Ecology and Management of a Forested Landscape: Fifty Years on the Savannah River Site

183218_Cloth 5/25/05 6:10 PM Page 1 ... FORESTS / ECOLOGICAL RESTORATION Advance praise for Ecology and Management...

Author: John Kilgo | John I. Blake | H. Ronald Pulliam

18 downloads 573 Views 15MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

183218_Cloth

5/25/05

6:10 PM

Page 1

...

FORESTS / ECOLOGICAL RESTORATION

Advance praise for Ecology and Management of a Forested Landscape

“Ecology and Management of a Forested Landscape is a unique chronicle of the successful ecological rehabilitation and restoration of a degraded, formerly agriculture-dominated system, starting with research and moving through adaptive natural resource management. With a case-study approach containing applications and concepts extending beyond the southeastern United States, this book is invaluable to all ecologists—from the academic to the practicing land manager.” —W. Mark Ford, research wildlife biologist, USDA Forest Service, Northeastern Research Station, West Virginia “The Savannah River Site is a priceless model of ecological recovery and restoration. It provides hard evidence of how a mutually beneficial relationship between humankind and natural systems might develop. This book’s clearly stated goals and objectives are admirably supported by data that cover large temporal and spatial spans.” —John Cairns Jr., University Distinguished Professor of Environmental Biology Emeritus, Virginia Polytechnic Institute and State University JOHN C. KILGO is research wildlife biologist, USDA Forest Service, Southern Research Station, Center for Forested Wetlands Research. JOHN I. BLAKE is assistant manager of the research program with the USDA Forest Service, Savannah River.

Washington • Covelo • London www.islandpress.org

Ecology and Management of a Forested Landscape

“The history of ecological research at the Savannah River Site is testimony to the power of long-term studies, interdisciplinary collaboration, and the application of basic science to land management challenges. This volume wonderfully documents that history and provides a comprehensive review of our current understanding of the dynamics and functioning of this diverse landscape.” —Norman L. Christensen Jr., professor of ecology and founding dean, Nicholas School of the Environment and Earth Sciences, Duke University, North Carolina

KILGO BLAKE

Ecology and Management of a Forested Landscape Fifty Years on the Savannah River Site

Edited by John C. Kilgo and John I. Blake

All Island Press books are printed on recycled, acid-free paper. Cover design: Amy Stirnkorb Cover photo: John Kilgo

Foreword by H. Ronald Pulliam

About Island Press Island Press is the only nonprofit organization in the United States whose principal purpose is the publication of books on environmental issues and natural resource management. We provide solutions-oriented information to professionals, public officials, business and community leaders, and concerned citizens who are shaping responses to environmental problems. In 2005, Island Press celebrates its twenty-first anniversary as the leading provider of timely and practical books that take a multidisciplinary approach to critical environmental concerns. Our growing list of titles reflects our commitment to bringing the best of an expanding body of literature to the environmental community throughout North America and the world. Support for Island Press is provided by the Agua Fund, The Geraldine R. Dodge Foundation, Doris Duke Charitable Foundation, Ford Foundation, The George Gund Foundation, The William and Flora Hewlett Foundation, Kendeda Sustainability Fund of the Tides Foundation, The Henry Luce Foundation, The John D. and Catherine T. MacArthur Foundation, The Andrew W. Mellon Foundation, The Curtis and Edith Munson Foundation, The New-Land Foundation, The New York Community Trust, Oak Foundation, The Overbrook Foundation, The David and Lucile Packard Foundation, The Winslow Foundation, and other generous donors. The opinions expressed in this book are those of the author(s) and do not necessarily reflect the views of these foundations.



r

Fifty Years on the Savannah River Site

Edited by John C. Kilgo and John I. Blake

r Foreword by H. Ronald Pulliam

Washington • Covelo • London

ip.kilgo.cx.i-400_436-482

6/23/05

2:20 PM

Page vi

Copyright (c) 2005 Island Press All rights reserved under International and Pan-American Copyright Conventions. No part of this book may be reproduced in any form or by any means without permission in writing from the publisher: Island Press, 1718 Connecticut Ave., Suite 300, NW, Washington, DC 20009. ISLAND PRESS is a trademark of The Center for Resource Economics. Copyright is claimed in the work of I. Lehr Brisbin Jr., Kurt A. Buhlmann, William D. Carlisle, Michael B. Caudell, Brent J. Danielson, J. Whitfield Gibbons, Judith L. Greene, Nick M. Haddad, Charles H. Hunter Jr., Paul. E. Johns, Robert A. Kennamer, Yale Leiden, Barton C. Marcy Jr., John J. Mayer, Tony M. Mills, William F. Moore, Eric A. Nelson, Sean Poppy, Travis J. Ryan, David E. Scott, Barbara E. Taylor, Tracey D. Tuberville, Lynn D. Wike, Christopher T. Winne, in the foreword, and the index to the Island Press edition. In accordance with Federal law and U.S. Department of Agriculture policy, this institution is prohibited from discriminating on the basis of race, color, national origin, sex, age, or disability. To file a complaint of discrimination, write USDA, Director, Office of Civil Rights, Room 326-W, Whitten Building, 1400 Independence Avenue, SW, Washington, DC 20250-9410 or call (202) 7205964 (voice and TDD). USDA is an equal opportunity provider and employer. Product or trade names may be registered trademarks, and are given only to identify materials used. Mention of specific products or trade names should not be considered an endorsement or recommendation by the authors. No claim to copyright can be made for original works produced by U.S. government employees as part of official duties. Original works by the U.S. government are in the public domain. Library of Congress Cataloging-in-Publication data. Ecology and management of a forested landscape : fifty years on the Savannah River Site / edited by John C. Kilgo and John I. Blake ; foreword by H. Ronald Pulliam. p. cm. Includes bibliographical references and index. ISBN 1-59726-010-X (cloth : alk. paper) — ISBN 1-59726-011-8 (pbk. : alk. paper) 1. Forest ecology—South Carolina—Savannah River Site. 2. Restoration ecology—South Carolina—Savannah River Site. I. Kilgo, John C. ( John Carlisle), 1967– II. Blake, John Irvin. QH105.S6E28 2005 333.75′153′097577—dc22 2004025494 British Cataloguing-in-Publication data available. Printed on recycled, acid-free paper Design by Paul Hotvedt Manufactured in the United States of America 10 9 8 7 6 5 4 3 2 1

Contents List of Figures and Tables ix Foreword xvii Preface xx Acknowledgments xxii

Chapter 1 The Savannah River Site, Past and Land-Use History 2 Industrial Operations and Current Land Use 12 Chapter 2 The Physical Environment Climate and Air Quality 20 Soils and Geology 30 Water Resources 41

19

Chapter 3 SRS Forest Management 57 Silviculture and Harvesting Activities 59 Prescribed Fire Management 75 Ecological Restoration 84 Chapter 4 Biotic Communities Plant Communities 106 Aquatic Invertebrates 161 Butterflies 175 Fishes 184 Amphibians and Reptiles 203 Nongame Birds 223 Nongame Mammals 253

103

Present 1

viii

Contents

Chapter 5 Threatened and Endangered Smooth Purple Coneflower 266 Sensitive Plants 275 Shortnose Sturgeon 282 American Alligator 285 Wood Stork 289 Bald Eagle 295 Red-Cockaded Woodpecker 301 Sensitive Animals 312

Species

Chapter 6 Harvestable Natural Resources Minerals 325 Commercial Forest Products 328 Fishery of the Savannah River 338 Small Game 341 Waterfowl 347 Wild Turkey 359 Furbearers 366 Wild Hog 374 White-Tailed Deer 380 Chapter 7

Conclusion

323

390

Appendix: Habitat Suitability Matrix for SRS Plants 401 Literature Cited 436 List of Reviewers 466 About the Authors 467 Index 469

264

List of Figures and Tables Figures Figure A. At the time of government acquisition, all towns and buildings were removed or demolished. xxi Figure 1.1. Streams and physiography of the Savannah River Site. 3 Figure 1.2. Pine savannas probably dominated most of the uplands in the area prior to European settlement. 4 Figure 1.3. Bottomland hardwood forests occurred on the floodplains of larger streams and rivers. 5 Figure 1.4. Pre-European vegetation types of the Savannah River Site. Color insert Figure 1.5. Cut-over condition of much of the Savannah River Site at the time of government acquisition. 11 Figure 1.6. Land use on the Savannah River Site in 1951. Color insert Figure 1.7. Satellite image of the Savannah River Site and surrounding region, March 1999. Color insert Figure 1.8. Land-use areas of the Savannah River Site. Color insert Figure 1.9. Aerial view of a developed area and surrounding forest on the Savannah River Site. 14 Figure 1.10. Size of the workforce on the Savannah River Site, 1987–2003. 16 Figure 2.1. Topographic relief on the Savannah River Site. 32 Figure 2.2. Geological stratigraphy and groundwater systems of the Savannah River Site. 34 Figure 2.3. General soil map of the Savannah River Site. Color insert Figure 2.4. Major streams, wetlands, and larger lakes of the Savannah River Site. 42 Figure 2.5. Relative mean monthly discharge for major streams on the Savannah River Site. 48 Figure 2.6. During reactor operations, the high flow rates and temperatures of reactor cooling water destroyed riparian vegetation in Fourmile Branch, Pen Branch, and Steel Creek. 51

ix

x

List of Figures and Tables

Figure 2.7. Aerial view of Ellenton Bay, a large Carolina bay bisected by a utility right-of-way. 55 Figure 2.8. Hydroperiods for fifty-six Carolina bays on the Savannah River Site. 56 Figure 3.1. Longleaf pine planted in an old field on the Savannah River Site, early 1950s. 61 Figure 3.2. Net number of acres planted 1953–2003 or seeded successfully 1960–1971 at the Savannah River Site for slash pine, loblolly pine, longleaf pine, and various hardwood species including cypress. 62 Figure 3.3. Longleaf pine planted in cutover scrub oak on the Savannah River Site, early 1950s. 63 Figure 3.4. Changes in silviculture and harvesting practices on the Savannah River Site 1952–2001. 71 Figure 3.5. Number of wildfires and average area per fire 1954–2002 on the Savannah River Site. 77 Figure 3.6. Trends in prescribed burning at the Savannah River Site, 1952–2002. 79 Figure 3.7. Under proper conditions, smoke from prescribed burning is carried upward and away from sensitive areas. 83 Figure 3.8. Locations of restoration projects on the Savannah River Site. 88 Figure 3.9. Aerial view of the Pen Branch corridor and delta on the Savannah River Site during reactor operations. 90 Figure 3.10. Degraded wetland areas of the Pen Branch corridor and delta on the Savannah River Site that were impacted by thermal releases from reactors and later restored as part of the mitigation effort. 91 Figure 3.11. Planting trees in the Pen Branch corridor on the Savannah River Site, 1993. 92 Figure 3.12. A drainage ditch from a Carolina bay on the Savannah River Site. 94 Figure 3.13. Aerial view of restored Carolina bays on the Savannah River Site. 98 Figure 3.14. Distribution of remnant and degraded savanna plant communities in relation to land-use and fire exclusion history, mapped for potential savanna restoration on a representative section of the Savannah River Site. 100 Figure 4.1. Forest land-use associations of the Savannah River Site. Color insert Figure 4.2. Potential vegetation types of the Savannah River Site. Color insert Figure 4.3. Pine savanna. 115 Figure 4.4. Sandhill woodland. 116 Figure 4.5. Forested Carolina bay. 123 Figure 4.6. Herbaceous Carolina bay. 126


xi

Figure 4.7. Longleaf pine plantation, two to three years old, with welldeveloped shrub-scrub understory. 128 Figure 4.8. Loblolly pine stand on an old-field site (“old-field pine”). 129 Figure 4.9. Mature loblolly pine stand with some understory development. 130 Figure 4.10. Mature slash pine stand with little understory but a hardwood midstory. 130 Figure 4.11. Upland hardwood forest. 131 Figure 4.12. Flooded swamp. 142 Figure 4.13. Bottomland hardwood forest with herbaceous understory. 149 Figure 4.14. Bottomland hardwood forest with switchcane understory. 150 Figure 4.15. Old-field conditions typical of rights-of-way and other open areas. 158 Figure 4.16. First-order (headwater) stream. 189 Figure 4.17. Third-order stream. 190 Figure 4.18. Terrestrial snakes associated with xeric upland habitats and mesic floodplain habitats on the Savannah River Site. 212 Figure 4.19. Aquatic snakes associated with stream systems and Carolina bays on the Savannah River Site. 213 Figure 4.20. Salamanders and frogs associated with Carolina bays on the Savannah River Site. 214 Figure 4.21. Turtles associated with Carolina bay wetlands on the Savannah River Site. 216 Figure 4.22. Locations of terrestrial refugia for wetland turtles in uplands surrounding Dry Bay on the Savannah River Site during autumn-winter, 1994–1997. Color insert Figure 4.23. Abundance of strong- and weak-excavating cavity-nesting birds and total bird species richness on plots with all coarse woody debris removed and with none removed on the Savannah River Site. 230 Figure 4.24. Abundance, species richness, and diversity of birds in three successional stages of bottomland hardwood forest on the Savannah River Site. 234 Figure 4.25. Probabilities of occurrence of four area-sensitive birds in bottomland hardwood forests of various widths on the Savannah River Site. 236 Figure 4.26. Number of shrub-successional bird species and total number of bird species in clear-cuts of various sizes on the Savannah River Site. 237 Figure 4.27. Densities of Bachman’s sparrows in clear-cuts isolated by various distances from areas with source populations on the Savannah River Site. 238 Figure 4.28. Number of small mammals captured in longleaf pine stands of various ages on the Savannah River Site. 257

xii


Figure 4.29. Number of cotton mice captured on plots where tornado damage created a pulse of dead wood in 1989 on the Savannah River Site. 261 Figure 4.30. Diversity and species richness of small mammals in three sizes of clear-cuts on the Savannah River Site. 262 Figure 5.1. Locations of smooth purple coneflower populations on the Savannah River Site. 269 Figure 5.2. The response of individual smooth purple coneflower plants to burning and cutting treatments at the Burma Road population area, Savannah River Site. 271 Figure 5.3. Flowering patterns of smooth purple coneflower following burning and cutting treatments at the Burma Road population area, Savannah River Site. 271 Figure 5.4. Potential shortnose sturgeon spawning habitat in the Savannah River adjacent to the Savannah River Site. 284 Figure 5.5. Population growth of American alligators in Par Pond on the Savannah River Site, 1972–1988. 287 Figure 5.6. Seasonal use of the Savannah River swamp system by wood storks, 1983–2002. 290 Figure 5.7. Average numbers of wood storks observed per aerial survey of the Savannah River swamp system, 1983–2002. 293 Figure 5.8. Locations of bald eagle nest sites and management areas on the Savannah River Site. 296 Figure 5.9. Number of groups and size of post-breeding-season population of red-cockaded woodpeckers on the Savannah River Site, 1975–2003. 304 Figure 5.10. Location of active and inactive red-cockaded woodpecker groups and recruitment stands within habitat management areas during 2001 on the Savannah River Site. 306 Figure 5.11. Artificial cavity inserts, developed at SRS, have become a critical tool in red-cockaded woodpecker recovery efforts rangewide. 307 Figure 5.12. A red-cockaded woodpecker cavity tree with an encroaching midstory below. 308 Figure 6.1. Volume of wood in softwoods and hardwoods sold on the Savannah River Site, 1955–2003. 335 Figure 6.2. Total value of wood sold for all species on the Savannah River Site, 1955–2000, and the average unit price of the wood sold during each year. 336 Figure 6.3. Habitats used by waterfowl and locations of nest boxes for breeding wood ducks and hooded mergansers on the Savannah River Site. 351 Figure 6.4. Population parameter estimates for female wood ducks using nest boxes on the Savannah River Site, 1979–1995. 354


xiii

Figure 6.5. Maximum numbers of ring-necked ducks, lesser scaup, buffleheads, and ruddy ducks observed per year during aerial surveys of Par Pond and L Lake on the Savannah River Site, 1982–2003. 358 Figure 6.6. Hunter recovery locations in the eastern United States of 594 ring-necked ducks originally banded on the Savannah River Site, 1985–2002. 359 Figure 6.7. Wild turkey observations recorded during South Carolina Department of Natural Resources summer brood surveys 1974–2003 on the Savannah River Site. 363 Figure 6.8. Number of Virginia opossum, raccoon, and striped skunk captured per year during the Small Furbearer Survey, Savannah River Site, 1954–1982. 367 Figure 6.9. Number of red fox, gray fox, and bobcat captured per year during the Small Furbearer Survey, Savannah River Site, 1954–1982. 370 Figure 6.10. Expansion of wild hog distribution on the Savannah River Site. 375 Figure 6.11. Estimated size of the deer population and number of deer harvested on the Savannah River Site, 1965–2003. 383 Figure 6.12. Relationship between the number of deer-vehicle accidents and (a) the estimated size of the deer population and (b) the size of the workforce on the Savannah River Site. 387

Tables Table 2.1. Mean monthly rainfall and extremes for the 773-A area at the Savannah River Site for the period 1952–2001. 22 Table 2.2. Predicted extreme precipitation recurrence estimates by accumulation period and observed extreme total precipitation received in the Savannah River Site region, August 1948–December 1995. 23 Table 2.3. Ranges for monthly mean, monthly high, and monthly low temperature and monthly mean, maximum, and minimum relative humidity, 1964–2001, from A Area at the Savannah River Site. 24 Table 2.4. Historical average pan evaporation at the Edisto Experiment Station, Blackville, South Carolina, 1963–1992. 25 Table 2.5. Monthly occurrences of tornadoes, hurricanes, thunderstorms, and snow or ice in the Savannah River Site region. 27 Table 2.6. Chemical characteristics of selected upland soils, by depth, on the Savannah River Site. 40 Table 2.7. Hydrologic characteristics of major streams on the Savannah River Site. 46 Table 2.8. Chemical characteristics of major streams on the Savannah River Site. 49

xiv


Table 3.1. Acreage treated by various silvicultural practices at the Savannah River Site 1952–2001. 65 Table 3.2. Pre- and postburn fuel loading and total fuel reduction. 80 Table 3.3. Observed annual mean twenty-four-hour PM10 values from three counties near the Savannah River Site. 84 Table 3.4. General ecological impacts from post-European settlement in the Central Savannah River Area and strategies for ecological restoration. 86 Table 3.5. Species richness for taxa in Pen Branch compared with disturbed post-thermal and late-successional forested reference sites at the Savannah River Site. 93 Table 3.6. Level of disturbance to surface hydrology by drainage ditches in isolated depression wetlands at the Savannah River Site in 2002. 95 Table 3.7. Effects of burning, harvesting, and harvesting plus burning on the average herbaceous species richness and percent wetland species occurring in Bay 93 on the Savannah River Site before and after closing the drainage ditch in 1994. 96 Table 3.8. Savanna grasses, composites, and legumes selected for experimental introduction to old-field pine sites at the Savannah River Site to establish founder populations. 101 Table 4.1. Extent of forest cover types on the Savannah River Site. 111 Table 4.2. Extent of vegetation types on the Savannah River Site. 114 Table 4.3. Percent basal area for species associated with sandhill woodland and remnant pine savanna communities on the Savannah River Site. 118 Table 4.4. Percent basal area for species associated with Carolina bay forests and savanna communities on the Savannah River Site. 124 Table 4.5. Percent basal area for species associated with upland oak-pine woodland and pine-hardwood forest communities on the Savannah River Site. 134 Table 4.6. Percent basal area for species associated with upland slope and hardwood communities on the Savannah River Site. 138 Table 4.7. Percent basal area for species associated with swamp communities on the Savannah River Site. 144 Table 4.8. Percent basal area for species associated with river and large stream bottom habitats on the Savannah River Site. 146 Table 4.9. Percent basal area for species associated with stream bottom communities on the Savannah River Site. 152 Table 4.10. Habitats of aquatic insects on the Savannah River Site. 162 Table 4.11. Habitats of aquatic arthropods on the Savannah River Site. 165 Table 4.12. Habitats of other aquatic invertebrates on the Savannah River Site. 166 Table 4.13. Conservation status of aquatic invertebrates of the Savannah River Site. 172


xv

Table 4.14. Butterfly species of the Savannah River Site, organized by family, with month and habitat of occurrence. 176 Table 4.15. Number of butterfly species on the Savannah River Site, by family. 183 Table 4.16. Fish species confirmed at the Savannah River Site. 185 Table 4.17. Relative density of fish in streams recovering from thermal impacts and in undisturbed streams on the Savannah River Site. 193 Table 4.18. Percent composition of fishes from Par Pond on the Savannah River Site, 1969–1980. 199 Table 4.19. Number of fish (and percent composition) captured in two studies of Carolina bays and isolated depression wetlands on the Savannah River Site. 201 Table 4.20. Habitat characterizations and rarity rankings of amphibians and reptiles of the Savannah River Site. 205 Table 4.21. A typology of species rankings for amphibians and reptiles on the Savannah River Site based on geographic range, habitat specificity, and local population size. 210 Table 4.22. Bird-habitat matrix for the Savannah River Site, South Carolina. 240 Table 4.23. Typical avian communities associated with six common habitats on the Savannah River Site. 228 Table 4.24. Taxonomic listing and conservation status of the mammals of the Savannah River Site. 254 Table 4.25. Primary habitats of nongame mammals of the Savannah River Site. 258 Table 4.26. Levels of foraging bat activity over nine habitats on the Savannah River Site. 260 Table 5.1. Number of ramets for three smooth purple coneflower populations on the Savannah River Site, 1988–2003. 270 Table 5.2. Sensitive plants occurring on the Savannah River Site, with their global and state ranking and number of populations for each species in 1990, 1995, and 2000. 276 Table 5.3. The Nature Conservancy and South Carolina Department of Natural Resources rarity and vulnerability rankings used on the Savannah River Site. 278 Table 5.4. Estimated population size and sex ratios of American alligators in Par Pond on the Savannah River Site 1972–1974 and 1986–1988. 286 Table 5.5. Wood stork use of the Savannah River swamp system, 1983–2000. 292 Table 5.6. Number of nestlings fledged by bald eagle nesting pairs on the Savannah River Site, 1986–2000. 299 Table 5.7. Numbers of red-cockaded woodpecker fledglings and groups on the Savannah River Site, 1990–2003. 304

xvi


Table 5.8. Acreage receiving midstory control and prescribed burning for red-cockaded woodpecker management on the Savannah River Site, 1990–2003. 309 Table 5.9. Number of red-cockaded woodpeckers translocated to the Savannah River Site, 1986–2000. 310 Table 5.10. Number of southern flying squirrels removed from redcockaded woodpecker cavities on the Savannah River Site, 1986–2003. 312 Table 6.1. Chemical formulas of minerals occurring at the Savannah River Site. 327 Table 6.2. Standing volume of pine and hardwood at the Savannah River Site at intervals, 1952 to 2001. 330 Table 6.3. Approximate distribution of the total forest area by stand age class and major commercial forest type using the Savannah River Site periodic stand mapping database. 331 Table 6.4. Estimated total number of trees by species and diameter class on the forested land area on the 2001 Savannah River Site in 1992. 333 Table 6.5. Comparative volume, value, and revenue sold from selected clear-cut or regeneration sales versus thinning or partial-cut sales 1987–1996 on the Savannah River Site. 336 Table 6.6. Area raked, total sales revenue, and unit value per acre for pine straw harvest at the Savannah River Site, 1991–2000. 337 Table 6.7. Estimate of percentage of fish species harvested from New Savannah Bluff Lock and Dam on the Savannah River during the 1999 access creel census. 340 Table 6.8. Christmas Bird Count data for small game birds at the Savannah River Site, 1979–2002. 342 Table 6.9. Small game harvest at Crackerneck Wildlife Management Area and Ecological Reserve, Savannah River Site, 1984–2003. 344 Table 6.10. Locations on the Savannah River Site where waterfowl and other selected aquatic birds have been observed, 1952–1997. 349 Table 6.11. Number of wild turkeys trapped on the Savannah River Site by the South Carolina Department of Natural Resources for off-site restocking programs, 1978–2000. 361 Table 6.12. Wild turkey harvest data recorded on Crackerneck Wildlife Management Area and Ecological Reserve, 1983–2003. 362 Table 6.13. Causes of mortality among 132 radio-instrumented wild turkeys on the Savannah River Site and the Crackerneck Wildlife Management Area and Ecological Reserve, 1998–2001. 363 Table 6.14. Annual number of beaver trapped on the Savannah River Site, 1983–2003. 369 Table 6.15. Number of wild hogs removed annually from the Savannah River Site, 1965–2003. 377

Foreword In 1539, Hernando de Soto and his band of six hundred soldiers, gold seekers, and Indian guides set out to explore the interior of what is now the southeastern United States. De Soto and his men traveled north and east from Florida and across the upper coastal plain of Georgia before crossing the middle Savannah River into South Carolina. Although their exact route is unknown, they would have passed through a heavily forested landscape, perhaps following Indian trails and sticking, as much as possible, to the open, sandhill scrub forest and longleaf pine–dominated uplands, avoiding the more difficult terrain of the tupelo-cypress swamps and bay forests of the bottomland floodplains. Though no doubt grand by modern-day standards and magnificent to behold, the forests encountered by De Soto had already been modified for centuries by Indians seeking to improve their hunting grounds and increase the abundance of edible berries and other wild foods. But the changes wrought by Native Americans were relatively minor compared to what was to come. Four hundred years after De Soto’s travels, the uplands of the upper coastal plain had been almost entirely cleared for intensive agriculture, and even much of the swampy lowlands had been drained and cleared. These dry, infertile lands provided a farmer little yield and a difficult life, however, so by the mid-twentieth century, many farmers had left, leaving the patchwork of abandoned farms and secondgrowth forests still seen throughout most of the upper coastal plain today. Can land degraded by centuries of poor agricultural practices be restored to something approaching its original productivity and diversity? This book tells the remarkable story of fifty years of natural resource management and restoration of the forested landscape of the Savannah River Site (SRS). In 1950, the Atomic Energy Commission began purchasing land and relocating thousands of descendants of the original European settlers who had cleared the land and tried to eek out a living from it. xvii

xviii

Foreword

Shortly afterward, researchers from the Universities of Georgia and South Carolina and the Philadelphia Academy of Sciences were invited to work on the site, and the USDA Forest Service began an aggressive program to replant and restore the forests. As a result of these efforts, the Savannah River Site is one of the best-studied ecological research sites in North America, and an amazing diversity of native flora and fauna exist in what was once corn and cotton fields, pastures, and degraded and poorly managed forests. Editors John Kilgo and John Blake have assembled a talented group of authors, all of whom are intimately familiar with the subject matter of their chapters. Some authors are university faculty who for years have traveled back and forth from schools across the country to work at the Savannah River Site because of the unique research environment the site offers. Others are permanent residents working on site at Westinghouse, the U.S. Forest Service, or the University of Georgia’s Savannah River Ecology Laboratory. Their collective knowledge of the history, ecology, and management of the Savannah River Site is itself a unique resource, and this book serves to make their knowledge and experience available to others. Today, most of the original forest traversed by De Soto is gone. In 1989, in “Longleaf pine and wiregrass: Keystone components of an endangered ecosystem” (Nat. Areas J. 9:211–213), Reed F. Noss estimated that less than 30 percent of bottomland and riparian forests and only 14 percent of longleaf forests remain in the Southeast and only 3 percent of longleaf habitat survives as old growth. Some of the unique species of the southeastern forests (e.g., Carolina parakeet, ivory-billed woodpecker, and Bachman’s warbler) are gone forever, but—though many of the remaining species are threatened or endangered—much of the original diversity of the region has survived. Our ability to ensure the long-term viability of the region’s biological diversity depends on three critical steps: (1) inventorying the existing diversity of native species, (2) determining the habitat requirements of the threatened species, and (3) restoring habitats and managing them to provide for the habitat requirements of native flora and fauna. In summarizing fifty years of research into the biotic communities and native species of the Savannah River Site, this book provides a comprehensive overview of the forest management practices that can support long-term forest recovery and restoration of native habitats. The success of the management efforts at SRS is attested to by the 103 species of reptiles and amphibians, 87 fish species, 69 species of dragonflies and damselflies,

Foreword

xix

99 species of butterflies, 64 rotifer species, and literally thousands of other species that still exist there. Not only the presence of species but also their habitat requirements have been documented in detail, even for often ignored groups such as aquatic invertebrates. As a result of reintroducing or regenerating appropriate native species, restoring natural hydrological cycles in the lowlands and regular burning in the uplands, controlling non-native invasive species, and carefully regulating hunting and fishing, the native flora and fauna of the Savannah River Site is flourishing. Our ability to preserve the native biological diversity of the southeastern United States, or any other region of the world, over the next thousand, or even hundred, years is still uncertain. There are those who feel we have done too little too late, and the loss of habitat and poor management practices of the past combined with our ignorance and greed in the future will inevitably lead to massive losses of biological diversity. This book stands as a counterargument to that bleak and gloomy view of the future and provides a concrete example of the role that good science combined with good management can play in ensuring that our descendants will be able to enjoy the splendors of nature that have delighted our own generation. H. Ronald Pulliam Regents Professor of Ecology University of Georgia August 12, 2004

Preface In 1950, the United States Department of Energy (then the U.S. Atomic Energy Commission) began purchasing the land that became the present Savannah River Site (SRS). All residents were removed (figure A), and in 1951 the government closed the site to the public to begin work on production of nuclear weapons materials. At the time, abandoned agricultural fields dominated upland areas, and the SRS and the USDA Forest Service initiated an aggressive reforestation program. Concurrently, the primary site contractor at the time, E.I. DuPont de Nemours Co., subcontracted researchers from the University of South Carolina, the Philadelphia Academy of Sciences, and the University of Georgia (which would eventually establish the Savannah River Ecology Laboratory) to initiate baseline ecological surveys of the site. Since that time, researchers from those organizations and many others have intensively studied and monitored the natural resources of the SRS. The initial inventory of the fauna and flora established both a baseline for future comparison and a philosophy of stewardship for resources that persists today. Although management objectives have changed, the SRS goal for stewardship has remained focused upon innovative leadership in resource management through sound scientific and technical strategies. In 1972, the Department of Energy designated the SRS as the nation’s first National Environmental Research Park, a place where the effects of human impacts on the environment could be studied. The SRS has provided excellent opportunities for research within that concept. The comprehensive nature and scope of information on the ecology of the site and its resources is unparalleled. The SRS has made this information available to the public through numerous professional journals, reports, and publications by the Savannah River Ecology Laboratory, the Savannah River Technology Center, the South Carolina Archeology Research Program, the U.S. Forest Service, xx

Preface

xxi

Figure A. At the time of government acquisition, all towns and buildings were removed or demolished (J. Kilgo).

cooperating universities, and other agencies. The SRS has periodically published overviews of the natural resources in various formats. However, no publication has integrated information on ecology, natural resources, and management practices, and various public groups have expressed a desire to obtain that relevant scientific and technical information about the site in a single document. This book tells the story of the fifty-year period after human residents moved from that 310-square-mile tract of land in the South Carolina coastal plain. Human impact has continued, to be sure. The SRS workforce approached twenty-five thousand at its peak in 1991. Nuclear reactors and related facilities have been constructed, as well as several large cooling reservoirs, and environmental contamination has occurred (there are sites on SRS designated under the provisions of the Resource Conservation and Recovery Act and the Comprehensive Environmental Response, Compensation, and Liability Act). These impacts have generally been localized within the site, however; industrial development (not including rights-of-way and reservoirs) constitutes less than 3 percent of the site’s area, and surface contamination exists in only 0.6 percent of the area. The SRS manages its forests on a far longer rotation length than most managed lands in the Southeast. Thus, the vast majority of the land area of SRS has suffered relatively minimal human impact in the past fifty years. We hope that this book will provide its readers with a better understanding of the plant and animal populations and communities present on the SRS and the effect on them of fifty years of land management by the Department of Energy.

Acknowledgments This work was supported by the U.S. Department of Energy–Savannah River Operations Office through the U.S. Forest Service–Savannah River (USFS-SR) under Interagency Agreement No. DE-AI09-00SR22188, which also supported authors from USFS-SR. Authors from the Savannah River Ecology Laboratory (SREL) were supported by the Environmental Remediation Sciences Division of the Office of Biological and Environmental Research, U.S. Department of Energy, through Financial Assistance Award No. DE-FC09-96SR18546 to the University of Georgia Research Foundation. Authors from Westinghouse Savannah River Company were supported by the U.S. Department of Energy under contract DE-AC0996SR18500. Authors from the U.S. Forest Service Southern Research Station (USFS-SRS) were supported by that agency. Many individuals generously contributed their time, efforts, and ideas to make this book possible. Elizabeth LeMaster, formerly of USFS-SR, was instrumental in the original conception of the book. Special thanks are offered to Dumitru Salajanu and Andrew Thompson (USFS-SR) for creating most of the maps used herein, to David Scott for providing many of the photographs, and to Kim Hale for support in putting it all together. Donald Von Blaricom (Strom Thurmond Institute, Clemson University, South Carolina) provided figure 1.3 and associated image analysis. Deno Karapatakis (SREL) provided figure 1.4. Dean Fletcher (SREL) provided the list of SRS fishes in chapter 4. Kay Franzreb and Chuck Daschelet (USFS-SRS) collected much of the unpublished red-cockaded woodpecker data in chapter 5. The late Tom Lloyd provided invaluable assistance with the forest inventory data in chapter 6. Finally, we wish to thank the multitude of land management professionals, from many organizations, whose diligent work during the past fifty years has resulted in the unique resource that is the Savannah River Site.

xxii

1

r

The Savannah River Site, Past and Present Land-Use History David L. White

Industrial Operations and Current Land Use John I. Blake, John J. Mayer, and John C. Kilgo

The land area now owned by the U.S. Department of Energy and known as the Savannah River Site (SRS) has been occupied by humans for about 11,500 years. In the section titled “Land-Use History,” David White describes the vegetation of the area prior to European settlement and then provides a brief overview of the area’s long and varied history, with an emphasis on the impacts of humans upon the landscape. Native Americans influenced the landscape through their use of fire and agriculture. Around 1700, Savannah Town was established as the first European settlement in inland South Carolina, approximately 20 km north of the present SRS. Although residents grazed cattle and hogs in the woodlands and began to affect native wildlife populations, agriculture was not well established until the late 1700s, after which, land clearing increased dramatically. Timber and cotton became the dominant products of the area. By 1950, when the government acquired the land, much of the site had been cut repeatedly and most of the uplands were in agricultural fields or bare ground. The SRS contracted the U.S. Forest Service to reforest the site in 1951. Today, the SRS is almost completely forested 1

2


and contrasts greatly with the surrounding landscape, which is dominated by agriculture and suburban development. (The material in this section was condensed and summarized from White and Gaines, 2000.) In “Industrial Operations and Current Land Use,” John Blake et al. first outline in general terms the primary missions, activities, and infrastructure of SRS. They then describe the land-use zones, including habitat management areas for the endangered red-cockaded woodpecker (a primary habitat management area, a supplemental habitat management area, and an other-use area), the Crackerneck Wildlife Management Area and Ecological Reserve (managed cooperatively by the South Carolina Department of Natural Resources), and the research set-aside areas. Collectively, these areas form the framework within which SRS land management is conducted.

Land-Use History David L. White Creation of the 80,267-ha (198,344-ac or 310-mi2) Savannah River Site (SRS) by the U.S. Department of Energy (DOE, formerly the Atomic Energy Commission, AEC) in 1951 set the stage for a dramatic change in land use. Construction of nuclear production facilities and the reforestation of abandoned farmland and cutover forests affected SRS ecosystems in profound ways. The construction and operation of nuclear facilities from 1953 to 1988 directly impacted about 4,000 ha (9,884 ac) of land, created almost 2,000 ha (4,942 ac) of cooling reservoirs, and released thermal effluent in all but one major SRS stream (Upper Three Runs). Nuclear facilities now on the site include five deactivated reactors, as well as facilities for nuclear materials processing, tritium extraction and purification, waste management, solid waste disposal, and power plants for steam generation and production of electric power (Noah 1995). This section describes the land that became the SRS and the historical uses of that land, focusing on agricultural and natural resource uses of the area. The SRS is located on the Upper Coastal Plain and Sandhills physiographic provinces, 30 km south of the Piedmont Plateau (figure 1.1). It is south of Aiken, South Carolina, and includes portions of Aiken, Barnwell, and Allendale Counties. Kolka et al. describe the soils and physiography of the SRS in chapter 2.

The Savannah River Site, Past and Present

3

Figure 1.1. Streams and physiography of the Savannah River Site.

Pre-European Settlement Vegetation For the past ten thousand years, oak and pine forests have dominated the SRS area. Pine species probably have dominated the uplands of the area for the past four to five thousand years (Watts 1971, 1980; Delcourt and Delcourt 1987). Views of pre- or early-settlement forests in the Central Savannah River Area (CSRA) and adjacent regions from the 1700 and 1800s help characterize the distribution of plant communities in the region (Von

4


Figure 1.2. Pine savannas probably dominated most of the uplands in the area prior to European settlement (J. Kilgo).

Reck 1733; Michaux 1805; Mills 1826; Lieber 1860; Sargent 1884; Cordle 1939; Bartram 1942; Bartram 1958; Lawson 1967; Drayton 1996). Generally, longleaf pine dominated the uplands (figure 1.2), while hardwoods, ranging from oak-hickory to cypress-tupelo forests, dominated the “clay land,” terraces, and flood plains (figure 1.3). Canebrakes in adjacent regions (Logan 1858; Lawson 1967) and the existence of remnant patches within the SRS suggest that these communities were common. Frost (1997) described composition and distribution of eleven presettlement vegetation types (figure 1.4, in color insert). He defined community types from soils, historical data, and remnant vegetation. Longleaf pine was dominant on 63 percent of SRS forests (80 percent of non-wetland areas). Swamps, bottomland, and bay forests occupied 22 percent of the site. Estimates of fire-return intervals ranged from one to three years on the Aiken Plateau to seven to twelve years on more fire-sheltered sites.

Land Use before 1950 The SRS area was used extensively by people prior to the establishment of the Site in 1951. I consider three broad time periods prior to 1951: preEuropean settlement, settlement to 1865, and 1865–1950.


5

Figure 1.3. Bottomland hardwood forests occurred on the floodplains of larger streams and rivers (J. Kilgo).

Pre-European Settlement Aboriginal people entered the SRS area about 11,500 years before the present (BP), though early use was sporadic and transient and probably concentrated along bottomlands and terraces adjacent to streams (Sassaman et al. 1990; Sassaman 1993). Sustained seasonal habitation of the area began between 9,800 and 8,000 years BP, with winter residential bases along the first terrace of the Savannah River near the mouths of major tributaries. Although use of the region may have declined between 8,000 and 6,000 years BP with a warming and drying climate, aboriginal populations began to increase again around 6,000 years BP. By 3,000 years BP, hunting parties used the Aiken Plateau at least seasonally (Sassaman 1993), and between 3,000 and 2,500 years BP, occupation of the Aiken Plateau became more intensive and perennial. Population density apparently fluctuated until the mid-1400s, when a significant portion of the aboriginal population is thought to have abandoned the CSRA, probably as a result of political actions of chiefdoms outside the immediate area (Sassaman et al. 1990; Anderson 1994). A severe drought in the mid1400s also may have affected the distribution of aboriginal populations (Stahle and Cleaveland 1992; Anderson 1994). When Hernando de Soto passed through the middle Savannah River valley in 1541, he found no

6


people in five days of travel from present-day Greensboro, Georgia, to the Savannah River and beyond, further supporting the contention that significant aboriginal populations were absent in the CSRA during the two centuries preceding European settlement. Native Americans had significant impacts on the southeastern landscape through their use of fire and agriculture. They used fire extensively for hunting and land clearing, although the extent of its historical use at the SRS is not known. In contrast to fires ignited by lightning strikes, which are most frequent during the spring and summer, Native Americans set fires during the fall, winter, and spring. Alteration of fire season and frequency, especially on the more mesic part of the landscape, may represent the largest-scale impact on the landscape by Native Americans in the region (White 2004). Native American agriculture apparently did not begin in the CSRA until approximately 800 years BP (Sassaman et al. 1990), later than elsewhere in the Southeast, and its extent is not known. Areas along streams were used most extensively, corn, beans, and squash being the main crops. Land clearing involved various ways of killing trees followed by burning. Native Americans practiced field rotation but not crop rotation. Generally, aboriginal agricultural techniques were much less erosive and damaging to the soil than those associated with Europeans after settlement (Herndon 1967; Trimble 1974). The population declines during the 1400s and 1500s probably had a significant impact on fire dynamics, the area cleared for cultivation, and the level of hunting pressure, but the degree of impact is not known. Thus, the CSRA landscape first described by explorers and settlers in the late 1600s resulted from a combination of natural disturbance patterns and, to a lesser extent, those brought about by Native Americans.

Settlement to 1865 Savannah Town, 20 km (13 mi) northwest of the current SRS boundary and just south of Augusta, Georgia, became the first inland settlement in South Carolina around 1700 and served as an important trading post. Whether the proximity to Savannah Town directly affected the SRS area is not known. The earliest land plats on the present-day SRS date from the 1730s (Brooks and Crass 1991), but settlement of the area did not occur until the 1760s (Brooks 1988). Woodland cattle grazing probably occurred in the SRS between the 1730s and the 1760s, but the dates and extent are not known (Brown 1894; Meriwether 1940; Brooks 1988). The


7

predominant land use before 1780 was woodland cattle grazing and scattered small-scale farming. Crop cultivation and timber cutting prior to 1780 was limited and occurred primarily along streams and terraces (Brown 1894). Planters grew rice and indigo to an unknown extent. Cowpens were common in the SRS area in the 1700s (Brown 1894; Bartram 1942). They were mostly 40 to 160-ha (100–395-ac) cleared areas with enclosures for cattle, horses, and hogs and buildings for the cowpen keepers (Dunbar 1961). Cattle also grazed the uncleared upland forests, bays, and bottomlands along streams. They used savannas in summer and cane swamps in winter. The widespread abundance of cattle likely impacted native grazers, cane and other forage plants (see the appendix for scientific names of plants), and soil erosion and water quality along streams and near cowpens. Hogs were abundant in the region (Schoepf 1911; Frost 1993), but their abundance in the CSRA was not documented until 1825 (Mills 1826). Cattle and hog abundance peaked in 1850. Hogs directly impacted the regeneration and survival of longleaf pine (Schoepf 1911) and competed with species that were dependent on hardwood mast. Several local (Mills 1826; Brown 1894) and regional (Ashe 1682; Von Reck 1733; Logan 1858; Chapman 1897; Bartram 1958; Lawson 1967) references cite an abundance of gray (Canis lupus) and red wolves (Canis rufus), panthers (cougar, Felis concolor), and “wild cats” (bobcat, Lynx rufus), as well as game species, notably white-tailed deer (Odocoileus virginianus) and wild turkey (Meleagris gallopavo). Bison (Bison bison) were also probably abundant based on their numbers above (Logan 1858) and below (Von Reck 1733) the SRS. Tarleton Brown (1894), who lived near the SRS in 1769 and later along Lower Three Runs, and Mills (1826) describe the abundance of certain predator and game species and the constant effort to eliminate the former. Logan (1858) characterized the dynamic relationship between the decline of the native fauna, the process of settlement, and the extensive peltry trade with Native Americans in the South Carolina upcountry (Piedmont). Much of this information is relevant to the SRS area. South Carolina passed laws to control or eliminate predators from 1695 to 1786 (Heaton 1972). Bison and the large predators were the first species eliminated, largely before 1800. White-tailed deer, black bear (Ursus americanus), beaver (Castor canadensis), and other species were reduced dramatically before 1800; other species such as the raccoon (Procyon lotor), opossum (Didelphis virginiana), muskrat (Ondatra zibethicus), and squirrel (Sciurus spp.) suffered declines throughout the 1800s. By 1900, the Carolina parakeet (Conuropsis carolinensis) and the passenger pigeon (Ectopistes migratorius) were extinct or

8


near extinction (Salley 1911), as was the ivory-billed woodpecker (Campephilus principalis), though due to habitat destruction as opposed to direct harvest. Establishment of grain and sawmills on SRS streams increased in the late 1700s. From 1780 to 1865, there was a dramatic increase in cotton farming, and by 1825 cotton and lumber were the primary staples in the CSRA. From 1825 to 1860, the amount of improved land (defined in the 1850 census as “only such as produces crops, or in some manner adds to the productions of the farmer”) increased from 4 percent to 31 percent of the total, so that in 1860, about 70 percent of the land on farms was woodland. Though many swamps, bays, and creek bottoms of the Upper Coastal Plain were cleared, drained, and cultivated between 1845 and 1860 (Hammond 1883), SRS swamp forests along the Savannah River in the 1840s were relatively intact, with only patchy human disturbance (Ruffin 1992). However, timber and fuelwood harvests in the upland forests were substantial before 1865. Sawmills were abundant on SRS streams (Brooks and Crass 1991; Ruffin 1992). Lumbermen released floodgates on SRS streams to facilitate transport of rafts of lumber to Savannah. The 1840 census indicates that forests within the Barnwell district were used more than those in surrounding counties, or in many areas of the southeastern United States. Demands on forests included the 1833 construction and operation of the Charleston to Hamburg (North Augusta) Railroad, Savannah River steamboats, and domestic fuelwood use.

1865–1950 Following the Civil War, a cycle of poverty, cotton dependence, and land abuse developed in the South and persisted for most of the period from 1865 to 1950. Increased pressures on the land for production of cotton and other crops, naval stores (tar, pitch, and turpentine), fuelwood, and timber left only scattered patches of relatively untouched land. A significant shift in settlement toward the upland sandhills and an increasing trend away from watercourses occurred in the SRS after 1865 (Brooks and Crass 1991), corresponding to an increased emphasis on cotton production and a decrease in available farmland. Within the CSRA, land-use intensity peaked in the 1920s with the peak in cotton production and following extensive forest cutting. Approximately 30 percent and 45 percent of Aiken and Barnwell Counties, respectively, was improved land (mostly cultivated) during


9

most of the period from 1900 to 1950, with cotton and corn production accounting for the majority of cultivated land. “Shifting agriculture,” the abandonment of “worn out” land for “new” land, prevailed in the nineteenth and twentieth centuries. The abandoned land eventually reverted to forest. As a result, estimates of land under cultivation at any time mask or underrepresent the cumulative impacts of cultivation on the landscape. During this period, most of the SRS consisted of relatively small, dispersed farms, largely related to the increase in tenant farming after 1865. Tenancy peaked in 1925, and erosional land use increased with tenancy (Trimble 1974). Mechanization of southern agriculture did not occur until the 1930s and came even later to most of the farms of the SRS (Cabak and Inkrot 1996). While soil erosion increased after 1870, it was probably not extensive until after 1900. However, based on local soil descriptions for the SRS area (Carter et al. 1914; H. H. Bennett 1928; Rogers 1990), severe erosion was not common, and even moderate erosion was not extensive. Drainage and cultivation of upland depressions and bays in Barnwell County were uncommon before 1912 (Carter et al. 1914) but increased rapidly after 1930. An estimated two thirds of depression wetlands on the SRS ultimately were drained, primarily for agricultural purposes (see chapter 3). Agricultural chemical use in the SRS area increased significantly in the late 1800s with the dramatic increase in fertilizer use (South Carolina Department of Agriculture, Commerce and Industries and Clemson College 1927). With the arrival of the boll weevil in South Carolina in 1917, farmers initiated applications of calcium arsenate, and by the 1930s most CSRA farmers were “mopping” cotton crops with calcium arsenate, water, and molasses (Brunson 1930; South Carolina Extension Service 1940, 1946; A. Barker, Allendale, S. C., pers. comm.). This mixture was the predominant pesticide used in the area until the late 1940s, when farmers began using DDT and other organic pesticides for a variety of cotton pests (Boylston, Nettles, and Sparks 1948; South Carolina Extension Service 1951). Forest use, in the form of land clearing, logging, and turpentining, increased dramatically between 1865 and 1950. U.S. Census records and other records (Frothingham and Nelson 1944) suggest that naval stores production peaked in CSRA counties between 1880 and 1890 after the statewide peak in 1879. Statewide production fell sharply after 1890 but increased again after 1920. Longleaf pine was still quite prevalent in CSRA forests in the 1880s (Anonymous 1867; Hammond 1883), and loggers did not cut much of

10


the river swamp until about 1900 (Fetters 1990). Between 1910 and the early 1930s, extensive railroad logging occurred within the SRS. At least nine companies logged the SRS with at least 22 km (14 mi) of rail line along the swamp, 40 km (25 mi) along Upper Three Runs, and unknown amounts along other streams. Between 1880 and 1925, the area of woodland on farms decreased from 65 percent to 33 percent. By 1938, logging had impacted 70 percent of the Savannah River swamp with additional operations occurring between 1938 and 1950 (Mackey and Irwin 1994). In the late 1940s, sawtimber and pulpwood harvests throughout Aiken and Barnwell Counties were extensive (McCormack 1948). Other significant drains on forest resources included harvests for fencing, fuelwood, and the railroads. Use of the yellow pines and other species as fuelwood continued until the 1890s, but nationally and regionally the railroads’ impact peaked in the 1880s. Initial clearing for construction alone yielded an estimated 3 to 12 ha of cleared line per kilometer of rail (11–48 ac per mile; derived from Derrick 1930). Within the SRS, rail lines were built after the Civil War. The railroads brought increased use of longleaf pine and swamp forests, creating new land for crops and eventually creating settlements and towns, from which many agricultural and timber products flowed. The rather rapid decline of longleaf pine during the late nineteenth and early twentieth centuries resulted from a combination of factors, including hogs, destructive wildfires, and naval stores activities (Ashe 1894). Hog saturation densities in Barnwell County were high enough between 1840 and 1900 to severely impact longleaf pine establishment (Frost 1993). A decline in fire frequency after 1880, related to passage of stock laws, further impacted establishment of longleaf pine. After 1880, pressures on the land from agriculture and wood use, coupled with fire suppression efforts of the 1930s, drastically reduced the once extensive longleaf pine forests in the SRS and throughout the rest of the South.

Land Condition in 1951 and 2001 After the Atomic Energy Commission acquired the SRS in 1951, it authorized the U.S. Forest Service to manage most of the land and to act as consultant to the AEC and the DuPont Company, the project contractor (Savannah River Operations Office 1959). Much of the site had been cut repeatedly, and the timber was of little value (figure 1.5). A 1951 forest inventory conducted for a real estate appraisal classified about 48,724 ha (120,400 ac) as forest land, including 25,643 ha (63,365 ac) as


11

Figure 1.5. Cut-over condition of much of the Savannah River Site at the time of government acquisition (U.S. Forest Service files).

pine, 10,296 ha (25,443 ac) as hardwood, 11,021 ha (27,233 ac) as swamp, and 1,764 ha (4,358 ac) as plantation (U.S. Army Corps of Engineers 1951). The remaining 32,265 ha (79,727 ac) were in agricultural land. These figures include existing roads, buildings, and other infrastructure and therefore overestimate actual vegetated areas. Recent analysis (Sumerall and Lloyd 1995; White 2004) of an orthorectified mosaic of 1951 aerial photos (figure 1.6, in color insert) yielded results comparable to the inventory appraisal and estimates by the Savannah River Operations Office (1959). Agriculture accounted for 38 percent of SRS land. Most of this was cropland or recently plowed ground. The majority of the uplands were in agricultural fields and bare ground. The two forested land classes consisted of “forest,” which represented mostly intact forest, much of which was distributed along streams and the Savannah River (44 percent), and “regenerating forest,” which represented regenerating woody vegetation from abandoned agricultural land and cutover forests (18 percent). The initial focus of management was to reforest abandoned farmland, and by 1960, the Forest Service had planted 24,000 ha (59,304 ac; see chapter 3 for details). Forested land increased dramatically between 1951 and 1988 (White and Gaines 2000). In 2001, virtually all of the SRS was

12


forested; only 12 percent of the forest stands were less than ten years old, and 72 percent were more than thirty years old. Satellite imagery of the region illustrates the impacts of reforestation of the SRS (figure 1.7, in color insert). The green, forested SRS contrasts sharply with the surrounding landscape, dominated by agriculture and urbanization.

Industrial Operations and Current Land Use John I. Blake, John J. Mayer, and John C. Kilgo The management of natural resources at the Savannah River Site (SRS) has been variously executed over the years to meet conservation and restoration objectives, to provide research and educational opportunities, and to generate revenue from the sale of forest products. However, these management activities have been implemented under the constraints imposed by the Site’s nuclear mission and the objectives for which the SRS was established. This management challenge has been further complicated by the vast area encompassed by the Site, as well as the complex spatial mosaic of operational facilities and natural features. This section provides a general description of both the operational infrastructure and the land-use framework within which natural resource management activities occur.

SRS Background and Operations The SRS is one of several government-owned, contractor-operated sites within the U.S. Department of Energy’s nuclear defense complex. It is managed as a controlled area with limited public access. It was constructed during the 1950s to produce basic materials (e.g., plutonium-237 and tritium) used in nuclear weapons. Responsibility for these activities was initially assigned to the Atomic Energy Commission, whose mission was later assumed by the Department of Energy. Following the end of the Cold War, the Site’s mission changed to stewardship of the nation’s nuclear weapons stockpile, nuclear materials, and the environment (Mamatey 2004). Activities associated with the nuclear mission at SRS occur in several industrialized or developed areas located around the site. There are five nuclear production reactors; two chemical separations facilities; a heavy water extraction plant; a nuclear fuel and target fabrication facility; a tritium extraction facility; waste processing, storage, and disposal facilities;


13

and various administrative support facilities. The production reactors, the heavy water extraction plant, and the nuclear fuel and target fabrication facility are no longer operational. The last reactor was shut down in 1988. Several of these latter facilities have been decommissioned, and the remainder are scheduled to be decommissioned by 2026 (Austin, Noah, and Nelson 2003). SRS facilities are located in twenty separate developed areas around the site, which encompass a total of 1,781 ha (4,403 ac). The administrative areas are situated around the periphery of the site, while the industrialized operations areas (e.g., nuclear reactors, separations and waste management facilities) are in the inner core of the 803-km2 (310-mi2) footprint, with sufficient buffer lands to protect both the surrounding communities and the security of these classified operations (figure 1.8, in color insert). Additionally, remote facilities, less than 1 to 2 ha (1–5 ac) in size, are scattered around the site. They include power substations, sanitary wastewater treatment facilities and lift stations, cooling water intake and pump stations, field laboratories, maintenance buildings, and various security facilities. Perimeter security barricades control personnel and vehicle access. The infrastructure necessary to support these various administrative and operations areas is massive. Site utilities provide electricity, steam, cooling water, domestic water, service water, and sanitary waste treatment. The SRS has an extensive internal transportation infrastructure, which consists of approximately 225 km (140 mi) of primary roads and 2,253 km (1,400 mi) of secondary roads (including logging roads and jeep trails). Recent traffic flow on primary roadways has been in the thousands of vehicles per hour during periods of worker shift change. The SRS has a railway system consisting of approximately 96 km (60 mi) of track. It also has used the Savannah River to transport large, heavy loads to the site. The various pipelines, transmission lines, roads, and railways all have maintained rights-of-way associated with them (Noah 1995). Buffer zones between industrialized areas and surrounding undeveloped habitats are minimal (figure 1.9). Most transitions are abrupt, with maintained lawns or parking lots ending at the forest edge. Due largely to the close proximity of industrialized and undeveloped areas, the industrialized areas are used by various wildlife species. The presence of a number of medium-sized species (e.g., opossum, eastern cottontail, gray fox, and raccoon) within facility areas demonstrates that perimeter fences do not effectively deter wildlife movement. Mayer and Wike (1997) documented 153 species in and around developed portions of the site. However, they considered most (58.3 percent) uncommon in these areas, and

14


Figure 1.9. Aerial view of a developed area and surrounding forest on the Savannah River Site (Westinghouse Savannah River Co. files).

introduced or invasive species made up 50 percent of the abundant species. Foraging and feeding were the most commonly observed activities. Of the eight subhabitats surveyed, landscaped areas away from buildings and structures were the most heavily used. Potential impacts to humans from such urban wildlife include contaminant transport, physical injury, disease transmission, and destruction of property. Potential impacts to wildlife in these areas include physical harm and contaminant exposure (Mayer and Wike 1997). In an effort to fulfill its nuclear operations in a safe, secure, and environmentally responsible manner, the SRS has operated an extensive environmental monitoring program since 1951. Both on-site and off-site locations and media are monitored for potential impacts. Monitoring programs cover a suite of potential contamination pathways, including surface water, groundwater, drinking water, ingestion, contact, and air.


15

Annually, thousands of samples (air, water, soil, sediment, food, vegetation, and animal tissue) from both within and around the site are taken to support different analyses, and the potential human dose impacts are calculated for the different pathways. In 2003, the estimated dose to the maximally exposed individual from all pathways was 0.19 millirem (mrem; Mamatey 2004), which is 0.05 percent of the dose (360 mrem) received annually by people from natural and other manufactured sources of radiation (e.g., x-ray, television; Arnett and Mamatey 2000). Screening of both aquatic and terrestrial biota doses for 2003, the most recent year available, resulted in all sampled sites passing the pathway screening (Mamatey 2004). The SRS has significant social and economic effects on the area outside of its boundary. It contributes to South Carolina and Georgia through employment and purchasing and through educational, research, technology transfer, business development, and community assistance programs. The site is located in the Central Savannah River Area, consisting of eight counties in South Carolina and Georgia. The region contains eight county governments and thirty-eight incorporated municipalities. SRS employment has varied over the life of the Site, with a maximum of 38,582 employees during the peak construction period in 1952. During the early 1990s, the SRS was the largest single employer in South Carolina (Reed et al. 2002; Grewal and Noah 2004). However, employment has declined in recent years with the Site’s reduced post–Cold War missions (figure 1.10). Stewardship plans for the SRS have been developed for the next fifty years. In the near term, work will continue to improve environmental quality, clean up legacy waste sites, and manage any future waste produced from Site operations. This effort will include the construction of new facilities, retooling of existing Site facilities for new missions, and reconfiguration of the Site to a form that is more conducive to meeting mission requirements. In the decades ahead, SRS will consolidate its functions toward the center of the site. As new missions are funded, facilities will be placed near areas of current industrialization to minimize maintenance costs, infrastructure needs, and developmental and environmental impacts. Natural resource management is an integral component of the SRS Long Range Comprehensive Plan (U.S. Department of Energy 2000). Specifically, the plan defines three natural resource goals: demonstrate excellence in environmental stewardship; provide natural resource information critical to the Department of Energy’s science base; and provide cost-effective, flexible, and compatible programs to support SRS missions.

16


Figure 1.10. Size of the workforce on the Savannah River Site, 1987–2003.

Current cleanup efforts at many Department of Energy sites, including SRS, cannot restore those federal lands to acceptable levels for unrestricted public use. This is due in part to the nature of the contamination and the lack of proven cleanup and treatment technologies. Some hazards may require attention for many centuries. Consequently, long-term stewardship will be needed at those sites to ensure that the selected remedies will remain protective for future generations (U.S. Department of Energy 2000).

Natural Resource Management Because the SRS conducts natural resource management within the framework of several land-use areas (see figure 1.8), knowledge of the objectives for those areas is important in understanding SRS land management. The SRS Long Range Comprehensive Plan (U.S. Department of Energy 2000), the Land Use Baseline Report (Noah 1995), and the Natural Resource Management Plan (U.S. Department of Energy 2005) provide overviews of land-use conditions, strategies, and activities. More detailed information on specific management objectives and practices within particular zones can be found elsewhere (NUS 1984; Davis and Janecek 1997; Edwards et al. 2000; Caudell 2000). Here we provide general background information on natural resource management in the major land-use areas and the rationale for partitioning the site.


17

The various programs and entities with land-use areas include the redcockaded woodpecker (Picoides borealis) management program, the Crackerneck Wildlife Management Area and Ecological Reserve, and the Department of Energy Set-Aside Program. Although other endangered and threatened species occur on SRS, the red-cockaded woodpecker recovery program influences the largest portion of the landscape (Edwards et al. 2000). About two thirds of the upland forest areas are managed for this species and for the associated fire-maintained savanna conditions that support a great diversity of species. In the mid-1980s, the first woodpecker management plan delineated the SRS roughly as a donut shape, with the outer perimeter as the recovery area and the core containing the industrial areas. In 1997, a new plan detailed the current red-cockaded woodpecker habitat management areas (see figure 1.8). Primary factors considered from a landscape perspective included minimizing smoke problems from prescribed burning, optimizing savanna restoration opportunities through compatibility with ecological land classification, increasing management flexibility, and retaining prime industrial development sites. The plan incorporated the Department of Defense concept of including a “supplemental habitat management area” where lower woodpecker population densities are accepted to achieve greater flexibility. The woodpecker management plan provides specific guidelines on the kind and amount of timber harvest, development, and other activity allowed in each zone (Edwards et al. 2000). Within the industrial core or “Other Use Area” (figure 1.8) are most of the original industrial facilities. Infrastructure developments that dissect the area heavily impact wildlife (Mayer and Wike 1997) and other natural resources. They include transportation, power, and communications facilities; monitoring equipment; soil and groundwater closure projects; and support facilities. In order to minimize mission conflicts, there is a need to maintain industrial management flexibility and to limit natural resource goals in this zone. However, at least one population of an endangered plant, numerous sensitive species, and considerable wetland habitat occur near the industrial facilities. The South Carolina Department of Natural Resources, in conjunction with the U.S. Department of Energy, manages the Crackerneck Wildlife Management Area and Ecological Reserve primarily as wildlife habitat to enhance recreational hunting, fishing, and nonconsumptive use (Caudell 2000). Objectives are similar to those on many state lands and wildlife management areas. The Crackerneck area encompasses about 4,450 ha (11,000 ac) of wetland and mesic land with predominately pine forest,

18


bottomland hardwood, and cypress-tupelo swamp habitats. Prior to SRS establishment, agriculture and logging activities heavily impacted this zone. No industrial facilities exist within it. Forest and wildlife management activities include traditional practices designed to enhance wildlife habitat for game species, such as frequent burning, maintenance of food plots, thinning of pine stands, creation of edge habitat, and protection of mast-producing oaks. The Savannah River swamp and the Lower Three Runs corridor are designated as separate zones. Resource management objectives are primarily wetland protection, access control, and minimization of contaminated sediment movement. Frequent flooding and wet soils limit access. Although logging impacted these areas prior to 1951 and reactor operations after 1951, limited timber harvesting or silviculture still occurs. Management activities that occur often include restoration programs, such as the Pen Branch restoration project (see chapter 3). The Department of Energy Set-Aside Program is implemented through designated land-use areas that cover about 5,665 ha (14,000 ac) in multiple parcels. Activities are restricted to nonmanipulative research and monitoring (Davis and Janecek 1997). A wide range of land uses, including logging, impacted the individual areas prior to 1951, but most have suffered relatively minimal disturbance since that period. The setaside areas cover a range of ecological conditions. They include unique ecological areas such as Carolina bays and major stream systems (e.g., Upper Three Runs and Meyers Branch), as well as old fields and experimental sites. The SRS began selecting set-aside areas in the 1950s for protection from land management. In addition to meeting research and monitoring objectives, these areas provide habitat for a number of sensitive plants and animals. The streams and wetlands frequently provide baseline data on metals, radioactive elements, and organic compounds on noncontaminated sites and serve as reference areas for assessing biological impacts from industrial facilities. Identification of SRS land-use area objectives and boundaries, as well as evaluation of activities compatible with those objectives, is a continually evolving process. Land management objectives must not compromise the evolving missions of the Site. In addition, land management activities on site, as elsewhere, are subject to applicable federal laws and regulations governing land use. While these varied objectives and constraints present a challenge to land management on SRS, they are designed to allow for compatibility between the primary SRS missions and the responsible stewardship of the vast natural resources of the site.

2

r

The Physical Environment Climate and Air Quality John I. Blake, Charles H. Hunter, Jr., and Bruce A. Bayle

Soils and Geology Randall K. Kolka, Gary Sick, and Bobby McGee

Water Resources Randall K. Kolka, Cliff G. Jones, Bobby McGee, and Eric A. Nelson

Climate, soils and topography, and water constitute the physical environment of any area and serve as the natural framework in which terrestrial and aquatic plants and animals must function. In the first section of this chapter, “Climate and Air Quality,” John Blake et al. describe the climate of the SRS as humid subtropical, with a mean annual temperature of 18°C (64°F) and a mean annual precipitation of 1,225 mm (48.2 in). They present trends and ranges in precipitation, temperature, and relative humidity and discuss conditions that create inversions or fog or affect atmospheric stability. Lightning, wind, storms, and other disturbances, as well as acid deposition and ground-level ozone concentrations, are considered for their impacts on natural resources. In the second section, Randall Kolka et al. describe the soils and geology of the SRS. The topography ranges from gently rolling to flat, and elevation ranges from 20 m to 130 m (66–427 ft) above sea level. Soils on the uplands are sandy. Intensive farming prior to SRS establishment significantly impacted 19

20


the soils of the area. Seven soil associations occur on the SRS: ChastainTawcaw-Shellbluff, Rembert-Hornsville, Blanton-Lakeland, Fuquay-BlantonDothan, Orangeburg, Vaucluse-Ailey, and Troup-Pickney-Lucy. The authors discuss the physical and chemical properties of upland and wetland soils and describe soil restoration and watershed maintenance efforts at SRS. Kolka et al. then describe the streams, waterbodies, and groundwater resources of the SRS in the third section, “Water Resources.” Wetlands and aquatic systems occupy more than 20 percent of the SRS, and nearly all drain to the Savannah River. After discussing general stream hydrology and chemistry, the authors describe characteristics of the major streams, impoundments, and isolated wetlands. Streams include the Upper Three Runs–Tinker Creek system, Beaver Dam Creek, Fourmile Branch, Pen Branch, and the Steel Creek–Meyers Branch system. Major impoundments include Par Pond and L Lake. Isolated wetlands include 343 Carolina bays and depression wetlands.

Climate and Air Quality John I. Blake, Charles H. Hunter, Jr., and Bruce A. Bayle The Savannah River Site (SRS) conducts intensive meteorological data collection, climate analysis, and modeling of the atmospheric transport of air pollutants to support safety, public health, and facilities design (Hunter 1999). This information and expertise provide resource managers and scientists with unparalleled capability. Climate is defined as the statistical weather characteristics of an area. These include such variables as average precipitation, frequency of extreme events, diurnal temperature range, number of frost-free days, and storm occurrence. In combination with soils and topography, climate establishes the natural environmental framework for both terrestrial and aquatic plants and animals. Resource management practices—whether aimed at conservation, restoration, research, exotic pest management, stream monitoring, harvesting, road construction, fire management, or erosion control—must take climatic factors into consideration when evaluating alternatives. Climate influences the physical range or geographical limit of many plant and animal species and associated communities (Barbour and Billings 1988). In part, those limits determined the natural assemblages of plants and animals prior to European settlement and now determine those species that can potentially be restored, managed, and sustained at the SRS. For example, slash pine (see appendix for scientific names) is no

The Physical Environment

21

longer planted here because the natural northern limit for that species is south of the SRS. While it survived well initially, experience over the last fifty years has shown it to be more susceptible to ice storms and stem diseases than native species such as loblolly, longleaf, and shortleaf pine. In contrast, although the SRS is at the edge of the range of the American alligator and the gopher tortoise, it is within their historical ranges and should be able to support viable populations of those species.

General Description The climate of the SRS is humid subtropical with a mean annual temperature of 18°C (64°F) and a mean annual precipitation of 1,225 mm (48.2 in). Geographical position heavily influences the climate. The SRS is approximately 160 km (100 mi) from the Atlantic Coast and a similar distance from the mountains. It is south and east of the Appalachian Mountains, which provide protection from colder and drier polar air masses that penetrate the region. As a result, the SRS rarely experiences snow or icing conditions compared with areas farther north. Because the “Bermuda high” pressure (Atlantic subtropical anticyclone) system generally weakens in the fall and winter, air masses that dominate during that period are drier and cooler. During the summer and early fall, because of the proximity of the SRS to the coast, the persistent Bermuda high dominates weather conditions, and temperatures are often greater than 32°C (90°F) with high humidity. These warm, humid conditions result in frequent afternoon thunderstorms, lightning, and occasional tornadoes or tropical storms.

Precipitation Important precipitation variables include monthly mean precipitation and extremes, the maximum precipitation and its recurrence interval, and the spatial variability and distribution of precipitation events. While mean annual precipitation is approximately 1,225 mm (48.2 in), extreme droughts occurred in 1954 (732 mm, or 28.8 in) and 2001–2002 (915 mm, or 36.0 in), and extreme wet years occurred in 1964 (1,866 mm, or 73.5 in) and 1972 (1,625 mm, or 64.0 in). According to monthly precipitation characteristics from 1952 to 2001 (table 2.1), precipitation tends to be distributed somewhat evenly throughout the year at SRS, but April, May, October, and November are typically drier than other months. In contrast, coastal regions in the South tend to have peak precipitation

22


Table 2.1 Mean monthly rainfall (depth in equivalent mm) and extremes for the 773-A area at the Savannah River Site for the period 1952–2001 Max Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

Min

Mean

Depth

Year

Depth

111.8 110.5 124.0 82.6 93.7 115.8 130.8 123.7 103.6 73.9 66.3 88.1 1225

254.5 202.4 278.4 208.3 276.8 291.6 291.6 313.4 221.2 498.3 197.6 242.6 1866

1978 1995 1980 1961 1976 1973 1982 1964 1959 1990 1992 1981 1964

22.6 15.5 23.1 14.5 5.1 22.6 22.9 26.4 12.0 0.0 5.3 11.7 732

Year 1981 2000 1995 1972 2000 1990 1980 1963 1985 1963/2000 1958 1955 1954

Source: Savannah River Technology Center, Atmospheric Technologies Group.

periods during the summer and somewhat drier winters. The variability in monthly precipitation is equally important. Maximum precipitation is 2 to 2.5 times the mean, whereas the minimum is as little as 10 to 20 percent of the mean. October has a maximum of 498 mm (19.6 in) and a minimum of 0 mm. On average, about seventy-six rain events each year have precipitation above 2 mm (0.08 in), which is the estimated canopy retention capacity for forest vegetation surfaces. Precipitation events below that amount will generally not rewet the soil. Events greater than 20 mm (0.8 in) in a twenty-four-hour period are fairly common, occurring an average of twenty times per year, and rain greater than 50 mm (2.0 in) can be expected at least once a year. The SRS region has experienced at least one one-hundred-year precipitation event in the last fifty years (table 2.2). Precipitation events of more than 100 mm (3.9 in) per twenty-four-hour period can be expected every five to ten years. Spatial variability in precipitation is also an important climate characteristic. Spatial gradients of precipitation often exist in mountainous terrain. At SRS, no such gradient is evident, either in total rainfall or seasonal means, but there is tremendous spatial variation in precipitation from individual storms. For example, over a seven-year period, measure-


23

Table 2.2 Predicted extreme precipitation recurrence estimates (mm) by accumulation period and observed extreme total precipitation received in the Savannah River Site region, August 1948–December 1995 Accumulation period

Predicted extremes Recurrence interval (yrs) 10 25 50 100 1000 Observed extremes Columbia, SC Augusta, GA SRS

3-hour

6-hour

24-hour

48-hour

83.8 101.6 116.8 129.5 188.0

91.4 111.8 127.0 144.8 210.8

127.0 154.9 175.3 198.1 292.1

165.1 200.7 218.4 238.8 NAa

127.8 108.0 132.1

134.4 114.3 147.3

194.6 217.7 187.7

NA 283.2 259.1

Source: Savannah River Technology Center, Atmospheric Technologies Group. a Not available.

ments at gauging stations on SRS had a mean precipitation of 6.8 mm (0.3 in) and an average standard deviation among locations of 4.4 mm (0.2 in), or approximately two thirds of the mean. Distance between locations and season determines reliability of individual storm precipitation measurements. Variability among measurements was less than 10 percent for observations within 0.5 to 1 km of each other during summer, but about 60 percent for measurements 2 to 5 km apart. In winter, comparable values were less than 5 percent and about 20 percent.

Temperature and Humidity The SRS experiences a range in average temperature, extreme high and extreme low temperatures, and corresponding minimum, maximum, and average humidity (table 2.3). Data over a thirty-seven-year period indicate climate ranges, but greater extremes are expected over longer periods of record. From June through September, extreme high temperature can exceed 40°C (104°F). Below-freezing temperatures can occur from late October through early April, but extreme low temperatures of –19.4°C (–3°F) and –13.9°C (7°F ) were observed in January and December, respectively. The pattern of humidity indicates the shift in air masses that

24


Table 2.3 Ranges for monthly mean, monthly high, and monthly low temperature and monthly mean, maximum, and minimum relative humidity, 1964–2001, from A Area at the Savannah River Site Temperature range in °C Month January February March April May June July August September October November December

Average

High

1.7–12.8 4.4–12.8 10.0–18.3 15.6–21.1 18.3–26.7 23.9–29.4 26.7–29.4 23.9–29.4 21.1–26.7 15.6–23.9 10.0–18.3 4.4–12.8

13.3–30.0 16.6–30.0 23.3–32.2 28.3–37.2 31.1–37.2 33.9–40.6 33.9–41.6 32.2–41.6 32.2–40.0 27.8–35.6 22.2–31.7 18.9–27.8

Low (–19.4)–0.6 (–11.1)–0.0 (–11.6)–4.4 (–1.7)–7.2 3.3–12.2 8.9–21.1 14.4–21.1 13.3–20.6 4.4–18.3 (–2.2)–7.8 (–7.8)–3.9 (–13.9)–(–1.7)

Percent humidity Average 70 65 71 56 63 75 75 78 78 74 70 70

Minimum Maximum 51 44 40 36 40 44 47 50 48 45 46 48

86 84 86 88 93 95 96 97 93 90 87 91

Source: Savannah River Technology Center, Atmospheric Technologies Group.

dominate the SRS over the year. Humidity is generally highest in midsummer; less humid periods in the spring and fall correspond to months with lower than average precipitation.

Evapotranspiration and Soil Water Deficits Evapotranspiration (ET) represents the combined amount of water lost to the atmosphere by surface evaporation from soils and transpiration from vegetation surfaces. It is important for terrestrial vegetation because it represents the amount of water that precipitation must replace during a given period to prevent a soil water deficit. A deficit period during a drought can reduce growth and cause mortality. This relationship is especially important for vegetation with poorly established roots systems, such as newly planted seedlings, or vegetation growing on very sandy soils with limited water-holding capacity. ET also regulates the seasonal cycle of wetting and drying in isolated wetlands, bottomland forests, and swamps. As it increases, soils dry and open water is reduced. The average monthly and daily open-pan evaporation for Blackville, South Carolina, provides an index of the potential ET (table 2.4). Total annual pan evaporation is approximately 1,448 mm (57 in), or slightly


25

Table 2.4 Historical average pan evaporation at the Edisto Experiment Station, Blackville, South Carolina, 1963–1992 Pan evaporation (mm) Month January February March April May June July August September October November December

By month

By day

48 67 112 151 174 188 191 164 132 103 67 51

1.53 2.39 3.62 5.02 5.62 6.27 6.16 5.28 4.41 3.33 2.24 1.64

Note: No pan coefficient (e.g., 0.8) adjustment has been applied to these values.

more than the average rainfall of 1,225 mm (48.2 in). Pan evaporation in summer is two to three times the winter rate. In SRS forests, actual ET follows calculated ET, using the Priestly-Taylor method (Rebel 2004). Combining actual ET by closed forest vegetation with precipitation records and the water-holding capacity of a typical soil (Dothan) yields an estimate of the average soil water deficit at SRS. Although rainfall tends to be uniformly distributed throughout the year, the actual ET from April through September results in a seasonal deficit of about 485 mm (19 in). Measurable deficits also occur in the fall during extreme droughts such as in October and November of 2001.

Atmospheric Stability, Inversions, and Fog Atmospheric stability and the formation of inversions help predict dispersion of smoke from prescribed burning and occasional wildfires. The combination of fog and smoke produces extremely limited visibility and hazardous conditions on public roads. Augusta, Georgia had heavy fog about thirty days per year from 1951 to 1995. Fog occurred about three days per month in the fall and winter. Stable atmospheric conditions occur when the temperature change with elevation in the atmosphere is more than 0.1°C per 30 m (more than 1.8°F per 100 ft), with little wind.

26


Unstable conditions occur when the temperature change is less than –0.5°C per 30 m (less than 0.9°F per 100 ft), in moderate winds. The latter conditions favor the transport and dispersion of pollutants, like smoke, away from populated areas. Inversions occur when temperature increases with height above the ground, preventing mixing and dispersion of smoke or other emissions. Elevation at the base of the inversion in the atmosphere defines the mixing height. This elevation varies diurnally and seasonally. Annually, the average mixing height in the early morning is about 380 m (1,246 ft) and in the afternoon is about 1,500 m (4,921 ft). The elevation of the afternoon mixing height ranges from 1,100 m (3,609 ft) in winter to 2,000 m (6,562 ft) in summer, on average. In Georgia and South Carolina, inversions within 457 m (1,500 ft) of the ground happen on about 70 percent of the nights during the year (Langley and Marter 1973). This observation has been confirmed by meteorological studies with data from towers at SRS (Lavadas 1997).

Lightning, Wind, and Disturbance Natural disturbances linked to climatic phenomena are important factors controlling the ecology and management of natural resources (Pickett and White 1985). Design of the recovery efforts for the endangered redcockaded woodpecker must consider the occurrence of these events (Hooper and McAdie 1995). Wildfires, lightning, hail, tornadoes, hurricanes, strong winds, ice glazing, flooding, and catastrophic insect outbreaks all create early successional or open habitat. These events cause vegetation mortality that allows other plant and animal species to develop. Many species native to SRS depend on natural or anthropogenic disturbances to sustain their populations, most notably fire-maintained savanna communities (see chapter 4). Storm phenomena occur in every month at SRS (table 2.5). Thunderstorms are the most frequent recurring event, with a peak in midsummer. Lightning, high rainfall, hail, and strong winds disturb forested and wetland areas. Lightning frequently kills one or more trees directly, leading to bark beetle attacks that cause additional tree mortality (Outcalt 1999). These spots provide gaps in the forest canopy. Approximately 0.04 lightning strikes per hectare per year occur at SRS and cause mortality in mature longleaf pine of 0.2 trees per hectare per year, a significant mortality source in rotations of 80 to 120 years. High winds not associated with tornadoes or hurricanes can also cause significant blowdown and top


27

Table 2.5 Monthly occurrences of tornadoes, hurricanes, thunderstorms, and snow or ice in the Savannah River Site (SRS) region

Month January February March April May June July August September October November December Total

Tornadoes (number)a 14 17 21 26 27 14 13 18 8 7 27 8 200

Hurricanes (number)b 0 0 0 0 0 1 2 11 18 4 0 0 36

Thunderstorms (days)c 0.8 1.7 2.6 3.9 6.3 9.7 13.1 10.0 3.5 1.3 0.8 0.7 54.4

Snow/ice (max depth, mm)d 66 (1992) 356 (1973) 28 (1980) 0 0 0 0 0 0 0 trace (1968) 25 (1993)

Source: Savannah River Technology Center, Atmospheric Technologies Group. a Includes all tornadoes (F-0 to F-5) in a 2-degree square centered on SRS, 1951–1996; 95 percent were between F-0 and F-2. b Includes all hurricanes observed in South Carolina, 1700–1992. c Average days per month for Augusta, Georgia, 1951–1995. d Maximum depth of snow and ice pellets observed in Augusta, Georgia, 1951–1995.

breakage in trees. In November 1995, high winds blew down the estimated equivalent of one million board feet of trees at SRS. Damage was most severe on old-field–planted pines, whose agricultural plow layer restricted taproot development (Kormanik, Sung, and Zarnoch 1998). Hail damage can also be significant. A hailstorm in 2002 seriously damaged a large stand of trees on SRS near Jackson, South Carolina. Nine tornadoes have been recorded in close proximity to SRS since operations began in 1951. In 1989, a tornado swept through the southern portion of SRS, creating a path 26 km (16 mi) long and destroying almost 500 ha (1,235 ac) of forest. Four F-2 tornadoes struck forested areas during March 1991. Tornadoes occur in all months, but most are in March, April, May, and November. Hurricanes also cause significant damage in South Carolina, but because the SRS is inland, winds associated with tropical weather systems usually diminish below hurricane force before reaching SRS. While snowfall is uncommon, ice glazing does occur, and the

28


loading on branches and stems causes significant damage to tree crowns. The weight of 1 cm (0.4 in) of ice, which has a recurrence frequency of ten to twenty-five years, is enough to cause severe tree breakage. Ice glaze leaves a legacy of forked trees and bent stems.

Air Quality Air quality can impact natural resources, including soils, vegetation, aquatic organisms, and surface waters at SRS. Of the six criteria pollutants established by the U.S. Environmental Protection Agency (EPA) under the Clean Air Act, four are of primary concern to natural resource managers because of their effect on terrestrial and aquatic ecosystems: sulfur dioxide (SO2 ), nitrogen oxides (NOx), ozone (O3), and particulate matter (PM). Air quality in the Central Savannah River Area is generally considered acceptable with respect to current levels and trends of these four pollutants. From 1995 to 2000, South Carolina and Georgia state monitors registered almost no exceedances of the National Ambient Air Quality Standards (NAAQS) within or near the SRS (Aiken, Allendale, and Barnwell Counties, South Carolina; and Richmond County, Georgia). The one exception was in Richmond County, which exceeded the NAAQS for the O3 one-hour standard once in 1995 (U.S. Environmental Protection Agency 2001a).

Acid Deposition Acidifying compounds that are suspended in the atmosphere, such as sulfates (SO4 ), nitrates (NO3), and ammonium (NH4), can be deposited on forests through precipitation and fog or simply by settling out of the atmosphere. Sulfur dioxide (SO2 ) and nitrogen oxides (NOx), the precursors to those acidic compounds, are emitted primarily from coal-fired power plants, industry, and the transportation sector. Acid deposition can adversely impact both soils and streams. Soils with a low buffering capacity can exhibit a depletion of calcium, magnesium, and potassium, a decrease in pH, and an increase in available aluminum (National Science and Technology Council 1998). Depletion of macronutrients can lead to growth reduction in vegetation. Soil pH levels below 4.5 can increase aluminum uptake by plants. Both low pH and aluminum are toxic to aquatic organisms and vegetation. Despite the negative impacts of NO3 deposition, that compound can benefit vegetation by providing nitrogen, an essential nutrient whose absence often limits vegetation growth


29

in southeastern soils. The natural geologic parent materials, warm temperature, high rainfall, and long period of soil development in the Carolinas results in highly leached soils. In general, the upland soils of the SRS have low to moderate buffering capacity, naturally low cation exchange capacity (CEC), low pH, and low base saturation (calcium and magnesium as a percent of CEC), making them somewhat susceptible to acidification. Compared to most of the eastern United States, only moderate levels of sulfate and nitrate deposition occur within the upper coastal plain of South Carolina (National Atmospheric Deposition Program 2001). Emissions of SO2 and NOx, precursors to acidic compounds, originate primarily outside of South Carolina. Those precursors are transformed through atmospheric chemical reactions into secondary pollutants, sulfate and nitrate (SO4 and NO3), which can travel great distances from their point of origin. The SRS lies approximately midway between the closest monitoring sites at Santee National Wildlife Refuge, Clarendon County, South Carolina, and Bellville, Evans County, Georgia. The majority (43–49 percent) of both annual nitrate (7.9 kg per ha) and annual sulfate (10.7 kg per ha) deposition from rainfall occurs during summer (National Atmospheric Deposition Program 2001), coinciding with the peak for electric power demand. Sulfate deposition has been decreasing over the SRS area since approximately 1990 as a result of changes in the Clean Air Act. Nitrate deposition has remained constant since approximately 1990.

Ground-Level Ozone Ozone is a gaseous pollutant formed by a reaction between NOx and volatile organic compounds (VOC) in the presence of sunlight and warm temperatures. Ozone levels typically follow a diurnal pattern, being lowest during late night and early morning hours and highest during the afternoon. The trend in rural O3 has remained fairly constant since 1989 (U.S. Environmental Protection Agency 2001a), and the trend for the near future is expected to remain constant. NOx levels from vehicles and utilities have remained relatively constant over the past decade, and anthropogenic VOC levels have declined by 20 percent. The overwhelming majority of VOC emissions is from trees. Damage to sensitive vegetation occurs when hourly O3 exposures during the growing season are frequently above 0.05 ppm or when there are numerous hours in which O3 concentrations are greater than or equal to 0.10 ppm. However, other environmental conditions must be favorable

30


for O3 to enter a leaf. Soil moisture is perhaps the most important factor affecting O3 uptake by plants. During drought conditions, plants close their leaf stomata cells and prevent O3 from entering. Thus, damage to sensitive plants caused by O3 occurs most often during years with plentiful rainfall. The U.S. Forest Service Southern Forest Health Monitoring program annually assesses impacts of O3 to sensitive species in South Carolina. Visible symptoms do occur to varying degrees on sensitive species within the SRS. However, current research has been unable to establish that ambient O3 exposures are causing growth losses. Visible symptoms of O3 do not necessarily damage an individual species or the forest. Not all plant species have the same likelihood of damage when exposed to similar amounts of O3. For example, black cherry may suffer damage at lower O3 exposures than tulip poplar. Loblolly pine is normally considered very resistant to O3. At concentrations the SRS currently receives, we do not expect loblolly pine growth loss.

Soils and Geology Randall K. Kolka, Gary Sick, and Bobby McGee Knowledge of the soils and geology of the Savannah River Site (SRS) is critical to understanding the ecology and management of its natural resources. The geology of the site determines the topography and influences the water quality in various aquifers, as well as the transport of water, organic materials, ions, and metals to streams (Cooke 1936; Siple 1967). It is also the primary parent material for soil development. In combination with other environmental factors, soil properties influence vegetation composition and productivity (Row 1960; Whipple, Wellman, and Good 1981; Jones, Van Lear, and Cox 1984; Thompson and Lloyd 1995; Duncan and Peet 1996; Smith 2000); root disease (Witcher and Lane 1980); animal habitat; and management activities. Activities include planting and seeding, harvesting, vegetation control, fertilization, restoration, road construction, and prescribed burning (Hatcher 1957; Aydelott 1971; Wells et al. 1979; Bush et al. 1995). Various soil properties and topographic position affect the transport of nutrients, sediment, organic matter, and chemicals to wetlands and streams via surface runoff and subsurface flow (Dosskey and Bertsch 1994; Bush et al. 1995). The characteristic “blackwater” chemistry of SRS streams and isolated wet-


31

lands is a consequence of the soil chemical and transport properties within the drainages (Meyer 1986).

Land-Use Impacts to Soil before 1951 Soils vary naturally across the SRS because of differences in parent material and the effects of topography, climate, organisms (plants and animals), and time ( Jenny 1941). However, a major factor affecting the soils is the recent human-induced landscape disturbance since European settlement, including a wide range of agriculture and related activities (Cabak and Inkrot 1997; White and Gaines 2000; White 2004). With the exception of narrow terraces and floodplains, SRS soils are generally poorly suited for productive farming. On the uplands, the sandy texture and low organic matter, nutrient status, pH, and cation exchange capacity limit productivity (Rogers 1990; Brooks and Crass 1991). Significant portions of the SRS are cypress-tupelo swamp and too wet for cultivation. Early farmers often practiced a form of slash-and-burn agriculture on the uplands, abandoning cleared forests after a few years (Cabak and Inkrot 1997). To improve productivity, farmers applied fertilizers such as lime, phosphates, and nitrogen to cash crops like cotton and corn. The residual effect of these practices was enhanced subsequent pine tree growth (Bennett 1956). However, because of the soil’s physical and chemical properties, surface erosion, and the reduction in organic matter during cultivation (e.g., Smith 2000), additional nutrients have been leached away or transported to subsurface clays (Odum, Pinder, and Christiansen 1982). On many sites, a residual plow layer at shallow depths of 0.5 to 0.8 m (1.6–2.6 ft) is compacted sufficiently to impede tree taproot development and enhance windthrow (P. Kormanik, U.S. Forest Service, pers. comm.). At SRS, erosion and stream sedimentation were not as serious as in the Piedmont and mountains. Caved stream banks and erosion channels, formed from concentrated runoff, contributed to stream sedimentation prior to SRS establishment (U.S. Department of Agriculture 1951), and they are still evident.

General Physiography and Geology The northern part of SRS is located within the Aiken Plateau of the Sandhills physiographic province. The southern part is within the Pleistocene coastal terraces of the Upper Coastal Plain (see figure 1.1). The topography is gently rolling to flat, and elevation ranges from 20 to 130 m (66 to 427

32


Figure 2.1. Topographic relief on the Savannah River Site. MSL = mean sea level.

ft) above sea level (figure 2.1). The Aiken Plateau has sandy soils deeply incised by stream channels. These soils range in age from 10 million to 50 million years. Slopes range from gentle (1–2 percent) to moderately steep (30–40 percent). Some upland areas, as well as bottomlands along the major streams, are nearly level. Strongly sloping areas are adjacent to major drainage ways and their headwaters. The Pleistocene coastal terraces are flat to gently rolling; the Brandywine, Sunderland, and Wicomico terraces that generally parallel the Savannah River represent successive recessions


33

in sea level from ten thousand to one million years ago (Langley and Marter 1973). The oldest, Brandywine Terrace, is adjacent to the Aiken Plateau at 50 to 80 m (164–262 ft) elevation. The Sunderland Terrace lies between the Brandywine and Wicomico terraces at elevations ranging from 30 to 50 m (98–164 ft). The Wicomico comprises the current floodplain of the Savannah River at 20 to 30 m (66–98 ft) above sea level. Flat-bedded sedimentary deposits of Paleozoic, Mesozoic, and Cenozoic geologic formations underlie the SRS, but only Cenozoic-aged (Tertiary) deposits are at or near the surface (figure 2.2; Prowell 1996). The Tertiary formations are important aquifers for environmental management and environmental restoration activities, potable water supplies, and stream hydrology (Williams and Pinder 1990; Prowell 1996). Of these formations, only Paleocene deposits (65–54 million years BP) do not reach the surface at SRS, ranging from 30 to more than 90 m (98 to more than 295 ft) deep. Eocene deposits (54–38 million years BP) include the Huber and Congaree Formation (undivided), the McBean Formation, and the Barnwell Group. The Huber formation is fine to coarse, poorly sorted quartz in a matrix of white kaolin clay (Prowell 1996), whereas the Congaree formation is fine to coarse, moderately to well-sorted quartz sand in a matrix of gray kaolin clays. The Huber and Congaree Formation is approximately 15 to 30 m (49–98 ft) thick and is only at or near the surface along Upper Three Runs (Prowell 1996). The McBean Formation lies above the Huber and Congaree Formation with the basal part of the formation containing white to buff sandy limestone, calcareous sand, and dark olive-green marl deposits. Above the basal sediment, the formation contains moderately to well-sorted quartz sand and gravel. The McBean Formation is approximately 10 to 45 m (33–148 ft) thick and, like the Huber and Congaree Formation, is only at or near the surface along the Upper Three Runs drainage (Prowell 1996). The Barnwell Group, consisting of the Clinchfield Formation, Dry Branch Formation, and Tobacco Road Sand, lies above the McBean Formation. The Clinchfield Formation consists of sandy to calcareous clayey sand and is not present at the surface on the SRS. It ranges from 0 to 15 m (0–49 ft) thick and underlies other sedimentary deposits in the southeastern area of the SRS. The Dry Branch Formation contains well-sorted calcareous clays, kaolin clays, and sands and ranges from 0 to 30 m (0–98 ft) thick across the SRS, getting thicker toward the southeast. It is at or near the surface in all major SRS drainages (Prowell 1996). The Tobacco Road Sand consists of poorly to moderately sorted fine to very coarse sand and is 0 to 15 m (0–49 ft) thick, also becoming thicker toward the southeast.

Figure 2.2. Geological stratigraphy and groundwater systems of the Savannah River Site (U.S. Department of Energy 1998).

It is at or near the surface along all major drainages and extends into the uplands of the Aiken Plateau and Brandywine Terrace. Above the Eocene deposits are deposits of the Miocene (23–5 million years BP), Pliocene (5–1.8 million years BP) and Holocene periods (11,000 years BP to present; figure 2.2). Deposited during the Miocene, the Upland Unit has beds of gravel and poorly sorted sands, fine to coarse sand containing clays, and brightly colored (red, orange, yellow, or purple) massive sandy clay. From 0 to 25 m (82 ft) thick, the Upland unit is at or near the surface on most upland areas of the SRS, including portions of the Aiken Plateau and the Brandywine Terrace. Above the Upland Unit is dune sand of Pliocene age. The dune sands are eolian deposits consisting of moderately sorted medium sands devoid of clays. Dune sand deposits up to 10 m (33 ft) thick dot the landscape of SRS, becoming


35

more defined in the northeast part of the site. Two alluvial deposits comprise the Holocene deposits on site. Both are fine to coarse sands intermixed in a sparse clay matrix and range in thickness from 0 to about 15 m (49 ft). The older, more highly weathered alluvial deposit is associated with the Sunderland Terrace. The younger deposit, which occurs on the Wicomico Terrace, is the current floodplain of the Savannah River.

Soil Associations Approximately fifty distinct soil types exist on the SRS. Soil scientists often combine them to develop general soil associations that show broad areas of distinctive soils, relief, and drainage. Typically, a soil association consists of one or more major soil types and several minor soils. Such associations (figure 2.3, in color insert) can be used to compare the suitability of large areas for general land uses. Seven soil associations occur on the SRS (Rogers 1990). The first two associations comprise most of the floodplain, wetlands, and bottomlands along stream terraces and the Savannah River. The last five comprise the associations of the upland Sandhills and Coastal Plain.

Chastain-Tawcaw-Shellbluff Association Located on floodplains along the larger streams, Chastain-TawcawShellbluff soils occur in depressions and remnant sloughs from old stream channels. Slopes are generally 0 to 1 percent. The association includes approximately 60 percent Chastain, 20 percent Tawcaw, 15 percent Shellbluff, and 5 percent other soils. Chastain soils are poorly drained and clayey to a depth of about 1 m (3.3 ft). Tawcaw soils are somewhat poorly drained, clayey in the upper part, and loamy in the lower profile. Shellbluff soils are well drained and loamy to a depth of about 1 m. This association constitutes about 6 percent of the SRS.

Rembert-Hornsville Association Rembert-Hornsville soils are located on stream terraces adjacent to floodplains. Slopes are generally 0 to 2 percent. The association has approximately 30 percent Rembert, 18 percent Hornsville, and 52 percent other soils. Rembert soils are poorly drained, and Hornsville soils are moderately well drained. This association constitutes about 7 percent of the SRS.

36


Blanton-Lakeland Association Blanton-Lakeland soils are located on the broad stream divides in the uplands. Slopes are generally 0 to 10 percent. The association has approximately 40 percent Blanton, 20 percent Lakeland, and 40 percent other soils. Blanton soils are somewhat excessively drained, have sandy surface and subsurfaces, and loamy subsoil 1 to 2 m (3–7 ft) below the surface. Lakeland soils are excessively drained and are sandy throughout. This association constitutes about 18 percent of the SRS.

Fuquay-Blanton-Dothan Association Fuquay-Blanton-Dothan soils are located on the broad upland ridges. Slopes are generally 0 to 10 percent. The association has approximately 20 percent Fuquay, 20 percent Blanton (see above), 12 percent Dothan, and 48 percent other soils. The soils are generally well drained to excessively well drained. Fuquay soils are well drained with a sandy surface and loamy subsoil that contains iron-rich nodules of plinthite. Dothan soils are well drained with loamy subsoil that contains iron-rich nodules of plinthite. This association constitutes about 47 percent of the SRS.

Orangeburg Association Orangeburg soils also are located on the broad upland ridges. Slopes are generally 0 to 10 percent. The association includes roughly 70 percent Orangeburg soils and 30 percent other soils. Orangeburg soils are well drained with friable loamy subsoil. These soils generally have the highest clay content and cation exchange capacity among the upland soils (Looney et al. 1990). This association constitutes about 2 percent of the SRS.

Vaucluse-Ailey Association Vaucluse-Ailey soils are located on the uplands in scattered areas around the head and sides of small drainage ways. Slopes are generally 6 to 15 percent. The association includes approximately 25 percent Vaucluse, 15 percent Ailey, and 60 percent other soils. Vaucluse soils have loamy surface and subsurface layers of less than 50 cm (20 in). Ailey soils have a moderately thick sandy surface and subsurface layer. Both soils are well drained and have a loamy subsoil with a brittle layer. This association constitutes about 10 percent of the SRS.


37

Troup-Pickney-Lucy Association Troup-Pickney-Lucy soils are located on the steep slope uplands and on the floodplains along streams. Slopes range from 0 to 40 percent. The association has approximately 45 percent Troup, 10 percent Pickney, 10 percent Lucy, and 5 percent other soils. Troup soils are well drained, with a thick, sandy surface and loamy subsoil 1 to 2 m (3–7 ft) below the surface. Pickney soils are poorly drained with a thick black surface soil and are sandy throughout. Lucy soils are well drained with a moderately thick, sandy surface and subsurface layer, with loamy subsoil at a depth of 0.5 to 1 m (1.6–3.3 ft). This association constitutes about 10 percent of the SRS.

Soil Characteristics Because of the extent and importance of the wetlands on the SRS, we consider the distinct characteristics of wetland and upland soils.

Wetland Soils Wetland soils include typically flooded Fluvaquents along the Savannah River and deltas, as well as true organic soils (Medisaprists) in some of the Carolina bays and bottomland hardwood systems (Rogers 1990). The SRS contains approximately 16,000 ha (40,000 ac) of wetland or hydric soils. Carolina bays range in size from less than 0.1 ha (0.04 ac) to more than 50 ha (124 ac). Bay surface soils generally grade from well-drained sands on the exterior to sandy loams in the wetland center. A loamy to clayey horizon commonly underlies the sandy upper horizons. Bays with long hydroperiods may develop variable layers of peat on the surface (Schalles et al. 1989). Soils in bottomlands range from sandy well-drained soils at floodplain edges and in upper areas of watersheds to very poorly drained loamy and clayey soils in floodplains and stream deltas. Bottomlands with extended hydroperiods, commonly found in deltas and in some floodplains, may develop organic upper horizons of variable depth. Generally, past disturbance and the degree of flooding, inundation, and saturation of the soil dictate plant distribution and abundance more than physical or chemical properties (Whipple, Wellman, and Good 1981; De Steven and Toner 1997). On about 11,170 ha (28,000 ac) of the SRS, flooding is a hazard. Ecological (Whipple, Wellman, and Good 1981) and wetland (De Steven and Toner 1997) restoration studies have documented wetland soil properties at SRS. Dixon et al. (1997) and Schalles et al. (1989) measured

38


baseline chemical characteristics both on and off SRS. Wetland soils are highly variable, but the variables tend to align with major soil groups. In the surface, soil organic carbon ranges from 0.5 to 9.0 percent and pH ranges from 4 to 7.6. Nutrient status is generally high except in Carolina bays; for example, phosphorus ranges from 84 to 830 µg/g. Cation exchange capacity and the percent base saturation are also high, with the former ranging from 2 to 194 meq/100g. Nearly all chemical variables decrease with depth, except pH and in Carolina bays.

Upland Soils Most upland soils are Paleudults that are highly weathered, nutrient poor, and typically sandy. These soils are well drained to excessively well drained with a sandy surface layer of variable thickness over a loamy to clayey subsoil. The productivity of planted pines at SRS is directly related to the depth to the subsurface clay layer and the thickness of the surface organic horizon (Row 1960). In general, this relationship is also true for old-field vegetation (Odum 1960; Thompson and Lloyd 1995) and native communities ( Jones, Van Lear, and Cox 1984; Smith 2000). While water availability in the root zone is undoubtedly important, nutrition, particularly nitrogen, is the primary factor limiting growth on these sites (Birk 1983; McKee et al. 1986; Davis and Corey 1989; Allen, Albaugh, and Johnsen 2002). Several studies have measured physical properties of the upland soils. Surface soil texture ranges from 80 to 90 percent sand, with less than 5 percent silt and the balance in clay size particles (Odum 1960; Nutter 1979; Odum et al. 1982; Looney at al. 1990). At the lower limit of rooting for most trees, 1.5 to 2 m (4.9–6.6 ft), the clay content typically increases to 35 percent or more, with an additional 10 to 15 percent silt content, depending on the soil series (Nutter 1979). Bulk densities are normally low in the surface (1.1–1.4 g/cc) and rapidly increase to a depth of 1 m (1.6–1.8 g/cc; Davis and Corey 1989). Clays are primarily kaolinite (70–95 percent), with lesser amounts of vermiculite (1–30 percent) and minor quantities of illite (Looney et al. 1990). These soil textures generally result in very low water-holding capacity for plants during the summer. From the surface to a depth of 2 m (6.6 ft), the amount of water available to plants varies from 100 to 300 mm (3.9–11.8 in). As much variation in chemical properties exists within a given soil series as between soil series. In part, this variation comes from prior landuse history (Smith 2000), succession (Odum, Pinder, and Christiansen 1982), and amendments applied (Birk 1983; Davis and Corey 1989). In


39

contrast to wetland soils, surface mineral soil has very low organic matter, which varies from less than 1.0 percent to 4.0 percent (Odum 1960; Smith 2000), or from less than 0.1 percent to 1.45 percent organic carbon (Looney et al. 1990). Similarly, pH ranges from around 4 to 6.5 at the surface with little relationship to soil series; the maximum values are less than the maximum pH in wetland soils. On the whole, cation exchange capacity (1.2–4.5 meq/100g), base saturation (0.12–0.88 percent), and most major nutrients are lower (table 2.6) than in wetland soils. However, phosphorus is high for forest sites, particularly in the subsoil, presumably due to prior agricultural fertilization (Nutter 1979; Odum et al. 1982).

Soil Restoration and Watershed Maintenance Since only 1.3 percent of the SRS drains into the Salkehatchie basin, the impacts of soil and watershed management activities at SRS largely affect the Savannah River. Watershed degradation in industrial areas is generally caused by unhealthy or poor vegetation cover (from improper maintenance) or by impermeable surfaces (rooftops, paved and gravel areas, sidewalks, and compacted areas), both of which increase storm water runoff. A forest with a fully developed overstory allows almost no storm water runoff (Williams and Pinder 1990) or sediment transport (Yoho 1980; Patric, Evans, and Helvey 1984). Replacement of forests with lawns, utility corridors, and roadsides doubles the amount of runoff. If such areas are converted to impermeable surfaces, runoff is about eleven times as great (Natural Resource Conservation Service 1986). Natural channels that carry runoff from the watershed can be overtaxed and fail. In forested areas, poor vegetation cover follows disturbances such as road construction, timber harvest, prescribed burning, and vegetation control. These activities can lead to short-term increases in stream flow, sediment, and nutrient transport (Binkley and Brown 1993). The SRS uses best management practices to minimize the impact of normal forest management activities (National Council for Air and Stream Improvement 1994). Mitigative harvesting practices such as helicopter logging, low ground pressure equipment, and restricting harvests during wet weather reduce rut formation, compaction, and erosion. The SRS has left buffer strips along streams since the mid-1970s and has reduced mechanical site preparation (see chapter 3). Managers have routinely seeded secondary roads to grass, and water bars have been installed since the 1970s (Swift 1988). The area in stabilized log-loading decks and forest roads is minimal relative to industrial sites.

Loamy sand

Loamy sand Loamy sand

115–170

0–25

Sandy clay loam

90–114

26–75 77–257

Sandy clay loam

Sandy clay loam

54–89

Loamy sand

Sandy clay loam

Sandy clay loam

118–165

0–18

Sandy clay loam

75–117

19–53

Loamy sand

Sandy clay loam

0–30

Sandy clay loam

118–152

31–74

Sandy clay loam

82–117

Source: Nutter 1979. a CEC = cation exchange capacity.

Troup

Norfolk

Fuquay

Loamy sand

Sandy clay loam

Dothan

0–23

Series

Soil texture

24–81

Horizon depth (cm)

5.2 5.2

4.9

5.3

5.6

5.6

5.4

5.3

5.0

4.7

4.3

4.5

5.3

5.4

5.3

5.0

pH

0.15 0.12

1.15

0.09

0.20

0.30

0.40

0.65

0.20

0.25

0.38

0.76

0.18

0.21

0.40

1.45

Organic matter (%)

70 85

20

200

175

175

164

82

205

170

138

120

241

252

200

140

P

18 19

10

90

268

287

227

90

178

225

123

101

131

180

150

128

Ca

15 18

3

77

97

70

53

12

79

53

36

11

44

65

55

14

Mg

Total (µg/g)

Table 2.6 Chemical characteristics of selected upland soils, by depth, on the Savannah River Site

15 17

14

15

22

23

27

25

15

36

28

16

16

20

26

16

K

1.2 1.4

1.7

3.3

2.6

3.1

3.1

1.6

3.3

3.2

2.0

1.5

3.2

4.5

3.0

2.0

0.32 0.34

0.12

0.38

0.88

0.69

0.53

0.48

0.47

0.51

0.50

0.44

0.33

0.38

0.63

0.34

23 22

41

20

18

21

24

30

25

18

23

18

35

23

30

47

Total CECa % base Exchangeable (meq/100g) saturation Al (µg/g)

40 Ecology and Management of a Forested Landscape


41

From a watershed perspective, forest and protected areas contribute a relatively minor amount of surface runoff and sediment compared to developed areas (National Council for Air and Stream Improvement 1994). In 1973, severely disturbed sites such as gravel pits, spoil piles, and borrow pits from construction occupied about 809 ha (2,000 ac) on SRS (Beavers et al. 1973). Although direct sedimentation impacts were localized, there were concerns that these poorly vegetated sites could cause additional erosion, channel degradation, and deposition in low areas (Aydelott 1971). Therefore, most early soil conservation efforts were aimed at revegetating these sites (Hollod and Christensen 1983). Establishment and growth of vegetation on these sites was inadequate unless the sites were ripped to reduce compaction and organic amendments such as municipal waste were applied (Berry and Marx 1980). In 1991, the restoration program shifted its focus to the legacy impacts of SRS operation and of facility developments where off-site storm water effects could not be controlled. After 1992, due to new regulations, new construction and development projects incorporated measures to manage storm water runoff, surface drainage, and vegetation cover along roads. Vegetation maintenance activities include liming, fertilization, seeding, mowing, and aeration. Annually, SRS restores about 60 ha (150 ac) and maintains about 457 ha (1,129 ac) of disturbed sites. Because of the need for rapid establishment of cover, SRS uses quick-germinating annual grasses (e.g., ryegrass, browntop millet; see the appendix for scientific names) in a mixture of seeds that includes warm-season perennials (e.g., Bahia grass, Bermuda grass) and legumes (e.g., Trifolium clovers, vetches, partridge pea). Efforts are currently underway to evaluate various native grasses.

Water Resources Randall K. Kolka, Cliff G. Jones, Bobby McGee, and Eric A. Nelson The Savannah River Site (SRS) has freshwater resources that support rich aquatic communities. Over 20 percent of the SRS consists of wetlands and aquatic systems (Bowers et al. 1997). With the exception of the far northeastern corner that drains to the Salkehatchie River (1.3 percent of SRS), all surface water on the SRS drains to the Savannah River (figure 2.4). The Savannah River watershed drains approximately 27,000 km2 (10,420 mi2), including western South Carolina, eastern Georgia, and a

42


Figure 2.4. Major streams, wetlands, and larger lakes of the Savannah River Site.

small portion of southwestern North Carolina. The SRS makes up about 3 percent of the Savannah River watershed. In its middle and lower portions, where a 27-km (17-mi) reach forms the SRS boundary, the Savannah River is broad with extensive floodplain swamps and numerous tributaries. Several groundwater aquifers exist in the Tertiary and Upper Cretaceous sedimentary deposits that underlay the site. Groundwater is an important source of flow to riparian zones, streams, and wetland depressions on SRS. Five major stream systems originate on or pass through


43

the SRS (see figure 2.4). A smaller stream, Beaver Dam Creek, also originates on the SRS and ultimately drains to the Savannah River. Two large cooling water reservoirs, Par Pond and L Lake, are present in the southeastern portion of the site. A number of small artificial ponds ranging in surface area from 0.1 to 1.2 ha (0.04–0.5 ac) exist on the SRS. Some were farm ponds, constructed when the area was still under private ownership, whereas others resulted from cooling water canals or other SRS construction projects. More than three hundred isolated Carolina bays and wetland depressions (see chapter 3) exist at SRS, as well as numerous beaver ponds (Fitzgerald 1979; Snodgrass 1997). Bottomland hardwood and swamp wetlands are associated with the floodplains of the Savannah River and its tributaries.

Historical Impacts Prior to 1951 The historical impacts of post–European settlement land use on the natural streams, swamps, and isolated wetlands of the Southeast have been extensive (Mulholland and Lenat 1992). To promote navigation, several of the larger SRS streams, like Upper Three Runs, were cleared of debris in the 1850s. During the same period, numerous dams were constructed along major streams, including Upper Three Runs, Lower Three Runs, Fourmile Branch, Pen Branch, and Tinker Creek, to power sawmills and gristmills (Brooks and Crass 1991). Remnants of those dams still exist on SRS. About 162 ha (400 ac) of artificial ponds were present in 1951, generally at the headwaters of major tributaries. Logging occurred along all of the major streams and throughout the Savannah River swamp (Fetters 1990). The primary impacts of logging were altered drainage patterns that resulted from the construction of access roads, rail lines, and haul back lines, which often blocked natural channels or created new channels. Fur trapping reduced beaver populations and thereby the number of beaver ponds ( Jenkins and Provost 1964). Corralling of livestock adjacent to streams (e.g., Pen Branch) and farming in the uplands often resulted in loss of stream bank integrity and an increase in overland flow and sedimentation (Trimble 1975). Meyers Branch was dredged during the 1940s to improve drainage of the basin in an attempt to reduce habitat for malarial mosquitoes. Information about chemical inputs to these systems prior to 1951 is scarce, but DDT was used for mosquito control, arsenate insecticide was used for boll weevil control, and mercury from manufacturing facilities was discharged into the Savannah River (Gladden et al. 1985).

44


Isolated wetlands were also impacted. Although the exact number is unknown, many Carolina bays were drained and farmed (Kirkman at al. 1996). Drainage reduced the hydroperiod of the bays, and farming altered the soil and vegetation. One of the largest impacts on the hydrology of SRS resulted from construction of the Strom Thurmond Reservoir on the Savannah River above Augusta, Georgia. The reservoir stabilized the flow of the river and altered the frequency, timing, depth, and duration of natural flooding of the swamp (Sharitz and Lee 1986).

Groundwater Groundwater resources of the SRS are described in detail by Clarke and West (1997, 1998). The SRS is underlain by the Atlantic Coastal Plain Hydrogeologic Province, which includes three major aquifer systems and three confining units, all underlain by the Appleton Confining System (see figure 2.2). The Appleton Confining System separates the Atlantic Coastal Plain Hydrogeologic Province from the underlying Piedmont Hydrogeologic Province. The Atlantic Coastal Plain Hydrogeologic Province consists of unconsolidated sediments of Late Cretaceous and Tertiary origin. The uppermost unconfined Floridan Aquifer System includes the Steed Pond, Upper Three Runs, and Gordon Aquifers and ranges from 0 to 40 m (0–131 ft) below the surface of SRS (U.S. Department of Energy 1999). It is the primary source of groundwater contributing to streams (U.S. Department of Energy 1999; Arnett and Mamatey 1996). All three Floridan Aquifers are recharged directly through precipitation where they are at or near the surface and through leakage from both underlying and overlying aquifers. The Floridan Aquifer System is generally not a source of domestic or production water on the SRS because deeper aquifers provide a more abundant supply of higher-quality water (U.S. Department of Energy 1995b). Below the Floridan Aquifer System is the Crouch Branch Confining Unit of the Meyers Branch Confining System that separates the above aquifers from the Crouch Branch Aquifer of the Dublin Aquifer System (see figure 2.2). Recharge of the Crouch Branch Aquifer is from both overlying and underlying aquifers. Groundwater from the Crouch Branch Aquifer is pumped for both domestic and industrial uses, although the majority of groundwater used on the SRS is from the deeper McQueen Branch Aquifer (U.S. Department of Energy 1995b).


45

The McQueen Branch Confining Unit of the Allendale Confining System separates the Crouch Branch Aquifer from the McQueen Branch Aquifer (figure 2.2). Recharge of the McQueen Branch Aquifer is mainly from the overlying Crouch Branch Aquifer. The McQueen Branch Aquifer is the main production aquifer of the SRS. The SRS withdraws approximately 14.0 billion liters (3.7 billion gal) per year of groundwater for domestic and industrial uses (U.S. Department of Energy 1995b). Groundwater quality is strongly influenced by the mineralogy of aquifer geology. In general, groundwater from the aquifers on the SRS is low in dissolved solids and in some areas low in pH, which results in high corrosivity and dissolved iron concentrations (Arnett, Karapatakis, and Mamatey 1993). Solvents, metals, radionuclides, nutrients, and other chemicals generated during SRS operations contaminate groundwater at 5 to 10 percent of the site (U.S. Department of Energy 1995b). Generally, the contaminated areas underlie or are in the vicinity of facilities, and contaminants are the result of facility processing. Contamination is restricted to Floridan Aquifer System waters, with one exception where trichloroethylene and tetrachloroethylene have contaminated the Crouch Branch and McQueen Branch Aquifers in the northwest portion of the site (U.S. Department of Energy 1995b).

Streams The hydrology and chemistry of streams on SRS have important influences on stream ecology, and each stream has had a unique history.

Stream Hydrology The depths of perennial streams vary from less than 1 meter to several meters during nonstorm periods. SRS streams have numerous pools but few riffles. In general, they are shallow and have low gradients and very low perennial and intermittent drainage densities (see table 2.7). Low stream slopes are typical of the geomorphic nature of the Lower Coastal Plain and Sandhills region. Ephemeral streams (estimated at 5–25 percent of drainage) are not frequent at SRS due to its sandy soils and low relief. They occur primarily in low-lying areas or along steeper slopes associated with Vaucluse soils that border major streams. Flow rate, depth, width, submerged woody debris, sediment, and sunlight have significant effects on habitat and associated biological communities both on and off SRS

63,792 50.4

Upper Three Runs

1.22 None 0.64 0.28 0.67 0.42 2.0–3.7 2.9

13,342 77.3

Tinker Creek

7.64 1955–1985 0.69 0.29 0.93 0.41 1.3–34.0 15.3

5,865 100

Fourmile Branch

3.07 1954–1988 0.65 0.34 1.02 0.44 4.2–26.8 14.5

5,663 100

Pen Branch

Beaver Dam Creek 2,200 100 2.78 1952–1982 NAc NA 2.30 NA 27.1–64.2 35.3

Steel Creek 9,339 100 1.63 (6.10)a 1954–1968 0.47 0.13 1.70 0.46 3.4–17.0 11.2

0.50 (7.15)a 1958–1964 0.56 0.06 2.31 0.38 4.8–32.0 13.0

32.8

47,326

Lower Three Runs

(1993–2000). Beaver Dam Creek (1976–2000) is still affected by power generating operations. f Ratio of mean annual stream flow to annual precipitation over area for period of record.

Source: U.S. Geol. Surv. Reps. for South Carolina: http://sc.water.usgs.gov. Note: The last four variables were calculated using data from the lowest gauge on each watershed and thus do not represent the entire watershed. a Includes area impacted by L Lake, Par Pond, and Ponds 2, 5, B, and C. b Perennial and intermittent stream segments only. c Not available. d Perennial segments only. Portion below L Lake dam for Steel Creek and below Par Pond dam for Lower Three Runs. e Upper and Lower Three Runs (1975–2000), Steel Creek (1994–2000), Pen Branch (1993–2000), Fourmile Branch (1987–2000), and Tinker Creek.

2.25 None 0.32 0.17 Mean stream slope (%)d 6.82 Mean annual flow (m3/sec)e 0.37 Precipitation runoff ratiof Range of peak discharge (m3/sec) 12.1–56.6 Median peak discharge (m3/sec) 21.4

Industrialized area (% SRS) Thermal impacts (yrs) Mean density (km/km2)b

Watershed area (ha) Amount of watershed on SRS (%)

Hydrologic characteristic

Table 2.7 Hydrologic characteristics of major streams on the Savannah River Site


47

(e.g., Meffe and Sheldon 1988; Smock and Gilinsky 1992; Lakly and McArthur 2000; Fletcher et al. 2000). The percentage of the watershed area affected by industrial land use is greatest in the Fourmile Branch, Pen Branch, and Beaver Dam Creek watersheds. Average precipitation-runoff ratios in SRS streams range from 0.37 to 0.46 (table 2.7) and are typical for the fall line and Upper Coastal Plain/Sandhills region of South Carolina (Smock and Gilinsky 1992). Peak discharge is generally lower in that region than in other areas of South Carolina, particularly the Piedmont and mountain regions (Guimares and Bohman 1992). Because of channel characteristics, bed sediment and organic debris are readily resuspended and transported during storm flows. Though mean flow is greatest during winter (figure 2.5), catastrophic flooding and hence annual peak runoff can occur in any month. On streams not affected by industrial process water discharge, mean monthly flows typically are greatest in March but decline abruptly in April as evapotranspiration increases. As evapotranspiration declines in the fall and surface soils are recharged, mean flow increases. During reactor operations, about 10 percent (174,000 gal/min, or 658,662 l/min) of the flow of the Savannah River was used for cooling water (Dukes 1984). The impacts of this pumping on the streams and ponds are complex (U.S. Department of Energy 1997; Arnett and Mamatey 1999). Water pumped from the Savannah River was initially used to cool P, R, L, C, and K Reactors and was discharged directly to streams. Subsequently, water was directed to cooling ponds, such as Par Pond, constructed for that purpose. Water was later used in a recirculation system. Currently, river water is pumped to maintain reservoir levels in Par Pond and L Lake, thereby reducing exposure of contaminated sediment and maintaining the ecological benefits of these larger bodies of water.

Stream Chemistry The geology of the SRS region strongly influences water chemistry. The sedimentary material through which water flows and eventually reaches the surface consists primarily of acidic silica sands and kaolinitic clays. Flowing waters at SRS are characterized as blackwater streams. They typically have high dissolved organic matter and acidity and low buffer capacity and nutrient concentrations (Meyer 1986) compared to Piedmont streams or major streams draining the Gulf Coast (Smock and Gilinsky 1992). Bowers et al. (1997) summarized extensive stream chemistry and temperature data from SRS.

48


Figure 2.5. Relative mean monthly discharge for major streams on the Savannah River Site: Upper Three Runs (UTR) and Lower Three Runs (LTR; 1975–2000); Tinker Creek (TC; 1993–1995, 1999–2000); Fourmile Branch (FM; 1987–2000); Pen Branch (PB; 1993–2000); Steel Creek (SC; 1993–2000); and Beaver Dam Creek (BDC; 1976–2000) (U.S. Geological Survey Reports. for South Carolina, http://sc.water.usgs.gov).

As in most stream systems, dissolved O2 decreases in periods of low flow and increased water temperature during summer and fall (table 2.8). Dissolved organic carbon, dissolved solids, and pH fall within normal ranges for blackwater streams (table 2.8; Smock and Gilinsky 1992). Phosphorous levels are lower in SRS streams than in the Savannah River. The alkalinity of SRS systems is typically lower than in whitewater systems, and sulfate provides the major anion contribution to the system. Because bicarbonate and calcium are low, heavy metals can be more toxic to certain organisms in these systems (Specht and Paller 1995). Elevated heavy metal concentrations have been observed in only a few locations close to contaminated seeps (Bowers et al. 1997).

Major Streams and Watersheds Major streams that drain the SRS include the Upper Three Runs–Tinker Creek system, Fourmile Branch, the Steel Creek–Meyers Branch system, Pen Branch, Lower Three Runs, and Beaver Dam Creek.

0.05–1.1

0.05–4.5 0.2–0.7

Total sulfate (mg/l)

Total chloride (mg/l) Total calcium (mg/l)

1–13.8 1.2–2.3

1–5.6

NDi–0.16 0.3–5.6

1.0–12

5.0–12.5

17.0–40 2.0–6.0 1.0–69

5.2–8.0

7.1–24.4

Upper Three Runsb

3.3–10.0

2.4–10.7 2.6–4.2

3.0–12.9 1.0–6.4

4.0–19.0

4.7–11.0 3.0–8.1

4.4–13.2 ND–0.18

6.3–14.8

2.0–170 10.0–21.0 2.8–28

6.0–8.6

18

Pen Branchd

NA ND–0.13

6.4–12.7

30.0–135 6.0–17.0 50% drained by ditching selected native wetland vegetation Break down water-holding Probably >20 small dams and ponds on streams at SRS structures to reestablish

Restoration strategies

Examples specific to SRS

Table 3.4 General ecological impacts from post-European settlement in the Central Savannah River Area and strategies for ecological restoration

86 Ecology and Management of a Forested Landscape

SRS Forest Management

87

(chapter 4). Although the approximate area and distribution of hardwood stands is similar to that of pre-European settlement (Frost 1997), human activities have drastically altered the composition. Colonial settlers preferred oaks for barrels, furniture, and fuelwood. As a result, oaks declined to the point that they are no longer dominant in most hardwood stands on SRS. Certain red oak (e.g., Q. falcata) and hickory (e.g., Carya pallida) species sustained less dramatic reductions. Hardwood stands currently occur along stream corridors, near the Savannah River swamp on nontillable soils, and on sandhill soils too poor to farm. Upland stands often occur in small isolated remnants that escaped fire, in fencerows, and in stringers leading from stream corridors through the uplands. The current management objective is to maintain the existing percentage of land area in hardwood, mixed pine-hardwood, and bottomland swamp forest stands. Some area reduction within the primary red-cockaded woodpecker recovery zone (see chapter 1) may be offset by increases in the other zones through conversion of old-field pine to mixed hardwood-pine on mesic or wet soils. Restoration goals are to improve the quality of the species mixture in stands, particularly increasing white and red oaks, dogwood, holly, and other species that are soft fruit–producing. Experimental planting of various hardwood species in existing hardwood and old-field stands has occurred since the mid-1960s. However, poor stock quality, competition, and inappropriate site selections limited success. Since 1993, the SRS has planted 157 ha (388 ac) of hardwood on moist sites with a mixture of cherrybark oak, swamp chestnut oak, willow oak, green ash, white oak, and sycamore (figure 3.8a). Limited seed and seedling availability is a major constraint to planting more white and red oaks on suitable sites. Methods for restoring hardwoods include harvesting, usually clear-cutting small blocks of either pine or previously high-graded hardwood stands, followed by site preparation, which may include burning and herbicides. Enrichment planting of various oaks and other species is followed by competition release using mechanical cutting or spot treatment with herbicide. Natural regeneration of oaks is often unreliable due to previous removals, irregular acorn crops, and high acorn consumption by animals. The SRS developed a cooperative seed orchard to help supply southern red and white oak seed. Because size and root development of the bare-root stock is critical to long-term survival and growth (Kormanik, Sung, and Kormanik 1994), nursery managers carefully select the stock. Root competition and shading from overstory trees result in poor growth of species planted in the

Figure 3.8. Locations of restoration projects on the Savannah River Site: (a) mixed hardwood stands restored since 1993; (b) Carolina bays with restoration activity since 1989 (bays not labeled are the nineteen bays in the mitigation bank); (c) red-cockaded woodpecker habitat restored since 1983 by midstory removal and prescribed burning; and (d) sites selected for establishment of savanna plant populations in old-field pine stands. RCW = red-cockaded woodpecker (U.S. Forest Service, unpublished data).


89

understory (R. H. Jones, Virginia Polytechnic Institute, unpublished data), so planting in newly cleared areas free of competition is preferable.

Bottomlands and Riparian Zones: Pen Branch The SRS has a “no net loss” wetlands policy and a wetlands banking program to mitigate potential loss of wetlands on the site. Several wetland mitigation projects involving the creation, restoration, or enhancement of wetlands have been performed on SRS (Irwin et al. 1997). The Pen Branch restoration, required for the continued operation of K Reactor (U.S. Department of Energy 1991), exemplifies the mitigation process at SRS. The Savannah River swamp is a 3,020-ha (7,462-ac) forested wetland on the floodplain of the Savannah River at the SRS (see figure 2.4). Historically the swamp consisted of approximately 50 percent bald cypress–water tupelo stands, 40 percent mixed bottomland hardwood stands, and 10 percent shrub, marsh, and open water (Nelson, Dulohery et al. 2000). Major impacts to the swamp hydrology and vegetation occurred with the completion of nuclear production reactors in the early 1950s. Water was pumped from the Savannah River through secondary heat exchangers of the reactors and discharged into tributary streams that flowed into the swamp. From 1954 to 1988, SRS discharged hightemperature effluents in excess of 65° C (149°F) into one of the tributaries, Pen Branch, at rates often twenty to forty times greater than normal flow. The sustained increases in water volume resulted in overflow of the stream banks, erosion of the original stream corridor, and deposition of a deep silt layer at the confluence of Pen Branch and the river floodplain. The nearly continuous flooding of the swamp, the thermal load of the water, and the heavy silting resulted in complete mortality of the original vegetation in the Pen Branch corridor and in large areas of the river floodplain (figure 3.9). Once SRS reduced the pumping, natural reestablishment of early successional species like cattail, bulrush, buttonbush, pokeweed, blackberry, and black willow occurred in the affected areas. However, few volunteer seedlings of bottomland hardwoods or bald cypress were evident. Therefore, a mitigation action plan was formulated to guide the restoration of the degraded Pen Branch wetlands. The successful completion of the mitigation entails three strategies: (1) the rehabilitation of the Pen Branch corridor and delta by natural succession, (2) the reforestation of the corridor and delta by planting, and (3) the compensatory mitigation of other

90


Figure 3.9. Aerial view of the Pen Branch corridor and delta on the Savannah River Site during reactor operations (U.S. Forest Service files).

impacted areas on the SRS pending evaluation of the success of the first two approaches. From 1993 to 1995, the SRS planted approximately 75 percent of the affected Pen Branch floodplain area in bottomland hardwood tree species, keeping the remaining area (25 percent) unplanted for experimental purposes (figure 3.10). Three restoration approaches were formulated to address the differing conditions of the impacted floodplain. Approximately 8,700 seedlings were planted in the lower corridor (15 ha, or 37 ac) without any site preparation, and the delta (12 ha, or 30 ac) was planted after herbicide application in the absence of burning (figure 3.11). The upper corridor (24 ha, or 60 ac) was planted after the application of herbicide and a prescribed burn. Herbicide application and prescribed burning were performed to control a dense black willow overstory and to clear brush and vines from the planting area. Tree species included in the plantings were overcup oak, swamp chestnut oak, nuttall oak, willow oak, cherrybark oak, water hickory, persimmon, green ash, sycamore, swamp black gum, water tupelo, and bald cypress (Dulohery et al. 1995). While the stream structure and aquatic communities were not manipulated, the trees were expected to alter light, temperature, and organic debris (logs, leaf litter) in the stream favorably for fish and invertebrates. In


91

Figure 3.10. Degraded wetland areas of the Pen Branch corridor and delta on the Savannah River Site that were impacted by thermal releases from reactors and later restored as part of the mitigation effort.

addition, several areas of open water in the delta were left unplanted with cypress or tupelo to benefit wading birds, waterfowl, and alligators. The SRS developed an extensive research program to examine the restoration ecology of the Pen Branch system. A special edition of Ecological Engineering (Nelson, Kolka et al. 2000) outlines many of these studies. Tree seedling studies indicated that many site preparation techniques (burning, herbicides, thinning) did not significantly impact early growth or survival (Dulohery, Kolka, and McKevlin 2000). However, tree shelters

92


Figure 3.11. Planting trees in the Pen Branch corridor on the Savannah River Site, 1993 (U.S. Forest Service files).

and root pruning were effective silvicultural techniques that enhanced survivability in areas prone to stress from herbivory and competition (Conner, Inabinette, and Brantley 2000). A 1997 survey showed that water tupelo, green ash, sycamore, and persimmon had the highest survival in the upper corridor, while bald cypress survived best in the wetter lower corridor and river delta areas (Kolka et al. 1998). Although species abundance and, in some cases, diversity are higher in the Pen Branch floodplain than in the reference systems (table 3.5), the composition of plant and animal communities and key energy sources such as soil carbon and nutrients indicate that the Pen Branch floodplain remains an immature, early successional system but is moving toward recovery (Giese et al. 2000; Wigginton, Lockaby, and Trettin 2000). Pen Branch is currently functioning as a viable early successional wetland. Kolka et al. (2000, 2002) used measurements of hydrology, soils, vegetation, carbon and nutrient cycling, and animal communities to predict wetland function in response to the restoration. As a consequence, SRS wetland restoration research will serve as a template for future wetland restorations on site and elsewhere.


93

Table 3.5 Species richness, calculated as either total number of species observed or average number, for taxa in Pen Branch compared with disturbed post-thermal (20–30 years) and late successional forested reference sites at the Savannah River Site

vegetationa

Fourmile Branch Meyers Branch and Steel Creek and Tinker Creek

Pen Branch

Pen Branch

(unplanted)

(planted)

(post-thermal)

(reference)

81 2.35 22.4

79 2.86 NA

68 2.23 NA

63 2.96 17.6

44.4 8.2 21

44.2 9.7 18

NA 16.3 15

NA 19.1 12.5

Total Herbaceous speciesa Macroinvertebrate ordersb Herpetofaunal speciesc Avian speciesd Fish speciese a Giese

et al. 2000. and McArthur 2000. c Bowers et al. 2000. d Buffington et al. 1997. e Paller et al. 2000. b Lakly

Carolina Bays The SRS has several hundred Carolina bays or baylike depression wetlands, ranging from small (less than 0.1 ha or 0.25 ac) ephemeral bays to large (larger than 50 ha or 124 ac) bays that retain water for most of the year (chapter 2; Schalles et al. 1989). They serve as habitat for a wide range of rare plants and many vertebrates. The adjacent uplands also provide nesting sites for turtles and birds, as well as niches for facultative wetland plants. Although bays share some common plant and animal associates, the variability in composition between bays with similar soil, hydrology, and geomorphic conditions suggests that periodic rainfall, fire, and chance colonization also influence the observed flora and fauna (Greenberg and Tanner 2004). Predicting the restored structure and composition of the dominant vegetation of a disturbed bay is difficult, even using current topographic, soil, and hydroperiod conditions (De Steven and Toner 1997). In a specific restored bay, predicting the species of vertebrates and invertebrates, particularly rare or sensitive species, is even more difficult. Initial estimates by Kirkman et al. (1996), based on 1951 aerial photography, indicated that approximately two thirds of these isolated wetlands and nearly all of the associated uplands had been altered by human

94


Figure 3.12. A drainage ditch from a Carolina bay on the Savannah River Site. The bay is visible as the canopy opening in the background (U.S. Forest Service files).

activities such as draining (figure 3.12), farming, harvesting, and restriction of fire. However, beaver dams and other natural processes have closed the drainage ditches in some bays and natural recolonization has occurred without human intervention. Thus, the need for restoration is limited to those sites where the level of disturbance is such that recovery will not occur by natural processes alone. To identify sites effectively altered by drainage activities, SRS scientists recently considered information from geographic information databases, published reports, and field visits (table 3.6), in addition to the 1951 aerial photography (Kirkman et al. 1996). They determined that 195 (57 percent) of the 343 depression wetlands on SRS are not effectively drained. Nineteen bays were destroyed by construction activities in the early decades of Site operations. Of the remaining 129 bays, 4 were restored in the early 1990s, 16 are currently being restored, and another 3 are scheduled for restoration in 2006–2007. Field visits have yet to confirm the status of 92 bays with ditches evident in 1951. Prior to the initiation of restoration, the influence of residual overstory trees, burning, and soil disturbance on vegetation in bays was unknown. In December of 1989, three intact bays (Bays 56, 57, 58) were experi-


95

Table 3.6 Level of disturbance to surface hydrology by drainage ditches in isolated depression wetlands at the Savannah River Site in 2002 Status of hydrological disturbance 1951a

No ditch present in Ditch present in 1951, but no drainageb Ditch present in 1951, drainage confirmedc Ditch present in 1951, restoredd Ditch destroyede Ditch present in 1951, drainage status unknownf Total

Number

Percent

124 71 14 23 19 92 343

36.2 20.7 4.1 6.7 5.5 26.8 100

a No

evidence of drainage appeared in 1951 photograph, though some wetlands were probably farmed. b Ditches were filled through natural processes, or slope of drain was inadequate, for drainage. c Not all wetlands are potential restoration candidates due to proximity to site operations. d Includes four restored in the 1990s (Lost Lake, 170, 5119, and 93), sixteen restored in 2002, and three scheduled for 2006. e Destroyed in the early decades of SRS facility development and operations. fNot field checked to confirm condition.

mentally burned and tilled to test certain hypotheses. Soil tillage stimulated vegetation diversity, recruitment from the seed bank, and rare plant occurrence (Kirkman and Sharitz 1994). Active bay restoration (figure 3.8b) started with Lost Lake in the late 1980s and early 1990s (Halverson et al. 1997). Lost Lake is a bay impacted by the M Area waste retention basin overflow. Though farmers had previously drained Lost Lake, contamination from the basin required the bay to be redrained, the contaminated soil removed, and the area revegetated with native species. The hydrologic restoration was successful, but removal of soil probably had a detrimental effect; after restoration, reptiles have declined adjacent to the bay and non-native cattails have invaded (Halverson et al. 1997). Three drained bays (Bays 106, 170, and 5119) were restored in the early 1990s by harvesting the trees and plugging the ditches. However, for a variety of reasons (e.g., potentially limited seed bank, lack of soil disturbance, drought) few if any wetland plants naturally recolonized the areas, and the ditch plug on Bay 106 failed. In 1994, the drainage ditch of Bay 93 was closed, half of the wetland was harvested, and half of each portion (harvest/nonharvest) was burned. After four years, both harvesting

1.3

3.0 0.4

Harvest Harvest and burn

2.3

NS NS

15.1 15.0

1.3

NSb NS

Species richness

Wetland speciesa

1994

a

45.1 42.5

0

0

Wetland speciesa

Source: J. Singer, Savannah River Ecology Laboratory, unpublished data. Percent obligate and facultative wetland plants. b NS = not sampled.

1.4

Control

Species richness

Burn

Treatment

1993

13.1 10.7

1.0

1.2

Species richness

1995

57.7 61.8

0

0

Wetland speciesa

5.9 4.0

2.1

3.8

Species richness

1998

74.8 61.1

NS

NS

Wetland speciesa

Table 3.7 Effects of burning, harvesting, and harvesting plus burning on the average herbaceous species richness and percent wetland species occurring in Bay 93 on the Savannah River Site before and after closing the drainage ditch in 1994


97

and harvesting-plus-burning treatments increased wetland plant species richness (table 3.7; Singer 2002). In the late 1990s, an experimental approach was developed to restore several bays in conjunction with a wetland mitigation banking program. In 1997, SRS established a wetland mitigation bank to compensate for unavoidable wetland losses from future authorized construction and environmental restoration (U.S. Department of Energy 1997). The bank will not only hasten mitigation efforts with respect to regulatory requirements and implementation, but also will provide on-site and fully functional mitigation in advance of impacts. Using information and techniques from previous SRS work (as outlined above), researchers and managers identified nineteen Carolina bays in the nonindustrialized management area of SRS as candidates for restoration (see figure 3.8b). All nineteen bays possessed an active drainage ditch and a vegetation composition characteristic of a disturbed wetland system. Of the nineteen bays, sixteen (totaling approximately 20 ha, or 49 ac) were restored in 2001 by plugging the ditches and altering the vegetation. The remaining three bays serve as nonrestored controls in the interim. Undisturbed bays of similar size were used as reference sites. Several alternatives for restoring bays and adjacent uplands are being compared in a factorial design. On the SRS, two principal upland habitats commonly occur with Carolina bays: fire-managed, open-canopy pine savannas and relatively unmanaged, unburned, closed-canopy mixed pine-hardwood forests. To gain a better understanding of the relationship between buffer zone management and wetland properties, these two upland management alternatives are being examined as longterm goals. Bay-margin treatments were applied to a 100-m (328-ft) radius, from bay rim into the upland (figure 3.13). With each of these two upland alternatives, the bays were organized such that two wetland vegetation types (herbaceous and forested) were established, thus creating four bay-margin community combinations. Approximately 10 percent of the interior of herbaceous bays was planted with obligate wetland grasses (Panicum hemitomon and Leersia hexandra). The remaining area was not planted, but natural succession was encouraged through soil scarification. Forested bays were planted throughout their interior with swamp tupelo and bald cypress. Planting was initially successful, and most of the bays exhibited an increased hydroperiod during the first year of recovery compared to the control and reference systems. By 2002, however, all of the study sites,

a

b

Figure 3.13. Aerial view of restored Carolina bays on the Savannah River Site with (a) a mixed pine-hardwood margin (unthinned) and (b) a pine savanna margin (thinned), 2001 (Westinghouse Savannah River Co.).

including reference wetlands, had dried in response to a severe regional drought. Nevertheless, several species of amphibians, birds, and bats continued to respond positively to the treatments. Monitoring of biotic and abiotic conditions will record progress for five years, 2002–2006, to determine the final net improvement for each wetland.


99

Savannas Frost (1997) estimated that fire-maintained savanna communities historically occupied 80 percent of SRS uplands. Grass and herbaceous species originally dominated these communities, which had a pine overstory with scattered fire-tolerant hardwoods (e.g., Q. incana, Q. stellata, Q. marilandica). National programs are conserving and restoring these communities for their tremendous species richness of plants, as many as one hundred species per 0.1 ha (0.25 ac; E. W. Kjellmark, P. D. McMillian, and R. K. Peet, University of North Carolina, unpublished data). In addition, savannas provide habitat for several vertebrate species of concern in South Carolina. These include the gopher tortoise (see tables 4.20 and 4.22 for scientific names not given), gopher frog, pine snake, southern hognose snake, Bachman’s sparrow, northern bobwhite (Colinus virginianus), prairie warbler, and red-cockaded woodpecker. These communities depend on frequent fires to maintain the vegetation complexes. In 1951, many relict savanna plants occurred only along roadsides and in isolated woodlots (W. Batson, University of South Carolina, pers. comm.). Many vertebrate species persisted in clear-cut or heavily thinned stands that simulate the understory vegetation structure of native savannas (Krementz and Christie 1999). In 2001, the Department of Energy approved a plan to restore the gopher tortoise, and approximately one hundred tortoises were reintroduced on SRS (see chapter 4). Through the 1980s, forestry activities indirectly facilitated restoration and recovery of the savanna communities. The area of longleaf pine more than doubled, and over 12,141 ha (30,000 ac) of scrub oak received stem injection to release seeded or natural longleaf. In 1977, the prescribed burning program was greatly expanded to reduce fuel loading. Managers removed undesirable midstory hardwoods with chemical and mechanical treatments to improve red-cockaded woodpecker habitat (see figure 3.8c). The combined effects of harvesting and burning resulted in favorable conditions for savanna flora and fauna (Harrington and Edwards 1999; Johannsen 1998). In 1991, in systematic surveys of the upland pine forests, botanists identified state- and federally listed plant populations (chapter 5). In 1992, managers integrated red-cockaded woodpecker recovery with restoration of the savanna system as a whole. Research has established the composition of the pre-European landscape and general distribution of fire savannas (Frost 1997); the land-use history at SRS (White and Gaines 2000); and the distribution of savanna plant communities with respect to soil, topography, hydrology, and

Figure 3.14. Distribution of remnant and degraded savanna plant communities in relation to land-use and fire exclusion history, mapped for potential savanna restoration on a representative section of the Savannah River Site (C. Frost, The Nature Conservancy, unpublished data).


101

Table 3.8 Savanna grasses, composites, and legumes selected for experimental introduction to old-field pine sites at the Savannah River Site to establish founder populations Species Andropogon tenarius Anthaenantia villosa Aristida beyrichiana Aristida purpurascens Aster concolor Aster tortifolius Baptisia lanceolata Baptisia perfoliata Berlandiera pumila Carphephorus bellidifolius Chrysopsis gossypina Coreopsis major Desmodium strictum Eriogonum tomentosum Eupatorium album Eupatorium cuniformis Eupatorium curtsii

Species Galactia macreei Lespedeza hirta Liatris elegans Liatris secunda Liatris tenuifolia Nolina georgiana Petalostemum pinnatum Pityopsis graminifolia Polygonella americana Schizachyrium scoparius Silphium compositum Sorghastrum secunda Sporobolus junceus Stylisma patens Tephrosia florida Vernonia angustifolia

landform (Duncan and Peet 1996). The Nature Conservancy has helped map and classify fragments of the remnant savanna communities with respect to their restoration potential (figure 3.14). Current savanna restoration strategies consist of three components. First, prescribed burning is the key ecological process across the landscape, in conjunction with heavy thinning and midstory control, which stimulates grass and herbaceous species abundance. Second, after removing appropriate mid- and overstory trees, managers burn isolated fragments of intact remnant savanna communities ranging from less than one acre to several acres; this process will increase the abundance and flowering of the understory grass and herbaceous plants already present. Finally, managers and researchers have established local founder populations of rare or uncommon grass and herbaceous species (table 3.8) on old-field pine sites, which have poor seed banks after two hundred years of intensive agricultural use. These populations will ideally recolonize the landscape through dispersal to nearby areas where favorable establishment conditions have been created. Research is evaluating nursery procedures to grow approximately 150,000 individuals of thirty savanna species for

102


experimental transplanting to fourteen old-field pine sites (see figure 3.8d). These sites, which represent a range of fertility and moisture regimes, were heavily thinned and will be burned routinely. The research will assess demography and dispersal ability of each species. These studies will produce an operational plan to establish founder populations of rare and threatened species.

4

r

Biotic Communities Plant Communities Donald W. Imm and Kenneth W. McLeod

Aquatic Invertebrates Barbara E. Taylor

Butterflies Nick M. Haddad

Fishes Barton C. Marcy, Jr.

Amphibians and Reptiles Kurt A. Buhlmann, Tracey D. Tuberville, Yale Leiden, Travis J. Ryan, Sean Poppy, Christopher T. Winne, Judith L. Greene, Tony M. Mills, David E. Scott, and J. Whitfield Gibbons

Nongame Birds John C. Kilgo and A. Lawrence Bryan, Jr.

Nongame Mammals Susan C. Loeb, Lynn D. Wike, John J. Mayer, and Brent J. Danielson

103

104


The SRS supports a diverse array of plant and animal communities. This chapter describes those communities, exploring how they are arranged on the landscape and how the establishment of the SRS and its subsequent land management practices have affected each group. Most of the sections herein include a matrix of species-habitat relationships. Various workers have used many systems to classify SRS vegetation. In “Plant Communities,” Donald Imm and Kenneth McLeod describe how these ecosystems are distributed according to land-use histories and natural gradients of disturbance, topography, and soil features. They delineate seven major vegetation types: (1) remnant pine savannas and sandhill woodlands; (2) Carolina bays and other isolated wetland depressions; (3) upland pine forests; (4) upland hardwood and pine-hardwood forests; (5) bottomland and floodplain forests; (6) marshes; and (7) upland meadows, old fields, and industrial areas. Within each of these general vegetation types, they describe several subtypes, or communities, defined by ecological setting and dominant species. For example, Carolina bays and isolated wetland depressions may be wet savannas and meadows, dominated by herbaceous plants, or forested wetlands, dominated by various trees. Similarly, upland hardwood forests occur along stream drainages, around old-house sites, or in moist, fertile areas protected from frequent fire. Each setting supports characteristic species. Firemaintained upland pine savannas and sandhill woodlands are influenced by soil characteristics and fire frequency. A wide variety of forested floodplains, bottomlands, and swamps occur along major streams and the Savannah River, depending on hydrologic gradients and soil conditions. Some vegetation types are influenced more by past and current land management. Within certain ecological limits, a wide range of pine forest communities can develop on SRS uplands, depending on forest management practices such as planting, thinning (which affects density and species composition), timber rotation length, and prescribed burning. Vegetation in upland meadows, old fields, and industrial areas is influenced by seeding, planting, and mowing practices. Imm and McLeod provide detailed information on the species composition of the various communities that constitute each vegetation type. Aquatic invertebrates are abundant, productive, and rich in species in the streams, wetlands, impoundments, and other aquatic habitats of the SRS. Invertebrates play central roles in the functioning of those systems, processing both aquatic and terrestrial plant material and converting it to forms usable by many fish and other secondary consumers. They are potentially important in the transfer of contaminants, and their sensitivities to environ-

Biotic Communities

105

mental conditions make them useful in bioassessment protocols. Barbara Taylor describes the invertebrate communities of major aquatic habitats on SRS, including streams, impoundments, Carolina bays and isolated wetlands, and others. She discusses the effects of reactor operations (high water levels and temperatures) and other anthropogenic impacts on aquatic invertebrates. Finally, she outlines conservation concerns related to aquatic invertebrates. Little is known about most groups of terrestrial invertebrates on SRS (but see Van Pelt and Gentry 1985; Scheller 1988; Draney 1997). However, Nick Haddad provides a description of the butterflies known from the site, listing ninety-nine species that have been identified. He discusses their general habitat and relative abundance on SRS. In contrast, most vertebrate groups on SRS have been studied extensively. Barton Marcy lists the eighty-seven species of fish that have been collected on the SRS. In describing the major factors that affect fish distribution on the site, he considers stream order, depth, velocity, and other habitat features (e.g., water chemistry and substrate conditions), as well as such ecological factors as predation and competition. He then describes the composition of fish communities in each of the major SRS streams and water bodies. Kurt Buhlmann et al. list 103 species of reptiles and amphibians that have been documented on the SRS. The distributions of some of these species on SRS are limited because the site is on the edge of their range. Others are difficult to sample because of secretive habits. The distribution and abundance of most, however, are affected by species-specific habitat requirements, wetland hydroperiod, landscape structure, historic land-use patterns, and natural and anthropogenic disturbance. The authors provide examples of species limited by each factor. Finally, they discuss broad historical trends in SRS herptile populations. John Kilgo and Lawrence Bryan note that 259 bird species have been recorded on the SRS. The relative abundance of individual species has been dramatically affected by the establishment of the Site. They describe factors affecting SRS bird distribution, including season (as many birds are migratory), habitat type and successional stage, and landscape structure. They conclude with a discussion of historical trends in bird populations on SRS, observing that the site now supports more forest-associated birds and fewer field-associated birds due to the conversion of agricultural fields to managed forest. Susan Loeb et al. list all fifty-four species of mammals that currently occupy or recently occupied the SRS. This section focuses only on nongame mammals. The authors note that season affects the SRS distribution of only a few bats. Most species are affected by habitat type, various physical factors (e.g., soil type, coarse woody debris), and landscape structure.

106


Plant Communities Donald W. Imm and Kenneth W. McLeod The Savannah River Site (SRS) is a predominantly forested tract that lies below the Piedmont and north of the Savannah River. It is in the Sandhill and Upper Coastal Plain physiographic regions. Most of the area occupied by the SRS was once used for agriculture and is now forested with mid- to late-successional plant communities. This section will describe the general vegetation types on the varied landscape of the SRS, identify the factors that regulate or influence the dynamics of each type, and discuss the general impact of current forest management practices and past land-use activities. First, we discuss environmental factors that influence plant distribution (and, hence, community composition); second, we describe previously reported classifications of vegetation cover types on SRS; and finally, we present general descriptions of SRS vegetation types and plant communities. Each broad type of vegetation is typically composed of predictable suites of species, though the range of variation within a type may encompass several distinct subtypes, or “communities.” The appendix lists scientific names and the occurrence of other species in SRS vegetation types. Nomenclature follows Radford, Ahles, and Bell (1968). Tables, the appendix, and interpretations are based on published and unpublished data collected from SRS, review of other SRS studies, and information from the South Carolina Department of Natural Resources (SCDNR) and the Georgia Department of Natural Resources (GADNR) heritage programs.

Factors Influencing Plant Distribution Landscape position, soil type, past land use and disturbance, biological interactions, and chance determine the suite and proportion of species within a given area. Competition between individuals creates growth constraints through resource depletion. Other biological interactions— such as seed movement, predation, herbivory, and nutrient cycling—also affect plant growth and distribution. All of these factors determine the attractiveness and suitability for both existing and potentially existing plants and indirectly set the template for a natural progression of plant communities. Physical attributes also influence plant productivity and the development of plant associations. Such attributes include resource availability, past and present land management, and natural disturbance.

Biotic Communities

107

Resource Conditions Resource conditions, both median and extreme, during any given year or during the lifetime of a plant influence its survival, productivity, and competitiveness. Because much of the SRS has sandy soil surfaces within the rooting zone, capacity for holding resources (moisture, nutrients) and concentrations of ions are low. These conditions result in a magnified influence of soil organic material on soil chemistry. Topographic aspect and steepness influence the diurnal amount and duration of available nutrients during the growing season. Besides competition, tolerance of extreme or catastrophic conditions probably best defines the composition of the forests of SRS. Additional important adaptations for the local flora include the ability to become established through resistance, resilience, recruitment from a seed bank, or invasion via migration.

Land Management Animal migration patterns and historical vegetation influence pollinator activity, seed dispersion, and seed amount across the SRS. These factors are particularly important on the SRS because of the long history of intensive agricultural use (see chapter 1). Most upland areas of the SRS experienced agricultural clearing and planting during the twentieth century. Outside of agricultural weeds and roadside remnants, few species are present in upland forests relative to their presettlement diversity. Some hard and soft mast species (particularly trees) have invaded upland pine stands and in the absence of fire have remained competitive. Another persistent vegetation pattern attributed to human activity is the continued presence of artifial corridors. These corridors include old fence lines, persistent windrows, house places, cemeteries, and wood lots. Many permanent meadow corridors have remnant seed sources of species associated with the longleaf pine savanna. In the absence of fire, many of these species have limited numbers of individuals and seed. Therefore, few have spread into adjacent abandoned old fields and pine plantations. Several remnant populations of species associated with longleaf pine savanna communities have remained persistent in dry upland hardwood forests. Again, having been unburned, few species have spread into adjacent forested areas. Limited air movement due to the density of these forests has also restricted the dispersion of many light-seeded species in forested areas. Forest management plays the role of directing or accelerating the natural progression of plant community development. Subtle differences in

108


conditions that affect ecosystem development can lead to different vegetation types. No single soil or landscape condition is solely suited to one particular vegetation type. Similarly, one particular vegetation type is seldom restricted to a single soil or landscape setting. Over much of the SRS, soil conditions are nearly equally suitable for a variety of sustainable pine and hardwood communities. Management activities such as selective removal of trees and saplings, burning, and planting can shift the successional direction toward or away from certain forested communities and vegetation types. The success of these management activities depends on the match of vegetation conditions to the landscape or soil conditions. Well-developed or well-suited biological communities are often either resistant or resilient to changes invoked by forest management. For example, the conversion of established upland hardwood forests on fertile soils to pine savannas requires periodic mechanical and chemical treatments coupled with sustained burning and plant-introduction programs. Similarly, during early years, conversion from a pine forest on a sandy soil to upland hardwood forest requires periodic removal of invasive pine along with fire protection and plant introduction.

Disturbance The most common disturbance types on the SRS are flooding, drought, wind, and fire. Each of these four disturbance mechanisms has different types and magnitudes of influence on plant communities, and each is more likely to occur at particular landscape settings across SRS. Each disturbance benefits different types of species and provides opportunities for ecosystem reorganization.

Flooding Flood conditions that affect vegetation include season, duration, and rate of water movement. Flood frequency, depth, sedimentation rate, and water quality factors (e.g., composition, chemistry) have long-term effects on plant community development. Few species are tolerant of recurring floods during the growing season, but most tolerate flooding and soil saturation during the dormant season. River flooding is greatest February through April but can occur via tropical storms from August to October. Stream flooding can occur throughout the year, particularly during the summer months due to heavy localized precipitation. Flooding affects soil characteristics and understory vegetation. Flowing water transports sediments, organic debris, and living plants. Fine sediments and

Biotic Communities

109

organic material are easily displaced but not easily deposited. Shallowly rooted plants are easily uprooted. Inflexible plants, such as saplings and shrubs, incur greater damage from water flow. Floodwater also causes chemical changes. Upon flooding, soil oxygen dissipates within a few days, resulting in a decline of aerobic processes such as root respiration and bacterial activity. Most species will enter a period of dormancy, but some have specialized adaptations that allow for continued activity. Continued flooding can kill less adapted individuals and species.

Drought The SRS has a warm temperate climate with dry autumn months and occasional summer droughts (see chapter 2). During the afternoon hours of the summer months, plants reach wilting points due to high temperatures and low water availability in sandy soils. To reduce water loss, plants have various structural, morphological, and physiological adaptations. Some species are adapted to avoid water stress through dormancy. Drought conditions in the upland are often coupled with optimal moisture conditions in periodically flooded bottomlands, swamps, and isolated wetlands. Varied rainfall patterns and hydrology influence yearly and seasonal differences in germination success and survival, contributing to the complexity and diversity of these systems. Like flooding, frequent drought causes stress and reduces tolerance to disease and pest attack. It also has a predictable return frequency. Individuals less adapted to drought have reduced vigor, are less able to recover, and are more likely to die. By impacting juveniles and less adapted species, drought reduces the role of certain species in the future succession of a plant community.

Wind Wind disturbance associated with tropical storms may occur from late August to early October. Tornado disturbance may occur from late spring to early autumn. Strong winds can impact forest structure and species establishment. Strong winds blow down canopy trees, snap them at the base, bend them, and partially sheer off limbs. Besides damaging the canopy, however, hurricanes and tornadoes disperse seed regionally and locally. Wind moves new species into adjacent regions and nearby disturbed areas and enhances local genetic diversity. Smaller subcanopy individuals tend to be wind-sheered, bent over, or crushed by falling adjacent trees. Mortality from strong winds frequently occurs at the ridge of steep slopes, along forest-meadow margins, in saturated soils, or where unhealthy or recently overthinned trees occur.

110


Within such areas, the most susceptible individuals are shallowly rooted, taller than others in the canopy, or have weakened trunks. Certain species are inherently more susceptible because of wood strength and flexibility, architecture, and rooting patterns. Destructive winds during the dormant season primarily impact evergreen species due to added wind resistance of foliage. Fallen trees directly impact subcanopy and groundstory vegetation, and a disrupted canopy indirectly affects these layers.

Fire Fire historically occurred on SRS through lightning-caused wildfires and burning by Native Americans. Prescribed fire continues to regulate or redirect vegetation structure and plant composition. Many species and vegetation types depend on burning; yet fire disrupts biotic vigor, health, and survival. The influences of burning on plants are selective by species, size, and age. Fire impacts local competitive conditions and removes local insect pests. Burning can change ecosystem processes that regulate future seed pools, resource pools, and resource availability. Burning increases the availability of nutrients for plant uptake but reduces the total pool of nutrients within the ecosystem. The loss of organic material reduces moisture-holding capacities and can result in a xerification of uplands. The loss of surface organic material, which insulates the soil surface, elevates daytime soil temperatures and increases evaporative losses from the soil surface. The seasonality of fire also creates differences in burning effects. Burning has little impact on plants with buried or protected apical buds. Woody perennials in fire-prone habitats typically rely on root sprouting driven by stored carbon reserves and are most vulnerable to burning following spring bud-break during periods of depleted carbon reserves. Therefore, spring burning is a common strategy to reduce midstory hardwoods in pine savanna habitats.

SRS Vegetation Classifications Several vegetation classifications have been conducted on SRS. They have addressed historical conditions (Frost 1997), potential vegetation ( Jones, Van Lear, and Cox 1984; Duncan and Peet 1996; Imm 1996), and present conditions (U.S. Forest Service CISC; Pinder 1998). The U.S. Forest Service employs a standardized classification system called the Continuous Inventory of Stand Conditions (CISC), wherein aerial photography and periodic field checks are used to delineate landscapes into timber stands or

Biotic Communities

111

Table 4.1 Extent of forest cover types on the Savannah River Site CISC code 12 13 14 21 22 25 26 31 32 34 44 46 47 51 53 54 56 57 58 61 62 63 64 67 68 72 95 96 98 99

CISC forest cover type

Acreage

Percent of SRS

Shortleaf pine–oak Loblolly pine–hardwoods Slash pine–hardwoods Longleaf pine Slash pine Mixed yellow pine Longleaf pine–hardwoods Loblolly pine Shortleaf pine Sand pine Southern red oak–yellow pine Bottomland hardwoods–yellow pine White oak–black oak–yellow pine Post oak–black oak White oak–southern red oak–hickory White oak Tulip poplar–white oak–black oak Scrub oak Sweet gum–tulip poplar Swamp chestnut oak–cherrybark oak Sweet gum–willow oak Sugarberry–American elm–green ash Laurel oak–willow oak Bald cypress–water tupelo Sweet bay–swamp tupelo–red maple River birch–sycamore Water Grass Undrained flatwoods Brush species

40 4,888 251 42,028 18,042 271 539 61,942 81 147 424 4,275 601 44 6,188 22 63 449 16,142 218 11,297 920 2,797 6,639 2,574 60 4,131 12,350 596 533

0.02 2.46 0.13 21.17 9.09 0.14 0.27 31.20 0.04 0.07 0.21 2.15 0.30 0.02 3.12 0.01 0.03 0.23 8.13 0.11 5.69 0.46 1.41 3.34 1.30 0.03 2.08 6.22 0.30 0.27

Note: Forest cover types as delineated by the U.S. Forest Service Continuous Inventory of Stand Conditions (CISC) database. 1 ac = 0.405 ha.

areas of relatively homogeneous tree species composition and age. For each stand, forest managers determine the forest type, typically identified by one to three dominant canopy species. The CISC system standardizes forest types across North America and recognizes thirty on SRS (table 4.1). The system suffers from variability associated with multiple classifiers.

112


Additionally, some of the standardized types, primarily hardwood associations, do not precisely reflect SRS associations. However, because CISC is based on continually updated, ground-truthed data, it represents the most accurate description of actual current conditions at SRS. Jones, Van Lear, and Cox (1984), Frost (1997), and Imm (1996) used landscape features (e.g., topography, geomorphological classification, soil classification) to project the occurrence of general vegetation types across SRS (figures 4.1, 1.4, and 4.2, respectively). Jones, Van Lear, and Cox (1984) conducted their analyses prior to current red-cockaded woodpecker recovery initiatives and the increased emphasis on management for fire-dependent communities. They delineated existing natural vegetation from successional/management types. Frost’s (1997) classification system addressed plant communities prior to European settlement. Imm (1996) focused on in situ vegetation-type development under current management strategies. Though each made different assumptions, the three systems resulted in roughly 82 percent similarity in classification. All three classifications identified similar amounts of area for longleaf pine–sandhill scrub, cypress–tupelo swamp, Carolina bay, and upland hardwood slope communities. The collective area for bottomland and floodplain forest types was also similar among classifications. Several differences are noteworthy among the classifications. First, they disagreed over the dominant vegetation type associated with moist to mesic, well- to moderately well-drained, sandy loam surface soils. Frost’s (1997) classification assumed more frequent burning and thus favored firetolerant pine savanna associations, while Jones, Van Lear, and Cox (1984) and Imm (1996) favored mixed pine-hardwood associations. Second, they classified forest types associated with well-drained submesic sands of upland areas differently. Jones, Van Lear, and Cox identified upland loblolly pine as a sustainable forest type, Imm identified a mixed pine forest condition, and Frost associated these sites with mesic to submesic longleaf pine savannas. Again, Frost’s classification was based on pre-European settlement conditions in which frequent fire would have favored long-term dominance by longleaf pine. Jones, Van Lear, and Cox based their assumptions on site productivity and the existing conditions (infrequent fire, successional forest composition) prior to 1980. Imm based classification on site productivity and burning regime, wherein longleaf pine would be favored by periodic burning and would eventually become an important forest component, but where the burning frequency would still allow for continued codominance by loblolly pine. Finally, Imm identified a greater number of floodplain forest types than Jones, Van Lear, and Cox or Frost.

Biotic Communities

113

Jones, Van Lear, and Cox conducted fieldwork in the 1970s, while Imm’s fieldwork was some twenty years later. Thus, successional patterns may have led to further differentiations. Frost assumed that periodic burning occurred within the floodplain areas and influenced the forest composition and development. Generally, differences among the classifications focused on which pine would dominate and the relative abundance of hardwood. Pinder (1998) compared patterns of irradiance detected by satellite imagery to existing forest composition at known locations and extrapolated the correlations across SRS. According to Pinder’s analysis, pine forest dominates 42 percent of SRS. Frost (1997) classified 59 percent of SRS as longleaf pine savanna; Imm (1996) and Jones, Van Lear, and Cox (1984) identified 54 percent and 48 percent, respectively, of SRS as upland pine forest. The greatest classification differences between potential forests and Pinder’s (1998) existing habitats are the size and patterns of classified polygons. Nearly all of Pinder’s habitat polygons were less than 25 ha (62 ac), while the classification of potential forest types includes many very large areas of forest with similar composition. The present patchy pattern of habitats at SRS is due in part to human activities, such as logging and previous agricultural use. However, small patches were present in the pre-European landscape: Hardwood inclusions occurred in longleaf pine savannas, and pine inclusions occurred in bottomland hardwoods. Many soil types are well suited for a variety of forest compositions (see chapter 2) and, as evidenced by Pinder’s classification, do occur in mixed mosaics across the landscape. Other classifications of SRS vegetation have focused on specific ecosystems. For example, Gaddy (1994) and De Steven and Toner (1997) classified Carolina bays; Duncan and Peet (1996) and Smith (2000) studied pine and pine-oak communities; and Whipple, Wellman, and Good (1981), Golley, Petrides, and McCormick (1965), Jones, Van Lear, and Cox (1984), and Sharitz, Irwin, and Christy (1974a, b) described bottomland hardwood systems. Because these works were restricted to particular ecosystem types, they were not intended for broad-scale landscape classification.

SRS Vegetation Types Vegetation types are distributed across the landscape in a fairly predictable manner due to differences in history, resource availability, and disturbance patterns. At the tops of ridges, sandhill vegetation types dominate. The first transition along an elevational gradient is generally from the sandhill vegetation type to pine savanna, pine plantation, or

114


oak-pine woodland. As moisture and nutrient availability increase along the gradient, sandhill and pine savannas shift to other pine or pinehardwood communities. At depressions, pine and pine-hardwood communities change to Carolina bay and other isolated wetland communities. Near major stream corridors, steep contours support upland hardwoods associated with slopes. Finally, riparian forest types occur on the floodplain. Typically, the transition from pine and pine-hardwood communities to riparian forest communities is gradual. Increasing stream size toward the Savannah River results in expanded riparian area and increased topographic complexity, accompanied by a corresponding complexity in plant community. Smaller streams support unique seep, bog, and pocosin vegetation. In areas with nearly permanent flooding, bottomland and riparian forest types change to stream and river swamp communities. Below, we describe the general vegetation types defined by Imm (1996). The extent of each type on the SRS appears in table 4.2. This system provides greater ecological detail than the CISC system, which relies on only one to three dominant canopy species. Within each type, we also describe several communities. These communities were identified both from the literature and from discriminant analysis of unpublished data, the results of which appear in tables 4.3 through 4.9. Table 4.2 Extent of vegetation types on the Savannah River Site Vegetation type Stream swamp River swamp Bottomland hardwood Blackwater stream bottom Pine–bay hardwood Isolated wetland depressions Southern mixed hardwood Upland hardwood slope Upland pine–hardwood Yellow pine forest Longleaf pine forest Longleaf pine–scrub oak Udorthents Water

Percent of SRS 1.7 4.4 5.7 8.2 0.9 1.0 2.9 4.7 8.3 33.6 19.3 3.8 3.7 1.7

Corresponding CISC forest types 67, 68 67 46, 54, 61, 62, 63, 64, 72 31, 46, 56, 58, 62, 64, 68 31, 46, 68, 98 31, 46, 62, 64, 68, 95, 96, 98, 99 13, 46, 53, 54, 56, 62, 64 12, 32, 51, 53, 56 12, 13, 26, 32, 44, 47, 51 21, 22, 25, 31, 32 21 21, 26, 57 96, 99 95

Note: Vegetation types as delineated by Imm (1996). U.S. Forest Service CISC types included in each are also given.

Biotic Communities

115

Remnant Pine Savannas and Sandhill Woodlands In the early 1950s, limited remnant pine forests and sandhill woodlands harbored seed and isolated populations of species that were once commonly found across the SRS landscape. In most areas, the absence of fire and other land uses had removed pine savanna. Without sustained periods of frequent, low-intensity fires, fire-dependent pine savannas (figure 4.3) succeed to other communities or develop indirect losses in vigor and productivity. Both longleaf pine and wire grass influence fire behavior, are resistant to damage from fire, become competitively superior due to rapid growth response following fire and loss of competitors, and have improved reproductive and recruitment efforts following maintenance fires. Typically, fuel conditions beneath longleaf pine savannas are a continuous layer of accumulated dried grasses interdispersed with pine straw. These fuels dry rapidly and burn rapidly due to their dispersion, the unrestricted air movement, and rapid convective heat transfer via the open canopy. Nearly complete fuel consumption allows for germination from the seed bank and in situ dispersed seed. Though fire-dependent, fireadapted species do suffer some mortality during burn events, individuals that survive enjoy reduced competition and usually flourish. More important, buried seed of suitable species germinates following fire, resulting in a renewed forest floor composition of a wide range of species.

Figure 4.3. Pine savanna (D. Scott).

116


Figure 4.4. Sandhill woodland (D. Scott).

Thus, a periodic absence from burning results in a quick decline in new colonization by existing species and new colonials. Prolonged absence of burning results in the gradual loss of annuals from the seed bank and perennials from the typical continuous cover. These changes lead to an absence of exposed mineral soil suitable for colonization, a shift in fuel away from grasses toward pine straw, continued growth of sprouting hardwood shrubs, and ultimately the loss of the savanna structure. Sandhill woodlands are extremely dry and unproductive, and they produce a limited amount of highly flammable leaf litter that is interspersed with bare ground. Fires are therefore infrequent. However, due to litter accumulation over time, burning eventually removes most of the above-ground woody material, which is then replaced by rapidly growing sprouts with extensive root systems. In sandhill woodlands (figure 4.4), many tree species are resilient to fire. Turkey oak and other sandhill species tolerate burning by resprouting from carbon reserves in the root system. The understory composition and aged longleaf pine canopies of a few sites on SRS suggest that some very small areas experienced limited human disturbance. Frost (1997, 1998), Duncan and Peet (1996), and Smith (2000) identified some areas with intact understories and separated various suites of species along moisture and fertility gradients. The lack

Biotic Communities

117

of intact understories in fertile, moist areas likely reflects higher levels and a longer duration of agricultural use. Further, because moist soils tend to be more productive, they experience fewer fires and greater growth rates by invading hardwoods than better soils.

Pine Savanna and Sandhill Woodland Communities Shortleaf pine dominates the canopy on ridgetops and adjacent exposed slopes along major drainages (table 4.3, community 1). Pine ridges periodically burn and have relatively infertile sandy loam surface soils or soils with loamy horizons within 80 cm (31 in). Usually, shortleaf pine canopies incorporate other pines (loblolly, longleaf) or oaks (post, blackjack, scarlet, white, black). Usually shortleaf pine codominates with loblolly pine and longleaf pine on ridges and slopes with sandy loams. Understory shrubs include sparkleberry, blueberries, and mountain laurel. The forest floor has grasses, trailing arbutus, elephant’s foot, woodland coreopsis, asters, licorice goldenrod, blazing stars, and legumes. Infrequent pine-dominated slopes on SRS (table 4.3, community 2) have shallow inclines, submesic surface sands, and west- to southeastfacing aspects. Dominate species include loblolly pine, shortleaf pine, and some hardwoods. Pine slope understories have a sparse herb and shrub cover. Potentially, fire-adapted herbs could become established in these areas if burned. Bigleaf snowbell, fringe tree, and trailing arbutus are scattered along most pine slopes. Mesic pine savannas are composed of longleaf pine and loblolly pine (table 4.3, community 3). Most loblolly pine trees are younger and likely became established during periods of fire protection. Soils are usually loamy sands to a depth of 30 to 60 cm (12–24 in) and are underlain with either loams or clays. Shrubs are usually dense with scattered patches of little bluestem, panic grasses, and herbs. Near wetlands, many wetland herbs and ferns are also present. Dry longleaf pine savannas have a canopy of longleaf pine (table 4.3, community 4) with understory mixtures of blueberry, dwarf huckleberry, sparkleberry, grasses, and bracken fern. Soils are usually deep sands or loamy sands to depths of at least 80 cm (31 in). These soils are excessively drained and usually very limited in organic matter and clay content. Sprouts of many sandhill and drought-adapted woody plants are also present. Loamy pine savannas have sandy loam to loam surface horizons, often underlain with loamy subsoils at depths of less than 50 cm (20 in). In addition to longleaf pine, loblolly pine and shortleaf pine are codom-

Table 4.3 Percent basal area for species associated with sandhill woodland and remnant pine savanna communities on the Savannah River Site Species

1a

Pinus palustrus 1 Pinus taeda 4 Pinus echinata 78 Carya pallida Quercus stellata 3 Carya tomentosa 3 Quercus marilandica 1 Quercus falcata 1 Vaccinium arboreum 3 Quercus nigra 3 Prunus serotina 0 Quercus velutina 0 Cornus florida 2 Quercus alba 0 Nyssa sylvatica 0 Crataegus spp. 1 Diospyros virginiana Liquidambar styraciflua Ilex opaca 0 Carya ovalis 0 Quercus hemisphaerica 0 Quercus incana Quercus margaretta 0 Quercus laevis Rhus copallina 1 Sassafras albinum 0 Vaccinium stamineum 0 Prunus angustifolia Kalmia latifolia 1 Callicarpa americana Symplocos tinctoria Styrax grandiflora Chionanthus virginiana No. of plots 6 Basal area (m2/ha) 33.0

2b

3c

4d

5e

6f

7g

8h

9i

10j

0 50 30 0 1 2 0 2 2 1 0 1 1 1 1 0 0 2 1 1 2

59 31 1 0 1 0 1 0 1 0 1 0 0 0 1 1 0

88 1

32 31 13 1 0 3 0 1 1 0 1

87 1 3 1 2 0 3 0 0 0 1

54 33 9 1

10 13 4 10 1 2

46 0 0 3 0 1 0 0 1

3

0

0

0 0 1

0 1 2

0 1 5 42 0 0 0

11 13 65 0 0 1

0 0 0

0 0 0 1 1 0 0 0 0 0 1

0 0 0 2 2 1 0

0

0

1 0 0

4

0 0 1 0

1 2 5 0 0 0 0

0 2 0 0 0 2 0 0

0 0 1 1 0 0 0 1

1 0 1 0

1 1 0

1 1 0 3

1 0

1 3 1 1 0 1 3 1 0 1 0 0 12 11 15 2 0 0

1 0 1 0 1

0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 0 15 10 8 5 3 4 18 94 16 29.5 18.1 18.0 25.3 15.1 28.4 17.6 14.8 10.6

Note: Only those species accounting for >1 percent included; 0 indicates present but 1 percent included; 0 indicates present but 1 percent included; 0 indicates present but 1 percent included; 0 indicates present but 1 percent included; 0 indicates present but 1 percent included; 0 indicates present but 1 percent included; 0 indicates present but

Ecology and Management of a Forested Landscape: Fifty Years on the Savannah River Site

Ecology and management of a forested landscape: fifty years on the Savannah River Site

Management and ecology of river fisheries

Fifty Years of Invasion Ecology: The Legacy of Charles Elton

Ecology, Cognition and Landscape

The Landscape Ecology of Fire

Fifty Stories in Fifty Years

Fifty Stories in Fifty Years

Fifty Stories for Fifty Years!

Fifty Stories for Fifty Years!

Dynamic Landscape: Design, Ecology and Management of Naturalistic Urban Planting

Fifty Years of Public Service

Research and Development on a Salt Processing Alternative for High-Level Waste at the Savannah River Site

Israel's First Fifty Years

Key Topics in Landscape Ecology (Cambridge Studies in Landscape Ecology)

Harry Harrison - Fifty Stories in Fifty Years

The Beatles: fifty fabulous years

The Beatles: Fifty Fabulous Years

On the Nature of Prejudice: Fifty Years after Allport

Key Topics in Landscape Ecology (Cambridge Studies in Landscape Ecology)

Complexity in Landscape Ecology (Landscape Series)

Complexity in Landscape Ecology (Landscape Series)

Integrating Landscape Ecology into Natural Resource Management (Cambridge Studies in Landscape Ecology)

Geomorphology and River Management: Applications of the River Styles Framework

Savannah

Anthills of the Savannah

Fifty Years in Dyslexia Research

Principles and methods in landscape ecology, Second Edition (Landscape Series)

Fifty Years in Wall Street

Inequalities: Fifty years on from Hardy, Littlewood, and Polya

Savannah

Ecology and Management of a Forested Landscape: Fifty Years on the Savannah River Site