States. U.S. Geol . Surv. Water Supply Pap. 1817-F. Martin, L.D. 1980. Late Pleistocene faunal distribution and community evolution. Bull. Ecol. Soc. Am. 61(2): 120. Martin, P.S., and B.E. Harrell. 1957. The Pleistocene history of temperate biota in Mexico and Eastern United States. Ecology 38(3): 468-480. Meade, R.H. 1976. Sediment problems in the Savannah River Basin. Pages 105- 129 in B.L. Dillman and J.M. Stepp, eds. The future of the Savannah River. Water Resour. Res. Inst., Clemson Univ., Clemson, S.C. Meade, R.H., and S.W. Trimble. 1974. Changes in sediment loads in rivers of the Atlantic drainage of the United States since 1900. Pages 99- 104 in Effects of man on the inter- face of the hydrological cycle with the physical environment. Symposium. Sept. 1974: Internatl. Assoc. Hydrol . Sci. Publ. 113. Meeter, D.A., R.J. Livingston, and G.C. Woodsum. 1979. Long-term climatologi- cal cycles and population changes in a river-dominated estuarine system. Pages 315-338 in R.J. Livingston, ed. 119 Ecological processes in coastal and marine systems. Plenum Press, New York. Miller, M. , B. Hartman, and D. Dunford. 1977. Fish and wildlife values of the Apalachicola River and flood- plain. Pages 122-129 in R.J. Livings- ton and E.A. Joyce, eds. Proceedings of the Conference on the Apalachicola Drainage System, 23-24 April 1976. Fla. Mar. Res. Publ. 26. Fla. Dep. Nat. Resour. Marine Res. Lab., St. Petersburg. Miller, R.W., J.E. Schindler, and J.J. Alberts. 1975. Mobilization of mer- cury from freshwater sediments by humic acid. Pages 445-451 j£ F.G. Howell, J.B. Gentry and M.H. Smith, eds. Mineral cycling in southeastern ecosystems. U.S. Energy Res. and Devel. Adm. Tech. Info. Center. Mitsch, W.J. 1979. Interactions between riparian swamp and a river in south- ern Illinois. Pages 63-72 i_n R.R. Johnson and J.F. McCormick, tech. coords. Strategies for protection and management of floodplain wetlands and other riparian ecosystems. Publ. GTR-WO-12, U.S. For. Serv., Washing- ton,D.C. Mitsch, W.J., and K.C. Ewel . 1979. Compar- ative biomass and growth of cypress in Florida wetlands. Am. Midi. Nat. 101(2): 417-426. Mitsch, W.J., C.L. Dorge, and J.R. Weim- hoff. 1977. Forested wetlands for water resource management in southern Illinois. Res. rep. 132. Illinois VJater Resour. Center, Univ. of 111., Urbana. Monk, CD. 1966. An ecological study of hardwood swamps in north-central Florida. Ecology 47: 649-654. Morisawa, M., and E. LaFlure. 1979. Hy- draulic geometry, stream equilibrium and urbanization. Pages 333-350 j£ D.D. Rhodes and G.P. Williams, eds. Adjustments of the fluvial system. Kendall-Hunt Publ. Co., Dubuque, Iowa. 372 pp. Mulholland, P.J. 1979. Organic carbon cycling in a swampstream ecosystem and export by streams in eastern North Carolina. Ph.D. Dissertation. Univ. of N.C. at Chapel Hill. 152 pp. Mulholland, P.J., and E.J, Kuenzler. 1979. Organic carbon export from upland and forested wetland watersheds. Limnol. Oceanogr. 24(5): 960-966. Muller, R.A., and T.M. Oberlander. 1978. Physical geography today - a portrait of a planet. 2nd ed. Random House, New York. 590 pp. Mycielska-Dowgiallo, E. 1977. Channel pattern changes during the last glaciation and Holocene, in the northern part of the Sandomierz Basin and the middle part of the Vistula Valley, Poland. Pages 75-87 jji K.J. Gregory, ed. River channel changes. John Wiley and Sons, New York. 448 pp. Nelson, D.J., S.V. Kaye, and R.S. Booth. 1972. Radionuclides in river systems. Pages 367-387 in R.T. Oglesby, C.A. Carlson and J. A. McCann, eds. River ecology and man. Academic Press, New York. Nuritdinov, N., and B.B. Vartapetyan. 1976. Transport of oxygen from the overground parts into roots of cot- ton. Fiziol. Rast. 23(3): 622-624. Odum, E.P. 1969. The strategy of ecosystem development. Science 164: 262-270. Odum, E.P. 1977. The life support value of forests. Pages 101-105 jji Forests for people. Proc. Soc. Am. For., Soc. Am. Foresters, Washington, D.C. Odum, E.P. 1978. The value of wetlands: a hierarchical approach. Pages 16-25 in P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wetland functions and values: the state of our understand- ing. Am. Water Resour. Assoc, Minn- eapolis, Minn, Odum, E.P., J.B. Birch, and J.L. Cooley. 1981a. Lower Savannah River fresh- water wetlands: III. Comparisons of 120 giant cutgrass productivity in tidal and impounded marshes with special reference to waste assimilation and tidal subsidy. (Unpubl. ms.) Inst, of Ecology, Univ. of Ga., Athens. Odum, E.P. , J.E. Pinder III, T.A. Chris- tiansen. 1981b. Nutrient losses from sandy soils during old-field succession. (Unpubl. ms.) Inst, of Ecology, Univ. of Ga., Athens. Costing, Henry J. 1942. An ecological analysis of the plant communities of Piedmont, North Carolina. Am. Midi. Nat. 28: 1-126. Paerl, H.W. 1974. Bacterial uptake of dissolved organic matter in relation to detrital aggregation in marine and freshwater systems. Limno. Oceanogr. 19(6): 966-972. Parsons, K. , and C.H. Wharton. 1978. Macroinvertebrates of pools on a Piedmont river floodplain. Ga. J. Sci. 36: 25-33. Patrick, R. 1972. A commentary on "What is a River." Pages 67-74 in R.T. Oglesby, C.A. Carlson and J. A. McCann, eds. River ecology and man. Academic Press, New York. 465 pp. Patrick, P., J. Cairns, Jr., and S.S. Roback. 1967. An ecosystematic study of the fauna and flora of the Savannah River. Proc. Acad. Nat. Sci. Phila. 118 (5): 109-407. Patrick, W.H., Jr., and D.S. Mikkelsen. 1971. Plant nutrient behavior in flooded soil. Pages 187-215 jji Fer- tilizer technology and use. 2nd ed. Soil Sci. Soc. of Am., Inc., Madison, Wise. Penfound, W.T. 1952. Southern swamps and marshes. Bot. Rev. 18: 413-446. Perkins, H.F., C.B. England, and J. A. Gibbs. 1962. Some morphological, physical, chemical and clay mineral characteristics of several agricul- turally important Georgia Soils. Tech. Bull. N.S. 26. Ga. Agric. Exp. Stns., Univ. of Ga. College of Agric. , Athens. 37 pp. Perkins, H.F., R.A. McCreery, G. Lickaby, and C.E. Perry. 1979. Soils of the southeast Geogia branch experiment stations. Res. Bull. 245, Univ. of Ga. College of Agric. Exp. Stns., Athens. 51 pp. Peters, D.S., D.W. Ahrenholz, and T.R. Rice. 197&. Harvest and value of wetland associated fish and shell- fish. Pages 606-617 in P.E. Greeson, J.R. Clark and J.E. Clark, eds. Wet- land functions and values: the state of our understanding. Am. Water. Resour. Assoc, Minneapolis, Minn. Petersen, R.C., and K.W. Cummins. 1974. Leaf processing in a woodland stream. Freshw. Biol. 4(4): 343-368. Phung, H.T., and E.B. Knipling. 1976. Photosynthesis and transpiration of citrus seedlings under flooded condi- tions. Hortic. Sci. 11(2): 131-133. Pinder, J.E., and M.H. Smith. 1975. Fre- quency distributions of radiocesium concentrations in soil and biota. Pages 107-125 vn F.G. Howell, J.B. Gentry and M.H. Smith, eds. Mineral cycling in southeastern ecosystems. ERDA Symp. Ser. Conf. 740513. Ponnamperuma, F.N. 1972. The chemistry of submerged soils. Adv. Agron. 24: 29-96. Post, H.A., and A. A. de la Cruz. 1977. Litterfall, litter composition, and flux of particulate organic material in a Coastal Plain stream. Hydrobio- logia 55(3): 201-208. Potter, G. 1974. The population dynamics, bioenergetics and nutrient cycling of three species of mice in the Hubbard Brook Experimental Forest, New Hamp- shire, U.S.A. Ph.D. Dissertation. Dartmouth College, Hanover, N.H. 162 pp. Putnam, J. A. 1951. Management of bottom- land hardwoods. U.S. Dep. Agric. For. Serv. Southern For. Exp. Stn., Occas. Pap. 116. 51 pp. Putnam, J. A., G.M. Furnival, and J.S. McKnight. 1960. Management and 121 inventory of southern hardwoods. U.S. Dep. Agric. For. Serv. Agric. Hand- book 181. 102 pp. Radford, A.E., D.K.S. Otte, L.J. Otte, J.R. Kassey, and P.D. Whitson. 1980. Natural heritage: classification, inventory, and information. A.E. Radford, Dep. of Botany, Univ. of N.C. at Chapel Hill. 674 pp. Rahmatullah, F., M. Chaudry, and A. Rashid. 1976. Micronutrient avail- ability to cereals from calcareous soils. II. Effects of flooding on electrochemical properties of soils. Plant Soil 45: 411-420. Reddy, K.R., W.H. Patrick, Jr., and R.E. Phillips. 1976. Ammonium diffusion as a factor in nitrogen loss from flooded soils. J. Soil Sci. Soc. Am. 40(4): 528-533. Reuter, J.H., and E.M. Perdue. 1977. Importance of heavy metal -organic matter decomposition in natural waters. Geochim. Cosmochim. Acta 41: 325-334. Salisbury, F.B., and Plant physiology. Publ. Co., Inc. 436 pp. C.W. Ross. 1978. 2nd ed. Wadsworth , Belmont, Calif. Saucier, R.T., and A.R. Fleetwood. 1970. Origin and chronologic significance of late Quaternary terraces, Ouachita River, Arkansas and Louisiana. Geol. Soc. Am. Bull. 81: 869-890. Schlesinger, W.H. 1978. Community struc- ture, dynamics and nutrient cycling in the Okefenokee cypress swamp forest. Ecol. Monogr. 48: 43-65. Schumm, S.A. 1965. River metamorphosis. J. Hydraulics Div., Proc. Am. Soc. Civil Engineers Pap. 6352-95 (HYI): 255-273. Schumm, S.A. 1971. Fluvial geomorphology: the historical perspective. Chapter 4 jji R.H.W. Shen, ed. River mechan- ics. 2 vols. Based on lecture notes delivered at Institute on River Mechanics, Colo. State Univ. June 1970. Hsieh Wen Shen, Publ., Fort Collins, Colo. Reynolds, P.E., and W.R. Parrott, Jr. 1980. Hydrologic, soils and vegeta- tion measurements for a southern New Jersey hardwood swamp. U}^ G.S. Cox, ed. Proceedings Central Hardwood Con- ference III. Univ. of Mo., Columbia. Richardson, C.J., D.L. Tilton, J. A. Kadler, J. P. Chamie, and W.A. Wentaz. 1978. Nutrient dynamics of northern wetland ecosystems. Pages 217-241 jm R.E. Good, D.F. Whigham and R.L. Simpson, eds. Freshwater wetlands: ecological processes and management potential. Academic Press, New York. 378 pp. Rochow, J.J. 1974, Estimate of above- ground biomass and primary production in a Missouri forest. J. Ecol. 62: 567-577. Rosenau, J.C, G.L. Faulkner, C.W. Hendry, Jr., and R.W. Hull. 1977. Springs of Florida. Bull. 31 (Revised). Fla. Dep. of Nat. Resour., Tallahassee. 461 pp. Sepers, A.B.J. 1977. The utilization of dissolved organic compounds in aquatic environments. Hydrobiologia 52(1): 33-38. Shaw, S.P., and C.G. Fredine, 1956. Wet- lands of the United States. U.S. Fish Wildl. Serv., Circ. 39. 67 pp. Shelford, V.E. 1954. Some lower Missis- sippi valley biotic communities: their age and elevation. Ecology 35(2): 126-142. Shugart, H.H., D.E. Reichle, N.T. Edwards, and J.R. Kercher. 1576. A model of calcium cycling in an east Tennessee Liriodendron forest: model structure. 122 parameters, and frequency response analysis. Ecology 57: 99-109. Shure, D.J., and M.R. Gottschalk. 1976. Cesium-137 dynamics within a reactor effluent stream in South Carolina. Pages 234-241 jn C.E. Gushing, Jr., ed. Radioecology and energy re- sources. Spec. publ. 1, Ecol. Soc. Am. Dowdin, Hutchinson and Ross, Inc., Stroudsbury, Pa. Sigafoos, R. 1964. Botanical evidence of floods and flood-plain depositions. U.S. Geol. Surv. Prof. Pap. 485-A. 35 pp. Siolo, H. 1975. Amazon tributaries and drainage basins. Pages 199-213 jn^ A.D. Hasler, ed. Coupling of land and water systems. Springer-Verlag, New York. Sklar, F.H., and W.H. Conner. 1979. Ef- fects of altered hydrology on primary production and aquatic animal popula- tions in a Louisiana swamp forest. Pages 191-208 in J.W. Day, Jr., D.D. Gulley, Jr., R.E. Turner and A.T. Humphrey, Jr., eds. Proceedings Third Coastal Marsh and Estuary Man- agement Symposium. La. State Univ. Div. Continuing Ed., Baton Rouge. Slater, J.V. 1954. The quantitative eval- uation of dissolved organic matter in natural waters. Trans. Am. Micro. Soc. 73(4): 416-423. Sniffen, R.P. 1980. Distribution and abundance of invertebrates in a sea- sonally inundated swamp in North Carolina coastal plain [tentative title]. Ph.D. Dissertation. Univ. of N.C. at Chapel Hill. (In prepara- tion. ) Sniffen, R.P., L.A. Yarbro, and J.G. Smith. 1981. Bryophyte distribution and abundance in relationship to inundation frequency in a floodplain swamp. (Unpubl. draft.) Inst, for Coastal and Mar. Resour., E. Carolina Univ., Greenville. 15 pp. Society of American Foresters. 1967. Forest cover types of North America (exclusive of Mexico). Soc. For., Washington, D.C. 67 pp. An. Soil Conservation Service. 1975. Soil Taxonomy. Agric. Handbook 436. S.C.S. U.S. Dep. of Agric. Soil Conservation Service. 1977. Land treatment plan for erosion control and water quality improvement in the Obion-forked Deer Basin River. U.S.D.A., S.C.S. , 675 Courthouse Square, Nashville, Tenn. Soileau, L.D., K.C. Smith, R. Hunter, C.E. Knight, D.M. Soileau, W.E. Shell, Jr., and D.W. Hayne. 1975. Atchafa- laya Basin usage study. 1975; final report. Available from: U.S. Army Corps of Engineers, New Orleans, La. 85 pp. Staheli, A.C., D.E. Ogren, and C.H. Wharton. 1974. Age of swamps in the Alcovy River drainage basin. South- east Geol. 16(2): 103-106. Stankovic, V.S., and D. Jankovic. 1971. Mechanismus der fisch-producktion im gebiet des mittleren Donaulaufes. Arch. Hydrobiol. Suppl. 36: 299-305. Strahler, A.N. 1956. The nature of induced erosion and aggradation. Pages 621-638 in W.L. Thomas, Jr. ed. Man's role in changing the face of the earth. Univ. of Chicago Press, Chicago, 111. Stringfield, V.T., and H.E. LeGrand. 1966. Hydrology of limestone terranes in the Coastal Plain of Southeastern United States. Geol. Soc. Am. Spec. Pap. 93. New York. 46 pp. Stubbs, J. 1973. Atlantic oak-gum-cypress. Pages 89-93 jui Silvicultural systems for the major forest types of the United States. U.S.D.A. For. Serv. Agric. Handbook. 445 pp. Tanner, J. 1975. Ecological comparisons: Congaree and other bottomlands. Pages 104-107 in Congaree Swamp: greatest unprotected forest on the continent. S.C. Environ. Coalition, Columbia, S.C. 123 Teskey, R.C., and T.ri. Hinckley. 1S77. Impact of water level changes on woody riparian and wetland communi- ties. Vol. II: Southern forest region. U.S. Fish Wildl. Serv. Biol. Serv. Program FWS/OBS-77/59. Thorn, R.G. 1967. Coastal and fluvial landforms: Horry and Karion Counties, South Carolina. Tech. Rep. 44, Coastal Studies Inst., La. State Univ. , Baton Rouge. Thomas, J.D. 1975. Gammarid amphipods of the Barataria Bay, Louisiana Region. M.S. Thesis. La. State Univ., Baton Rouge. Thomas, J. P. 1966. Influence of the Altaniaha River on primary production beyond the mouth of the river. M.S. Thesis. Univ. of Ga., Athens. Thompson, M.T., and R.F. Carter. 1955. Surface water resources of Georgia during droughts of 1954. Ga. Dep. Mines, Mining and Geol. Inf. Circ. 17. Thorne, C.R., and J. Lewin. 1979. Bank processes, bed material movement and planform development in a meandering river. Pages 117-137 j£ D.D. Rhodes and G.P. Williams, eds. Adjustments of the fluvial system. Kendall-Hunt Publ. Co., Dubuque, Iowa. 372 pp. Thornes, J.B. 1977. Hydraulic geometry and channel change. Pages 91-100 jji K.J. Gregory, ed. River channel changes. John Wiley and Sons, New York. 448 pp. Tjnkle, D.W. 1959. Observations of rep- tiles and amphibians in a Louisiana swamp. Am. Midi. Nat. 62(1): 189-205. Todd, R.L., R.A. Leonard, and L.E. Ramus- sen. 1981. Effects of land use on nutrient cycling in agricultural watershed ecosystems. (Res. Proposal to NSF). Dep. of Agron., Univ. of Ga. , Athens. 96 pp. Trimble, S.W. 1970. The Alcovy River swamps: the result of culturally accelerated sedimentation. Bull. Ga. Acad. Sci. 28(4): 131-144. Trimble, S.U. 1979. Denudation studies: can we assume stream steady state? Science 88(4194): 1207-1208. Turner, R.E., S.W. Forsythe, and N.J. Craig. 1981. Bottomland hardwood forest land resources of the South- eastern United States. Pages 13-28 in J.R. Clark and J. Benforado, eds. Wetlands of bottomland hardwood forests. Proceedings of a workshop on bottomland hardwood forest wet- lands of the Southeastern United States held at Lake Lanier, Georgia, June 1-5, 1980. Developments in Agricultural and Managed-forest Ecol- ogy vol. 11. Elsevier Scientific Publ. Co., New York. U.S. Army Corps of Engineers. 1935. Altamaha, Oconee and Ocmulgee Rivers. Ga. House Doc. 68, 74th Congr. 1st Sess., U.S. Govt. Printing Office. U.S. Department of the Interior. 1978. The Atchafalaya, America's greatest river swamp. Presented to the Atchaf- alaya Basin Agency Manage. Group. U.S. Fish Wildl. Serv., Lafayette, La. U.S. Geologic Survey (USGS). 1977. Water data for Georgia. Water data rep. GA-77-1. U.S. Geol. Surv., Dep. of Interior, Washington, D.C. Vannote, R.L., G.W. Minshall, K.W. Cum- mins, J.R.D. Sedell, and C.E. Gush- ing. 1980. The river continuum concept. Can. J. Fish. Aquat. Sci. 37(1): 130-137. Veith, G.D., D.W. Kuehl, E.N. Leonard, F.A. Puglisi: and A.E. Lemke. 1979. Polychlorinated biphenyls and other organic chemical residues in fish from major watersheds of the United States 1976. Pestic. Monit. J. 13(1):1-11. Vester, G. 1£72. The physiology of flood tolerances in trees. Bot. Soc. Edinb. Trans. 41(4): 556-557. Walker, M. 1980. Utilization by fishes of a blackwater creek floodplain in North Carolina [tentative title]. M.S. Thesis. E. Carolina Univ., Greenville. (In preparation). 124 Walker, R.L. 1981. Distribution of oys- ter drills (Urosalpinx cinerea (Say)), in Wassau Sound, Georgia. Ga. J. Sci. 39: 127-139. Wallace, J.B., J.R. Webster, and W.R. Woodall. 1977. The role of filter feeders in flowing waters. Arch. Hydrobiol. 79: 506-532. Watts, W.A. 1971. Post-glacial and inter- glacial vegetation history of south- ern Georgia and central Florida. Ecology 52(4): 676-690. Wells, B.W. 1928. Plant communities of the coastal plain of North Carolina and their successional relations. Ecology 9:230-242. Wells, E.F. 1970. A vascular flora of the Uwharrie Wildlife Management Area Montgomery County, North Carolina. M.A. Thesis. Univ. of N.C. at Chapel Hill. 85 pp. Wharton, C.H. 1970. The southern river swamp, a multiple use environment. Off. of Res. and Serv. , Ga. State Univ. , Atlanta. 48 pp. Wharton, C.H. 1977. The natural environ- ments of Georgia. Georgia Geol. Surv., Dep. of Nat. Resour., Agric. Bldg., Martin L. King Blvd., Atlanta. [Editor's note: published under same title in 1978 by the Geologic and Water Resources Division and the Resource Planning Section, Office of Planning and Research of the Georgia Department of Natural Resources, Atlanta, Ga.] Wharton, C.H. 1980. Values and functions of bottomland hardwoods. Pages 341- 353 j_n Transactions 45th North Ameri- can Wildlife and Natural Resources Conference. Wildlife Management Institute, Washington, D.C. Wharton, C.H., and M.M. Brinson. 1979a. Characteristics of southeastern river systems. Pages 32-40 j_n R.R. Johnson and J.F. McCormick, tech. coords. Strategies for protection and manage- ment of floodplain wetlands and other riparian ecosystems. Publ . GTR-WO-12, U.S. For. Serv., Washington, D.C. Wharton, C.H., and M.M. Brinson. 1979b. Unpublished report to Environmental Protection Agency on the Tallahala Creek Reservoir Project, Jasper County, Mississippi. EPA Region IV, Atlanta. 87 pp. Wharton, C.H., and H.L. Ragsdale. 1979. The values of unmanaged national forests in the southern Appalachians. Unpublished report to Committee on RARE II, The Georgia Conservancy, Atlanta. 68 pp. Wharton, C.H., H.T. Cdum, E. Ewel, M, Duever, A. Lugo, R. Boyt, J. Bar- tholomew, E. Debellevue, S. Brown, M. Brown, and L. Duever. 1977. For- ested wetlands of Florida - their management and use. Div. of State Plan., State of Fla., Tallahassee. 347 pp. Wharton, C.H., V.W. Lambou, J. Newsom, P.V. Winger, L.L. Gaddy, and R. Mancke. 1981. The fauna of bottomland hardwoods in Southeastern United States. Pages 87-160 jn J.R. Clark and J. Benforado, eds. Wetlands of bottomland hardwood forests. Proceed- ings of a workshop on bottomland hardwood forest wetlands of the Southeastern United States held at Lake Lanier, Georgia, June 1-5, 1980. Developments in Agricultural and Managed-forest Ecology, vol. 11. Elsevier Scientific Publishing Co., New York. Whigham, B.F., and S.E. Bayley. 1978. Nutrient dynamics in fresh water wetlands. Pages 468-478 jin P.E. Creeson, J.R. Clark and J.E. Clark, eds. Wetlands functions and values: the state of our understanding. Am. Water Resour. Assoc, Minneapolis, Minn. White, D.C, R.J. Livingston, R.J. Bobbie, and J.S. Nickels. 1979. Effects of surface composition, water column chemistry and time of exposure on the composition of the microflora and associated macrofauna in Apalachicola Bay, Florida. Pages 83-116 vn R.J. Livingston, ed. Ecological processes in coastal and marine systems. Plenum Press, New York. 125 Whitehead, D.R., and E.S. Barghoorn. 1962. Pollen analytical investigations of Pleistocene deposits from western North Carolina and South Carolina. Ecol. Monogr. 32: 347-369. Whitlow, T.H., and R.W. Harris. 1979. Flood tolerance in plants: a state- of-the-art review. U.S. Army Engi- neers Waterways Experiment Stn. Tech. rep. E-79-2. Vicksburg, Miss. 161 pp. Whittaker, PwH., and P.L. Marks. 1975. Methods of assessing terrestrial pro- ductivity. Pages 55-118 vn H. Lieth and R.H. Whittaker, eds. Primary pro- ductivity of the biosphere. Springer- Verlag, New York. Windom, H.L., W.M. Dunstan, and W.S. Gard- ner. 1975. River input of inorganic phosphorus and nitrogen to the south- eastern salt-marsh estuarine environ- ment. Pages 309-312 jn E.G. Howell, J.B. Gentry, and M.H. Smith, eds. Mineral cycling in southeastern eco- systems. U.S. Energy Res. and Devel. Adm. , Tech. Inf. Center. Available from: NTIS, Springfield, Va. Winger, P.V. 1981. Physical and chemical characteristics of warmwater streams: a review. Pages 32-44 jr[ Warmwater Streams Symposium. Am. Fish. Soc. Winner, M.D., Jr., and C.E. Simmons. 1977. Hydrology of the Creeping Swamp watershed. North Carolina with refer- ence to potential effects of stream channelization. U.S. Geol. Surv. Water Resour. Invest. 77-26. 54 pp. Wolman, M.G., and L. Leopold. 1957. River floodplains: some observations on their formation. U.S. Geol. Surv. Pages 87- Prof. Pap. 109 in 282-C. Wolman, M.G., and J. P. Miller. 1960. Mag- nitude and frequency of forces in geomorphic processes. J. Geol. 68(1): 54-74. Woodall, W.R., J.G. Adams, and J. Heise. 1975. Invertebrates eaten by Altamaha River fish. Presentation; 39th Meet. Ga. Entomol. Soc, St. Simons Island, March 19-21. Woodwell , G.M. 1958. Factors controlling growth of pond pine seedlings in organic soils of the Carol inas. Ecol. Monogr. 38: 219-236. Yang, C.T., and CCS. Song. 1979. Dynamic adjustments of alluvial channels. Pages 55-67 j_n D.D. Rhodes and G.P. Williams, eds. Adjustments of the fluvial system. Kendall-Hunt Publ. Co., Dubuque, Iowa. 372 pp. Yarbro, L.A. 1979. Phosphorus cycling in the Creeping Swamp floodplain ecosys- tem and exports from the Creeping Swamp watershed. Ph.D. Dissertation. Univ. of N.C at Chapel Hill. 231 pp. Young, S.A. , W.P. Kovalek, and K.A. Del Signore. 1978. Distances travelled by autuHin-shed leaves introduced into a woodland stream. Am. Midi. Nat. 100(1): 217-222. Ziser, S.W. 1978. Seasonal variations in water chemistry and diversity of the phytophilic macroinvertebrates of three swamp communities in southeast- ern Louisiana. Southwest. Nat. 23 (4): 545-562. 126 APPENDIX SOILS OF SOUTHEASTERN FLOODPLAINS Tables A-1 through A-8 describe soil characteristics in sampled dominance types and river floodplain classes within the study area. Soil samples were collected non- randomly from a depth of 8 to 30 cm (3 to 12 inches) below surface litter zones, in the center of the most mature group of trees. No samples were taken from atypi- cal microtopographic relief features at the site, nor were samples taken where there was evidence of logging, vehicle passage, scour channels, or upland erosion sources. Mechanical analysis of percent clay, silt, and sand was by the Bouyoucos method. Samples with very high organic matter were subjected to hydrogen peroxide or ashed at 500°C for 4 hours prior to analysis. Organic matter was determined by the Walkley-Black method. Depending on the amounts of silt and clay present, organic matter may be overestimated. The error of overestimation due to water driven off from clay and silt was computed on the basis of 8% error with 5% clay-silt and 30% error with 85% clay-silt (Broadbent 1953; Klawitter 1962); corrected percent organic matter appears in parentheses in Tables A-1 through A-8. Macronutrient concentrations were determined by plasma emission spec- trometry, following extraction by the double-acid method. 127 o -a c . o (/I Q. s- !r 01 O > O S- r— 1 — o TO on o CD o •a a. •o o o o o o ai > o CO o o •a- CVJ 00 ca lO CSJ KO r>- ■a- en «a- ,-1 C~sJ Ol yD CSJ en cri r^ *i3 >=3- CO 00 ^i3 «3- CNJ ^ ^ CO en ■=3- ■—1 en O 00 <— t ro en o 00 ^3- Od O en ID 1^ o o ■a- CSJ in "a- CO CM O CM ^^ 00 r^ >a- t— < C\J »— I t— 1 00 o «3- CO o o f~-t 00 o CSJ 00 o UD "=j- ro tr. f^ en CO 1 — 1 en 00 ■=)• CSJ CSJ CSI r^ LO CSJ lO o CSJ CSJ ro CSI 00 en t— 1 1^ CO ■a- 1 — I a> ro IC — TO fO 00 o CM ^ ID CO ro CM en CO C3 •-< LO LO ID en >a- CO Q. :i. .— 0) E 1— Q. Q. U s_ •—■—>> OJ >> Q. E 0) 5 < — ly) 0) OJ , — Ol OJ >, r— S- C2. Ol +-> Q. U OJ OJ 3 CL ID = 1 CL +-" ■*-) 3 s +-> o 3 03 -*-> 1 — +-> 2 CL O. 0) E CL E Ci s- fO E aJ t3 3 ■*-> oJ en 2 j= tn tj OJ uo 01 2 Q O) ID -C ID •a O) aj 2 2 QJ Q- Ol i- CL >1 "O 1 — S- U 0) 01 Ol 1 X CL +-• t ■ ( — 3 S- Ol s en fO ■♦-> ID ID ID o 2 O o o u o -C ^ -C u 0) u (-) ID Ol ID 1 — 1 — ID C fD ID CL ID O. a. 1 O) X a. ^ •.- ID -C 0) OJ c 2 Ol Ol 3 t/1 (O sz f 0) -!-> u u (D 3 u Ol OJ +-> 1 — o 0) OJ ID ID .ir en Ol Q > i_) o o ID X3 o OJ Q. Ol OJ ID E 0) i- Ol o OJ o XJ +-> ai c ■!-> OJ o s- Ol ID (_ iX s_ o c u ■1 — c 128 0) o II T3 C to o II tn <-J -!-> c > 0) l/l c I. o Ol Q. a. o c o T3 a. T3 o o o o oo or Q. o x> T3 O O Ol o o "3- o "3- 00 CO o o U3 O •3- CO OJ >* «3 CO 00 LO lO US >* ^ CO o sf ^ VD C\J VD o o 00 cr> UD O^ LO *:3- LD r^ o ^o O^r-^r^co'^*^'— I"— 'CO CO ou .— I CSJ CM CO -— < oj LD n ^:1- C\J CM -Q -Q CO U3 CM U3 en CO o Ol Q. <: o E >> to JD ro , 0) 1— o c- O a. o 1 3 r^ o +-) OJ Ol o 1 — 1- o >1 3 3 na .c E IB ID S Ol en Ol a. o. 3 +-> 2 o •f- -^ Ol 1— .— Q. lO 3 3 IT3 E >> 3 cn o O a; Q. 3 CL 3 ■(-> CL E O) 3 ••- cn CO 01 0) U Ol S- s. o ■ u lO ■l-J .c I— IT3 j:^ 129 OJ o ^ CM ^3- <* O 00 o ^H .-H CNJ E O >-i o *~4 II oie: c • o o c >1 •r- (D > I/) O ■•^ OJ OJ fO OJ C ■M d; r— cu O S- O) OJ o c . u c •n- 1 s 3 lA ei (O o ro fO X3 .£Z O >i ro JQ h- CO 00 ■— I— HI 130 CD CO «3- 1.0 CO en en s_ o ro 03 O ro oj CM .— 1 — 1 CM ^ CO CTt «^ 3 O LO CSJ •-H CO CO 0 CNJ <.-> U3 ro "* "d- oi CO ^- CO LT) .— I t— I CM ^H CO > 3 o CO o '— Q.00 O C/1 .— ( I— 4 n CM -^ CM .—I O U^ r-H CM en o o ■a o o > •o o o > OJ X OJ 0) 3 S O ■— ■— Q OJ (O Q> -^ s- 0 ■L- 5 n3 3 1 — TO CJ fll OJ 0 QJ 01 CD OJ CJ n> 0 0 Dl ;j^ t31 en en C3 0 CD 0 0 0 0 0 0 OJ c cz =3 0 LO u OJ in >> s_ 131 01 s- o II >> u II •- O O M- -r- O 13 S ■— O 13 O > — 1 ^ (U -^ o s- (IJ i- 3 o Ul 3 1 — 03 03 U fU 0) 03 03 r— s- x> Q O -^ 3 f— OJ 13 a; 03 03 OJ 03 03 03 l- 5 u. a. D- 03 CJ 13 j:: e c 03 03 u >, u -:^ o .^ o u ■o +-> +-> c: S- 13 1 — c -1 — ■ ,— to 3 3 CO »— » _1 1 t_) t— o o 03 14- 0) a. >> 03 Ql o o CM o CM 4-> 1/1 c 03 c O I Q.C/3 E O u a; o (/I 1 s_ o U3 II ^Z U 1— o o CO o LO r^ ^ cr> lo CO LT) CO ur> CTi in CO 00 o s- 03 Q. CL ■a: 03 03 03 u 1 03 JZ >> > 13 o c ^ JO u ■M 13 o 03 03 s- 03 ■♦-> 03 s •f— h- 1/1 VI a. Q. ■a- U3 LO •— t t— I O Q. T3 O O -!-> Ol =3 c C •J- Ul 0) OJ .c o o 3 o CO Ol 0) 3 c ro Ol ■ — coastal plain streams, fauna, ecological zones, couplings, anaerobic gradient, nutrient cycles, flood stress c. COSATI Field/Group 18. Availability Statement Unlimited 19. Security Class (This Report) Unclassified 21. No o( Pages 133 + xii 20. Security Class (This Page) Unclassified 22. Price (See ANSI-Z39.18) See Instructions on Reverse OPTIONAL FORM 272 (4-77) (Formerly NTIS-35) Department of Commerce U.S. GOVERNMENT PRINTING OFFICE: 1982—570-800 (D