The Ecology and Silviculture of Oaks

The Ecology and Silviculture of Oaks The Ecology and Silviculture of Oaks Paul S. Johnson and Stephen R. Shifley US ...

Author: Paul S. Johnson | Stephen R. Shifley | Rogers Rogers

97 downloads 1200 Views 8MB 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

The Ecology and Silviculture of Oaks

The Ecology and Silviculture of Oaks

Paul S. Johnson and Stephen R. Shifley US Department of Agriculture Forest Service North Central Research Station Columbia, Missouri USA and

Robert Rogers College of Natural Resources University of Wisconsin/Stevens Point Stevens Point, Wisconsin USA

CABI Publishing

CABI Publishing is a division of CAB International CABI Publishing CAB International Wallingford Oxon OX10 8DE UK Tel: +44 (0)1491 832111 Fax: +44 (0)1491 833508 Email: [email protected] Web site: www.cabi-publishing.org

CABI Publishing 10 E 40th Street Suite 3203 New York, NY 10016 USA Tel: +1 212 481 7018 Fax: +1 212 686 7993 Email: [email protected]

© CAB International 2002. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners. A catalogue record for this book is available from the British Library, London, UK. Library of Congress Cataloging-in-Publication Data Johnson, Paul S. The ecology and silviculture of oaks / Paul S. Johnson and Stephen R. Shifley, and Robert Rogers. p. cm. Includes bibliographical references. ISBN 0-85199-570-5 (alk. paper) 1. Oak--United States. 2. Oak--Ecology--United States. I. Shifley, Stephen R. II. Rogers, Robert, 1941 - III. Title. SD 397.O12 J64 2001 634.97210973--dc21 2001035883 ISBN 0 85199 570 5

Typeset in Melior by Columns Design Ltd, Reading Printed and bound in the UK by Biddles Ltd, Guildford and King’s Lynn

Contents

Preface Acknowledgements Introduction Conflicting Environmental Philosophies Silviculture: a Consilient Discipline References

xi xiii 1 1 5 7

Part I. Ecology 1 Oak-dominated Ecosystems Introduction The Taxonomy of Oaks The Geographic Distribution of US Oaks Species ranges and groupings Distribution of oaks by hierarchically classified ecoregions Eastern Oak Forests The Northern Hardwood Region The Central Hardwood Region The Southern Pine–Hardwood Region The Forest–Prairie Transition Region Western Oak Forests The Southwestern Desert–Steppe Region The Pacific Mediterranean–Marine Region References

8 8 9 10 10 14 20 20 26 33 36 40 40 43 48

2 Regeneration Ecology I: Flowering, Fruiting and Reproduction Characteristics Introduction Flowering Male flowers Female flowers Factors Affecting Acorn Production Weather Premature abscission Variation in acorn production

54 54 55 55 58 61 61 62 64

v

vi

Contents

Acorn Predation and Dispersal Insects: destroyers of acorns Rodents: predation and dispersal Birds: predation and dispersal Oak Seedling Establishment Germination and initial establishment Early growth Seedling Sprouts Shoot dieback and root : shoot ratio Occurrence of shoot dieback Stump Sprouts and Related Growth Forms Definitions and origins Frequency of sprouting Sprout growth and mortality References

69 70 77 81 84 84 86 92 92 94 98 98 100 101 106

3 Regeneration Ecology II: Population Dynamics Introduction Regeneration Strategy Reproductive mechanisms: seeding and sprouting Accumulation of oak reproduction Fluctuation in population density Regeneration Potential Regeneration mode Modelling theory and objectives Stand-level regeneration models: purpose, problems and limitations References

117 117 118 118 121 142 147 148 155 157 158

Part II. Site Productivity and Stand Developmment 4 Site Productivity Introduction Measures of Site Productivity Relation of Site Productivity to Ecological Classification Productivity and Related Self-sustaining Properties of Oak Forests Effects of timber harvesting on site productivity Modifying site productivity through fertilization Methods of Evaluating Site Quality Site index Site evaluation alternatives to site index References

168 168 169 171 172 173 175 176 176 185 189

5 Development of Natural Stands Introduction Forest Canopy Layers Disturbance Disturbance type Disturbance size and frequency of occurrence Development of Even-aged Stands The stand initiation stage

194 194 194 195 196 196 199 201

Contents

The stem exclusion stage The understorey reinitiation stage The complex stage Development of Uneven-aged Stands Disturbance–Recovery Cycles References

vii

203 207 215 216 217 224

6 Self-thinning and Stand Density Introduction Self-thinning Reineke’s model The 3/2 rule Stand Density and Stocking Terminology Maximum and minimum growing space Stand density diagrams References

227 227 227 227 229 235 235 237 241 251

Part III. Silviculture, Growth and Yield 7 Even-aged Silvicultural Methods Introduction Natural Regeneration Methods The clearcutting method The shelterwood method The seed tree method Artificial Regeneration Methods Oak nursery stock Oak plantation establishment Enrichment planting Intermediate Cuttings Definitions and theory Application Special Problems: Reducing Insect and Disease Impacts Gypsy moth Oak decline Oak wilt Economic, Environmental and Social Considerations The clearcutting method The shelterwood and seed tree methods References

254 254 254 255 255 274 277 278 278 280 284 294 294 295 305 305 314 316 320 320 322 322

8 Uneven-aged Silvicultural Methods Introduction The Single-tree Selection Method Principles of application Specifying the distribution Applicability to oak forests Group Selection Method Economic, Environmental and Social Considerations References

335 335 337 337 342 352 365 372 374

viii

Contents

9 Silvicultural Methods for Multi-resource Management Introduction Oak Savannas Extent and characteristics Disturbance processes Managing oak savannas Managing Stands for Acorn Production Assessing and predicting acorn crops Effects of tree size and stand characteristics Guidelines for sustaining acorn production Old-growth Oak Forests Extent and characteristics Silvicultural options Forests in transition to old growth Old-growth forests at the landscape scale Aesthetics Stand-level aesthetics Landscape-level aesthetics References

380 380 380 380 383 385 391 392 395 399 403 403 404 408 409 410 411 415 417

10 Growth and Yield Introduction Growth of an Oak Annual phenology Diameter growth Height growth Survival rates Stand Growth Growth and yield in even-aged stands Growth and yield in uneven-aged stands Growth and Yield Models Modelling methods Stand-level models for oaks Stand table projection models Individual-tree-level models for oaks OAKSIM TWIGS Forest Vegetation Simulator Estimating ingrowth Model evaluation Volume Equations Regional Patterns in Oak Yield and Productivity References

424 424 425 425 426 432 437 439 439 446 447 447 448 451 451 452 453 455 455 459 459 460 461

Appendices Appendix 1. Common and Scientific Names of Species Appendix 2. Forest Cover Types of Eastern United States Dominated by Oaks or Oaks Mixed with Other Species Appendix 3. Forest Cover Types of Western United States Dominated by Oaks or Oaks Mixed with Other Species

468 475 480

Contents

Appendix 4. Formulae for Converting Site Index for Indiana, Kentucky, Ohio, and West Virginia Appendix 5. Converting Site Indexes for Four Regions Appendix 6. Converting Site Indexes, Yellow-poplar to Oak Appendix 7. Height/dbh Site Index Curves Index

ix

483 484 486 487 489

Preface

The earth is to be seen neither as an ecosystem to be preserved unchanged nor as a quarry to be exploited for selfish and shortrange economic reasons, but as a garden to be cultivated for the development of its own potentialities for the human adventure. (René Dubos, 1976)1

This book is written for forest and wildlife managers, ecologists, silviculturists, environmentalists, students of those fields, and others interested in sustaining oak forests for their many tangible and intangible values. The focus is on the oaks of the United States. Although the approach is fundamentally silvicultural, it is based on the premise that effective and environmentally sound management and protection of oak forests and associated landscapes should be grounded in ecological understanding. Although the subject is inherently scientific and technical, we have striven to make it generally accessible by minimizing the use of technical jargon. Where technical terms are necessary for efficient expression of concepts, we have first defined them. Much has been written about the ecology and silviculture of oaks. So much so that the related body of literature represents, in one sense, an informational ‘embarrassment of riches’. The embarrassment derives primarily from the paucity of synthesis within and across two broad fields of study. The first is ecology, which is the scientific study of the processes and relations among organisms and between 1Symbiosis

organisms and their environment including associated energy transformations. The second is silviculture, which is the art and science of producing, tending, and sustaining forests. Although the literature on North American oaks dates to the colonial period, most of it was written within the last 50 years, and a large proportion of that within the last 25 years. However, much of this literature resides in relatively obscure scientific and technical journals, proceedings of professional and scientific meetings, government publications, and other sources that are often difficult to locate and retrieve. But even with ready access to this disparate information, its synthesis into an holistic framework of knowledge is a daunting task. This book attempts to ease, if not eliminate, those problems. Although ecology has become a household word, silviculture has not. Nevertheless, silvicultural practices have shaped the character of the landscape wherever oaks and associated forests occur, which includes much of the United States. Those practices often have produced negative public reactions and sometimes even deleterious ecological consequences. Increasing economic demand for oak wood nevertheless makes timber harvesting and its aftermath an ever more conspicuous feature of the landscape. Moreover, the distinction between designed silvicultural practices and purely exploitative logging practices is not always apparent, especially to the public.

between the earth and humankind. Science 193(4252), 459–462 (1976).

xi

xii

Preface

Contemporary philosophies on how oak forests and associated resources should be managed range from narrowly preservationist or narrowly utilitarian to more inclusive and integrative multiple-value philosophies. One of the objectives of this book therefore is to present ecological and silvicultural concepts that can be used to address an array of problems defined by various perceptions of how oak forests should be treated. The current trend in managing forests and forested landscapes is away from a narrow focus on sustaining timber and other commodity outputs and towards a broader philosophy of sustaining desired ecological states. This shift in the forest management paradigm has been wrought by and is consistent with changing social values, scientific advances in ecology and society’s increasing awareness of environmental problems and expressed concerns on how those problems affect us collectively and individually. Consistent with the new paradigm, this book is designed and intended not so much as a how-to-do-it management manual as it is a source of ideas on how to think about oak forests as responsive ecosystems. Armed with that understanding, we believe managers and conservators of oak forests will be better positioned to adapt to changing social values and simultaneously to build and act on co-evolving ecological and silvicultural information. The book is divided into three sections. The first contains three chapters on the ecological characteristics and distribution of oak species and the various kinds of oak forests in the United States, differences among them and how they have been classified, their natural development, and the relation of oak forests to environment and related environmental concerns. The next two chapters on regeneration ecology provide the critical interface between oak ecology and silviculture. Understanding the regeneration ecology of the oaks is paramount to silviculturists because of widespread difficulties in regenerating and thus sustaining oak forests.

The second section comprises three chapters covering site productivity and stand development. An understanding of the productive capacity of oak forests is central to a broad spectrum of issues related to their management and potentialities, not only for timber but also for wildlife and other values. The chapters on stand development, self-thinning and stand density present concepts that are key to the application of silvicultural methods. The third section comprises four chapters on silvicultural methods and the growth and yield of oak forests. Silivcultural methods include traditional even-aged and uneven-aged methods as well as non-traditional methods for multi-resource management and conservation. Regeneration methods are discussed in relation to the apparent regeneration strategies that have evolved in the oaks and how those strategies vary among oak-dominated ecosystems. The approach to regeneration thus is less prescriptive and more ecologically principled than that typically presented in silviculture textbooks and ‘how-to’ guides. Throughout the book, accepted common names of trees follow Little’s (1979) Check List of Native and Naturalized Trees of the United States. Scientific names of trees and other organisms are listed in Appendix 1. We express our appreciation and indebtedness to all the ecologists, foresters, wildlife biologists, soil scientists, entomologists, pathologists and others, past and present, who have contributed to our collective knowledge of the oaks. We are hopeful that this compilation will make some small contribution to a more ecosystem-centred approach to managing and conserving oaks in the many forests and plant communities in which they occur. Paul S. Johnson Stephen R. Shifley Robert Rogers July 2001

Acknowledgements

The authors gratefully acknowledge the following for their invaluable assistance, without which this book would not have been possible. For their technical assistance and for creating a productive and congenial work environment, we are indebted to North Central Research Station technicians and support staffers Tim Bray, Dianne Brooks, Kenneth Davidson, Laura Herbeck, Kevin Huen, James Lootens, Marilyn Magruder, Allison Ramsey, Hoyt Richards and Neal Sullivan. Special thanks are accorded Lynn Roovers for countless hours spent on graphics and other details, and to William Dijak for his diligence and talent in preparing the maps for Chapter 1. For selflessly contributing their ideas, art work, photographs, reviews, and technical and scientific knowledge, we are deeply indebted to USDA Forest Service foresters and scientists Robert Cecich, Daniel Dey, Jeffrey Goelz, Gerald Gottfried, Kurt Gottschalk, David Graney, James Guldin, Jay Law, Edward Loewenstein, David Loftis, Robert McQuilkin, Ross Melick, Paul Murphy, Felix Ponder, Ivan Sander, Susan Stout, Richard Teck, Gary Z. Wang, Dale Weigel, Daniel Yaussy and John Zasada. We also thank all our partners in the University of Missouri School of Natural Resources for their valuable advice and support. Special thanks are accorded to Drs John Dwyer, H.E. Garrett, David Larsen, Bernie Lewis,

Nancy Loewenstein, Rose-Marie Muzika, Stephen Pallardy and Mr Dustin Walter for their reviews and ideas. A special thanks to Drs W. Carter Johnson, William Kearby and Walter Koenig for allowing us to reprint their marvellous photos, and Drs Jeffrey Ward (University of Connecticut), Carl Ramm (Michigan State University), William Parker (Ontario Forestry Research Institute), Willard Carmean (Lakehead University, Thunder Bay, Ontario) and an anonymous California reviewer for their constructive reviews and suggestions. Thanks to the Pioneer Forest of Salem Missouri and Leo Drey and Clint Trammel for allowing us to use their forest inventory data to evaluate uneven-aged oak silviculture. Thanks to the USDA Forest Service and the North Central Research Station Assistant Directors Donald Boelter and David Shriner, and Project Leader Frank Thompson, for allowing us the freedom to pursue and complete the project. Perhaps it goes without saying that the book could not have been written without the many researchers, past and present, who have contributed to the rich body of literature in oak ecology and silviculture; but we do not take their dedication and contributions to forest science for granted. Finally, we thank our families for their patience, forgone summer vacations and enduring support.

xiii

Introduction

Conflicting Environmental Philosophies Ecology is the scientific study of the interrelations among living things and their environment. Ecological knowledge effects an awareness of precarious interdependencies among the myriad organisms, large and minuscule, between organisms and non-living components of ecosystems, and the pervasive human impacts that threaten these relations. Ecology thus obviates our dependency on, and our relation to, natural processes and systems. Perhaps no science more so than ecology has generated more knowledge with implications relating to ethics, morality and human behaviour. In contrast, silviculture is the art and science of tending forests to meet human needs. Because silviculture is usually directly involved in the extraction of biomass, it produces disturbances along with associated ecological side effects. Silviculture is thus based on the planned use of controlled and directed disturbances to achieve defined human objectives. Ideally, it should be based on scientific principles which ensure that specified silvicultural goals are consistent with preserving or improving a forest’s ecological qualities, are compatible with its natural dynamic and thereby provide reasonable assurance of the forest’s sustainability. Like its parent discipline, forestry, silviculture evolved out of 17th century Europe in response to purely utilitarian needs, especially for the timber required

for sustaining the large naval armadas required for projecting colonial power in the late 18th century. Paramount among these concerns in Britain and France was a ready supply of pine and oak for ship masts and hulls. However, in the United States, serious concern over a declining forest resource did not occur until the late 19th century. By then the forests of eastern United States had been decimated by exploitative logging. A small but politically influential group of conservationists feared the same would happen to the western forests. This prompted the setting aside of forest reserves in the early 1890s from what remained of the public domain in the west. In 1897, the Organic Act was passed, which specified that the purpose of the reserves was ‘to improve and protect the forest within the reservation, or for the purpose of securing favourable conditions of water flows, and to furnish a continuous supply of timber for the use and necessities of citizens of the United States’ (United States Congress, 1897). This landmark legislation specified that the forest reserves were intended for managed use, not for wilderness preservation. Following the recommendations of the American Forest Congress of 1905, the reserves were transferred from the Department of Interior to the Department of Agriculture. Known as the Transfer Act, it provided that funds from the sale of products or the use of land in the reserves be used for managing and developing the forest reserve system. 1

2

Introduction

This change heralded the implementation of an ambitious programme of scientific forest management under the direction of Gifford Pinchot, the first Chief of the USDA Forest Service. At that time, forestry was virtually an unknown discipline in the United States and forestry curricula in US universities were just emerging. Although politically controversial in its day, the conservation movement was hailed by its founders as not only environmentally wise, but also economically beneficial (Pinchot, 1987). Pinchot and the founders of the early forest conservation movement envisioned a scientifically based forestry that would not only provide conservation benefits but would also result in the economic stability of rural communities in forested regions. Such benefits would accrue, they argued, from the application of scientifically derived sustained yield principles, which would ensure for perpetuity the even flow of timber and other commodities originating from the forest (Pinchot, 1987). Because the scientific underpinnings of sustained yield were largely invested in silviculture, and because silviculture has historically been justified on economic grounds, silviculture philosophically straddled agronomy (i.e. growing trees as crops) and economics. However, modern silviculture has been broadened to include not only sustaining timber yields, but also sustaining non-commodity values including oldgrowth forests, biodiversity, wildlife habitat and aesthetics. In this wider context, silviculture assumes application to a panoply of values that transcend economic utilitarianism. Despite the differences between the two disciplines, contemporary silviculture as it has been applied to most North American forests, remains naturally allied with and dependent upon ecology for much of its scientific underpinnings. The schism between silviculturists and some ecologists nevertheless runs deep. One source of this disunion emanates from the ecologists’ traditional focus on studying ecological processes in ecosystems largely unaffected or minimally affected by

humans and drawing conclusions therefrom. In contrast, silviculturists depend on scientifically based knowledge of disturbance-mediated mechanisms to control and direct forest ecosystem processes for human benefit. Recovery from such disturbances is predicated on the assumption that forests are inherently resilient, i.e. capable of rapidly returning to their previous or other silviculturally directed state. The silviculturist’s anthropocentric view of the forest is anathema to those who adhere to the biocentric view, which elevates nature to a position superior to human self-interest (Devall and Sessions, 1985; Chase, 1995; Ferry, 1995; Fox, 1995). The biocentrist’s agenda is centred on maintaining ‘natural’ ecosystems, including forest, in states free from human interference, and the need for establishing the pre-eminence of those states. From that perspective, human-mediated disturbance is seen as a disrupter of fragile ecosystems and the intended order of things. Moreover, such disruptions can potentially produce species extinctions and other irreversible environmental effects. The biocentric view therefore holds that the best way to preserve nature, wherever some vestige of it remains, is to leave it alone (Devall and Sessions, 1985; Chase, 1995). Humans are viewed as just one of many organisms in the biosphere no more important than any other – and like all component organisms should be subordinate to the healthy functioning of the interactive whole, i.e. the ecosystem. Biocentrism is therefore egalitarian among organisms and premised on an inherent right to life of all species and life forms. By extension, maintaining ecosystems in their ‘natural’ state becomes a social imperative. A biocentrist thus may view silviculture, along with other human interferences in the development of forests, as ecologically threatening, if not ruinous. The biocentric interpretation of the ‘message’ from ecology is thus at irreconcilable odds with the interpretation from silviculture. Biocentrism nevertheless now occupies a position of social and political prominence (Chase, 1995).

Introduction

The connections between ecology and silviculture none the less are apparent and important, especially when silviculture is applied to forests of natural origin. In that setting, silviculture by itself may not introduce new species or populations (i.e. new genetic material) from outside the forest. Human energy expenditures are often limited only to those required in cutting and removing trees. Such relatively non-intensive practices have characterized the silviculture applied to oak forests of the United States. There, oak silviculture has largely followed an ecological model whereby forests are managed by directing their continually changing states, or ecological successions, through manipulation of existing on-site vegetation and propagules. This approach relies on periodic timber harvesting and usually natural regeneration to maintain or periodically recreate desired ecological states. It contrasts with the more intensive agronomic model used in growing pine plantations and other monotypes. The latter approach usually depends on artificial regeneration, the introduction of new and ‘improved’ genotypes, exotic species, and other intensive and energy-expensive cultural methods like those used in agriculture, horticulture and agroforestry (growing trees intermixed with agricultural or horticultural crops). Nevertheless, the silvicultural methods that have been applied to oaks span the entire range of approaches from ecological to agronomic. In the public’s view, silviculture is an often confusing and controversial subject exacerbated by the claims of some environmentalists that it is an ecologically damaging enterprise that ‘seeks to accept “tree farms” in place of natural forests … The usual approach … is to seek ever more intensive management, which spawns even more problems’ (Devall and Sessions, 1985, p. 146). By comparison, there is seemingly little controversy and confusion over the reason to preserve something in its natural state free from human interference if it is otherwise threatened with extinction – even though the method or means of preservation may be debatable. Likewise, the reason for the cultivation and harvest of a corn field is easily understood and

3

accepted because of its purely utilitarian value, and its physical origins borne of human endeavour. Socially, silviculture is a more complicated issue. It is vulnerable in appearance, conceptually and often physically, seen as conforming to neither preservation nor agronomy. It is neither fish nor fowl, yet is often identified as disruptive if not exploitative of nature. To the non-silviculturist, application of the ecological model to silviculture may sometimes be difficult to distinguish from purely exploitative and environmentally damaging practices. However, such exploitation is not the intent of, nor does it constitute, silviculture. Silviculture is not synonymous with timber harvesting, yet is dependent upon it. The objective of modern silviculture is to create and maintain forests by design that produce material and nonmaterial benefits to humans without sacrificing their sustainability. Silvicultural intentions nevertheless are not ecologically infallible. A given silvicultural application, despite best intentions, may be inconsistent with ecological realities because of our incomplete knowledge and understanding of ecosystems. Poorly applied silviculture therefore can produce unintended and negative long-term ecological consequences. The possibilities for such outcomes impose serious responsibilities on silviculturists in the practice of their art and science. When silviculture is applied to ‘natural’ ecosystems, the intent, some would say, is to improve on nature by tinkering with it. But the biocentrist would argue that humans cannot improve upon nature – a notion consistent with the theological view that ‘man cannot improve upon God’s handiwork’. And much ecological knowledge and theory is purported to support that perception. Perhaps it is the proximity of the existing ‘near-natural’ state to the intended silviculturally created state that concerns those whose sentiments might be to ‘leave well enough alone’. The biocentrist might argue that silviculture promises only ‘a kinder, gentler rape of the forest’. These views may be further bolstered by an awareness of the shrinkage of natural ecosystems globally and its consequences.

4

Introduction

The fragmentation of today’s landscape into discrete blocks of forests spatially detached from human development may further reinforce the perception of the separation of humans and forest. This outlook is reflected in the Latin origin of the word forest, foris, which means outside. This etymology suggests a human view of forests evolving from deep historical and psychological roots, and one in which forests are functionally disconnected from humans. Even as late as the 18th century, the forest was perceived as something ‘beyond’ the boundary of European culture (Bonney, 1996). Today, most of the US population resides in urban areas. There, sources of basic human-sustaining resources are physically distanced from and foreign to everyday experience. A perception of forests as functionally and spatially distant from humans may be further reinforced by the common acknowledgement that, in some cases, the physical separation of humans from nature is necessary to preserve rare or endangered species and habitats. It is generally accepted that such preservation is a democratically mandated function of government. The results are commonly and favourably experienced annually by millions of visitors to national and state parks, wildlife refuges, and designated wilderness areas in national forests and other federal lands. It is generally understood that the role of humans there is restricted to that of protector and spectator, but not interloper. Contrasting with such models of the separation of humans and nature is the historical relation between humans and oaks, which are characterized by connectedness. From the oak’s perspective, those connections have produced both beneficial and harmful effects. The ecological evidence, as later discussed, nevertheless indicates that sustaining and thus preserving many oakdominated ecosystems will require human intervention. Humans and oaks have been closely associated throughout history. Before the arrival of Europeans, native Americans set fires, both accidentally and intentionally, which often burned out of control over enormous areas (Grimm, 1983;

Pyne, 1982, 1997; Guyette et al., 1999). Periodic fires were repeated over centuries in regions indigenous to the oaks, which includes much of North America. Their frequent occurrence created extensive areas of open-grown forests favourable to the survival of the relatively light-demanding but fire-tolerant oaks. It was a disturbance cycle that, in time and space, is unlikely to be repeated. Humans thus have had a prominent effect in shaping the nature and extent of the oak’s habitat, and perhaps even its evolution. But those events have been largely relegated to history. Much of what today remains of the oak forests of the United States is a legacy of an earlier disturbance history that was partially, if not largely, dependent on fire. Unlike the ecologist, the silviculturist has traditionally viewed forests from a utilitarian perspective that emphasized timber production. Accordingly, failure to harvest forests at their inherent sustainable capacity to produce wood (sustained timber yield) is deemed wasteful. A theological counterpart is seemingly expressed by the biblical admonition for man to exert dominion over the earth (Genesis 1:28). An economic analogue is expressed in Adam Smith’s 1776 treatise on the inherent value of the individual pursuit of economic selfinterest (Smith, 1870). Collectively, these beliefs and values, largely borne of the Enlightenment, have dominated the thinking and institutions of Western civilization for over 200 years. Self-interest prevails among private forest owners today, whether ownership goals are economic or non-economic. To a lesser extent, economic objectives dominate the management of many publicly owned lands, including the national forests. In the past, agency mandates, operating budgets and incentives tied to timber sales, produced powerful inducements to emphasize timber production, albeit within calculated sustained yield limits. Only within the last few decades has this philosophy been seriously challenged. Such material utilitarianism reduces forests to collections of trees having only commercial value. Other values are consequently diminished. The

Introduction

American conservationist, Aldo Leopold (1966, p. 251) expressed concern for this philosophy by asserting that ‘… a system of conservation based solely on economic self-interest is hopelessly lopsided. It tends to ignore, and thus eventually to eliminate, many elements in the land community that lack commercial value, but that are (as far as we know) essential to its healthy functioning.’

Silviculture: a Consilient Discipline The practice of silviculture therefore is caught in a web of competing values arising from different philosophies, ranging from biocentrism to economic utilitarianism. Unlike ecology, silviculture is directly connected to social institutions and conventions apart from science. Lying within its parent discipline, forest management, it is subject to the legal and social constraints of environmental law and policy operating within democratic processes (at least in the United States and other democratic countries where silviculture is practised). Within the context of democracy, silviculture is therefore socially integrative, i.e. in its application it must consider values borne of diverse social and political interests. Silviculture nevertheless lies at the core of forest resource management because its application results in direct physical action on the forest. This is also where fundamental scientific analysis is most needed. However, silviculture does not stand firmly by itself as a scientific discipline. This results in part from its strong connections to social and political institutions, and in part from its interdisciplinary qualities as a science. Within the biological domain of science, silviculture is most closely allied to ecology. However, it is also heavily dependent on plant physiology and genetics, plant pathology, entomology, and applied mathematics and statistics. Among the physical sciences, it borrows knowledge from geology, climatology, hydrology and soil science. It is also closely allied to other resource management disciplines including wildlife, fisheries, water and air

5

quality management. Silviculture therefore is inherently scientifically integrative. Silviculture consequently depends on linking knowledge and theories across many disciplines, both scientific and nonscientific, to form what Wilson (1998) terms ‘… a common groundwork of explanation’. If we accept that such linkages comprise consilience, we might consider that silviculture fits Wilson’s context, i.e. it comprises a hybrid domain of knowledge in which consilience is implicit. Because of silviculture’s socioeconomic connections, this consilience extends to other branches of learning including the social sciences and humanities. These connections can be represented by a series of concentric circles representing the social hierarchies within which silviculture exists. With silviculture at its centre, each ring of the social hierarchy bounds all the great areas of knowledge, including the biological, physical and social sciences as well as the humanities (Fig. I.1). This representation emphasizes the consilient nature of silviculture by placing it at the locus of all knowledge comprising its context. It represents an ideal, a unity of learning in which subjects that have been traditionally compartmentalized are breached in Wilson’s words, to ‘… provide a balanced clearer view of the world as it really is … A balanced perspective cannot be acquired by studying disciplines in pieces but through pursuit of the consilience among them … The enterprise is important for yet another reason: it gives ultimate purpose to intellect’ (Wilson, 1998, p. 13). Despite the complexities of silviculture’s complete context, our intent in the following pages is to present a synthesis of the ecological and silvicultural knowledge of oak forests in the United States. It is not to resolve the environmental issues surrounding oak forests, which fall into the social, legal, political and managerial domains represented by the concentric circles surrounding silviculture in Fig. I.1. The silvicultural context is nevertheless broad, and not limited to narrowly defined economic or commodity-production objectives. Consistent with the view of silviculture

6

Introduction

Humanities

Physical sciences cy ra oc l em ta D en licy nm po ro d vi an En w la

t st en re m Fo age an

m t st en re m Fo ge l a ta an en licy m nm po ro d vi n a En w cy la ra oc

em

D

Social sciences

Biological sciences

Fig. I.1. Silviculture’s relation to other disciplines. Concentric circles represent the social hierarchy within which silviculture exists in a democratic society. With silviculture at its centre, each ring of the hierarchy bounds all the major areas of knowledge, including the biological, physical and social sciences, and the humanities.

as a consilient discipline, we view the role of the silviculturist as just one of many possible players in the management of oak forests. Unlike the biocentrist, we infer no moral imperative to create or maintain oak forests in specified states other than those that are perceived, as best we can discern, as sustainable, beneficial, and pleasing to humankind, and that provide habitat for the many plant and animal species naturally associated with oaks. We believe these goals are consistent with the philosophy of land stewardship and wise use as proposed by earlier generations of conservationists, from which the more recent philosophy of ecosystem management has evolved. Our intention is to present information that can lead to an understanding of, and solutions to, silvicultural problems

related to oak forests. Moreover, we hope that this information fosters an informed and amiable dialogue and trust among foresters, land managers and owners, environmentalists, students and others interested in oak forests. The subject therefore is presented from a silvicultural perspective. The approach comprises a comprehensive view of forests as providing important social, spiritual and economic needs. Such an approach requires anticipating and managing for change, both predictable and unpredictable. This notion is consistent with Botkin’s (1990) call for a ‘new management’, wherein conservation and utilization of forest resources are compatible parts of an integrated ecosystem approach. It contrasts with the ‘old management’ in

Introduction

which conservation was too often subordinate to timber and other commodity production. The central concern of the new forest management, or ecosystem management (Salwasser, 1994), is the sustainability of forested ecosystems and associated human values in a continually changing mosaic of landscape patterns. The resulting

7

management and silviculture therefore must accommodate the complexities of the inevitably and continually changing ecological states that comprise a forested landscape. It also recognizes that such changes occur with or without human interference, and that we have both potentialities and limitations in controlling these changes.

References Bonney, W. (1996) Troping trees. In: Schultz, K.L. and Calhoon, K.S. (eds) The Idea of the Forest. Peter Lang, New York, pp. 119–146. Botkin, D.B. (1990) Discordant Harmonies. Oxford University Press, New York. Chase, A. (1995) In a Dark Wood: The Fight Over Forests and the Rising Tyranny of Ecology. Houghton Miflin, Boston. Devall, B. and Sessions, G. (1985) Deep Ecology. Gibbs M. Smith, Layton, Utah. Ferry, L. (1995) The New Ecological Order. University of Chicago Press, Chicago. Fox, W. (1995) Toward a Transpersonal Ecology. State University New York Press, Albany, New York. Grimm, E.C. (1983) Chronology and dynamics of vegetation change in the prairie-woodland region of southern Minnesota, U.S.A. New Phytologist 93, 311–350. Guyette, R., Dey, M. and Dey, D.C. (1999) An Ozark fire history. Missouri Conservationist 60, 4–7. Leopold, A. (1966). A Sand County Almanac. Ballantine, New York. Pinchot, G. ([1947] 1987) Breaking New Ground. Island Press, Washington, DC. Pyne, S.J. (1982) Fire in America. Princeton University Press, Princeton, New Jersey. Pyne, S.J. (1997) America’s Fires: Management on Wildlands and Forests. Forest History Society, Durham, North Carolina. Salwasser, H. (1994) Ecosystem management: can it sustain diversity and productivity? Journal of Forestry 92(8), 6–10. Smith, A. (1870) An Inquiry Into the Nature and Causes of the Wealth of Nations. London. United States Congress (1897) Surveying the public lands. US Statutes at Large 30, Ch. 2, pp. 32–36. Wilson, E.O. (1998) Consilience: The Unity of Knowledge. Knopf, New York.

1 Oak-dominated Ecosystems

Introduction A truly ecological perspective recognizes that humans and their activities are part of nature, and that enhancing all aspects of their lives – including their surroundings – begins with cooperation between individuals, based on mutual trust … Rather than halting or reversing disturbances, we should embrace change. Rather than excluding man from the garden, we should welcome his cultivation of it. (Alston Chase, 1995)

From earliest times, oaks have held a prominent place in human culture. Their uses have included wood for fuel, acorns for hog fodder and flour meal for human consumption, bark for tanning, wood strips for weaving baskets, charcoal for smelting ore, timbers for shipbuilding, mining timbers, railroad ties, pulpwood for paper, and lumber and laminates for furniture, panelling and flooring. Through the mid-19th century, oak was the wood of choice for shipbuilding in Europe and America. For that reason, oak forests and even individual trees were treated as critical national assets. During the Revolutionary War, the poor condition of the British fleet, which lacked replacements and repairs due to shortages of suitable oak timbers, may have contributed to the war’s outcome (Thirgood, 1971). In the 17th century, alarm over the depletion of timber supplies, especially oak, prompted passage and enforcement of laws mandating the protection, culture and establishment of forests in several European countries. In turn, those events influenced the development of scientific silviculture, as we know it today. 8

Modern as well as ancient man has benefited from the oak’s relation to wildlife. Wherever oaks occur as a prominent feature of the landscape, wildlife populations rise and fall with the cyclic production of acorns. Numerous species of birds and mammals are dependent on acorns during the food-scarce autumn and winter months. Even human cultures have relied on oaks as a staple food. Acorns were an important part of the diet of Native Americans in California before the 20th century (Kroeber, 1925) (Fig. 1.1). Today, the ecological role of oaks in sustaining wildlife, biodiversity and landscape aesthetics directly affects the quality of human life. The demand for wood products from oaks nevertheless continues to increase and compete with other less tangible values. Some have proposed that forests, including those dominated by oaks, are best allowed to develop naturally, free from human disturbance. What should the balance be among timber, wildlife, water, recreation and other forest values? Is there some middle ground that adequately sustains multiple goals? Informed answers and perspectives require an understanding of the ecology of oaks and the historical role that humans have had in that ecology, especially the comparatively recent role of humans in the ‘protection’ of oak forests from fire. A prerequisite to such understanding is a general knowledge of the oak’s geographical occurrence, taxonomic diversity, adaptations to diverse environments, and the historical changes in its environment.

Oak-dominated Ecosystems

Fig. 1.1. Native American collecting acorns as shown in Hutchings’ California Magazine in 1859. Acorns were a staple food of most California tribes before the end of the 19th century. They were gathered in conical woven baskets, which could hold a bushel or two of the nuts. Although the acorns of many species were eaten, favoured species were California black and California live oaks (Pavlik et al., 1991). After removing the shell (pericarp), acorns were ground into a flour, leached of tannins by soaking in running water, and then used to make a variety of foods including porridge and bread. Acorns were so highly valued that they sometimes provoked inter-tribal ‘acorn wars’. They were also widely utilized as food by Native Americans in the eastern United States. (Courtesy of the Bancroft Library, University of California, Berkeley.)

The Taxonomy of Oaks Taxonomically, the oaks are in the genus Quercus in the family Fagaceae (beech family). The Fagaceae probably originated in the montane tropics from which its members migrated and diverged into the current living genera by the late Cretaceous period (about 60 million years ago) 1

(Axelrod, 1983). By that time, mammals and birds had only recently evolved. Rapid speciation of oaks commenced in the middle Eocene epoch (40–60 million years ago). This was in response to the expansion of drier and colder climates, and subsequently to increased topographic diversity in the late Cenozoic era (< 20 million years ago) and fluctuating climates during the Quaternary period (< 2 million years ago) (Axelrod, 1983). Their fruit, the acorn, distinguishes the oaks from other members of the beech family (e.g. the beeches and chestnuts). With one exception, all plants that produce acorns are oaks. The exception is the genus Lithocarpus, which includes the tanoak of Oregon and California. Although represented by only one North American species, Lithocarpus is represented by 100–200 species in Asia (Little, 1979). Lithocarpus may be an evolutionary link between the chestnut and the oak (McMinn, 1964; cf., Miller and Lamb, 1985, p. 200). Worldwide there are about 400 species of oaks, and they are taxonomically divided into three groups: (i) the red oak group (Quercus section Lobatae1); (ii) the white oak group (Quercus section Quercus2); and (iii) the intermediate group (Quercus section Protobalanus3) (Tucker, 1980; Nixon, 1997). All three groups include tree and shrub species. The red oaks and white oaks include evergreen and deciduous species, whereas the intermediate oaks are all evergreen. The red oaks are found only in the Western Hemisphere where their north–south range extends from Canada to Colombia. In contrast, the white oaks are widely distributed across the Northern Hemisphere. The intermediate group comprises only five species, all of which occur within southwestern United States and northwestern Mexico. Many of the world’s oaks occur in regions with arid climates, including Mexico, North Africa and Eurasia, where they are often limited in stature to shrubs and small trees. About

Subgenus Erythrobalanus in earlier classifications. Subgenera Lepidobalanus and Leucobalanus in earlier classifications. 3 Subgenus Protobalanus in earlier classifications. 2

9

10

Chapter 1

80% of the world’s oaks occur below 35° north latitude and fewer than 2% (six or seven species) reach 50° (Axelrod, 1983). The most reliable distinction between the white oaks and red oaks is the inner surface of the acorn shell. In the white oaks it is glabrous (hairless) or nearly so, whereas in the red oaks it is conspicuously tomentose (hairy or velvety) (Tucker, 1980). In the intermediate group, this characteristic is not consistent among species. The leaves of the white oaks are usually rounded and without bristle tips whereas the leaf lobes of the red oaks are usually pointed and often bristle-tipped. To many silviculturists, ecologists and wildlife biologists, the most important difference between the white oaks and red oaks is the length of the acorn maturation period. Acorns of species in the white oak group require one season to mature whereas species in the intermediate and most of the red oak group require two seasons. The white oaks and intermediate oaks are characterized by the presence of tyloses (occlusions) in the latewood vessels (waterconducting cells) whereas tyloses are usually absent in the red oaks. These vesselplugging materials confer greater decay resistance to the wood of the white and intermediate oaks than the red oaks. Other morphological features that differentiate the three groups and species within them are presented in various taxonomic treatments (e.g. Tucker, 1980; Jensen, 1997; Manos, 1997; Nixon and Muller, 1997) and field identification guides (e.g. Miller and Lamb, 1985; Petrides, 1988; Petrides and Petrides, 1992). These sources also include range maps. In addition, the Silvics of North America, Vol. 2 (Burns and Honkala, 1990) provides information on the silvics and geographic ranges of 25 oaks. Of the more than 250 oak species occurring in the Western Hemisphere, the largest number occurs in Mexico and Central America. About ten species occur in Canada. For the United States species, the most complete and authoritative taxonomic treatment of the oaks is in the Flora of North America North of Mexico, Vol. 3 (Flora of North America Editorial

Committee, 1997), which lists 90 species of oaks native to the continental United States. However, we follow the taxonomic nomenclature of Little’s (1979) Checklist of United States Trees because of its widespread use in North American forestry literature. This checklist recognizes 58 native oak species plus nine varieties. Of these, about ten species are shrubs or shrub-like forms. More than 80 hybrids also have been described (Little, 1979; Tucker, 1980).

The Geographic Distribution of US Oaks Species ranges and groupings The oaks are widely distributed across the United States (Fig. 1.2). According to Little (1979), about 40 species and varieties occur east of the 100th meridian and about 30 species and varieties occur to the west. Only two species, chinkapin oak and bur oak, are common to both regions. Bur oak extends to the northwest whereas chinkapin extends to the southwest beyond the 100th meridian. The western oaks fall into three geographically distinct groups. One group is comprised of the west Texas oaks (nine species and varieties), and a second includes the southwestern oaks (16 species) that occur in New Mexico, Arizona, Utah, Colorado and Nevada. A third group is comprised of the Pacific Coast oaks (about 13 tree species plus several shrubby species) occurring largely in California, Oregon and Washington. Within the United States, numbers of oak species vary regionally. Based on a count of the number of oak species that occur within 6000 square mile areas, oak species ‘richness’ reaches a maximum of 20 species in the southeast (Aizen and Patterson, 1990) (Fig. 1.3). There, the ranges of several narrowly distributed North American oak species overlap with the ranges of several widely distributed species. Although the range of an oak species is positively correlated with its acorn size, the reason for this is unknown (Aizen and Patterson, 1990).


11

100 M21tb

M331

M241

211

332

241

263 M261

211

M333 M334

211

22tb

341

221a 341

342 261

211

M334

M332 M241

221a

251

M341

262

331

M341 M334

M341 342

M262

221b

M221

332 323

M222

311 M311

313 232

M311 314

321

252

M231 232 231

232 231

411

Fig. 1.2. The distribution of oaks in conterminous United States. The shaded areas represent the aggregated vegetation cover types within which oaks frequently occur as important species at a scale of 1 km2. The 100th meridian demarcates the approximate division between eastern and western oaks. Generated from Advanced Very High Resolution Radiometer satellite images (1990) and an associated system of land cover classification (USDA Forest Service, 1993; Powell et al., 1994). Ecoregion boundaries are from Bailey (1997). See Tables 1.2 and 1.3 for index to numbered ecoregions and oak species found in each. (Map compiled by W.D. Dijak, USDA Forest Service, North Central Research Station, Columbia, Missouri.)

Forest cover types (or simply cover types) are combinations of tree species that tend to spatially reoccur at stand-level scales (e.g. < 100 acres). The resulting categories are thus silviculturally useful in differentiating among different kinds of oak stands. Categorization of United States forests based on defined cover types was begun by the Society of American Foresters in 1929. There are 145 defined cover types in the United States and Canada (Eyre, 1980). These include 31 with ‘oak’ in the cover type name or in the list of species that define the type (Appendices 2 and 3). Of these, 23 oak types occur east and eight occur west of the 100th meridian. In addition, many of the non-oak cover types include one or more oak species as common associates. The geographic extent of individual cover types ranges from tens of millions of acres (e.g. the white oak–black oak–northern red oak cover type of the eastern US) to rel-

atively restricted areas (e.g. the northern pin oak cover type of the upper Lake States and the Mohr oak cover type of Texas and Oklahoma). Other types such as the live oak type of the South and the bur oak cover type in the Great Plains occur within long narrow belts associated with coastal plains and river corridors, respectively. Many of the western oak cover types, especially those in California, form belts that follow the Coastal and Sierra Nevada mountain ranges and foothills surrounding the Central Valley. Oaks occur in environments ranging from extremely wet and humid (e.g. the overcup oak–water hickory cover type of southern flood plains), to mesic (moist) upland forests receiving 50 or more inches of precipitation per year (e.g. the yellow–poplar–white oak–northern red oak cover type), to Mediterranean climates that receive 10 inches or less precipitation per year (e.g. the blue

12

Chapter 1

Fig. 1.3. The geographic distribution of numbers of oak species in eastern United States and Canada. The isolines were drawn from a grid comprised of 78 ⳯ 78 square mile cells within which the number of oak species were counted based on Little’s (1971, 1977) range maps. The greatest concentration of oak species (15 to 20) occurs in the southeastern United States where the ranges of several narrowly distributed species overlap the ranges of several widely distributed species. (Redrawn from Aizen and Patterson, 1990, used with permission.)

oak–digger pine cover type). Oaks occur in even drier climates where they form shrub vegetation such as the chaparral of southern California and the semi-desert scrub woodland vegetation of the interior southwest. Western cover types such as the canyon live oak cover type include closed-canopy stands in the northern part of their range and savanna-like woodlands in the south. Oak forests therefore range from closed canopy upland and lowland forests with trees greater than 120 ft tall to xeric (droughty) scrublands dominated by dwarf trees and shrubs. Some oaks, such as Georgia oak and McDonald oak are confined to very small geographical ranges and a narrow range of habitat conditions. Others such as white

oak are widely distributed and occur over a broad range of climates and habitat conditions. A species’ flexibility in occupying different habitats is implicit in the definition of species niche. The term denotes the specific set of environmental and habitat conditions that permit the full development and completion of the life cycle of an organism (Helms, 1998). The oaks occupy many niches because of the wide range of environmental conditions within which they can collectively occur. However, the niche of an individual species is more limited. Niche differentiation among the oaks and associated species is often evident from the way species segregate along environmental gradients such as the soil moisture gradient (Fig. 1.4). Oaks also differ in their ecologi-


cal amplitude, i.e. the range of habitat conditions that a species can tolerate (Allaby, 1994). The ecological amplitude of a species often forms a bell-shaped curve when illustrated diagrammatically (Fig. 1.4). However, some species, such as bur oak, occur in both bottomlands and dry uplands but are nearly absent at intermediate points along the moisture gradient (Curtis, 1959; Johnson, 1990). The species composition of forests is continually changing as a result of forces both internal (autogenic) and external (allogenic) to the forest. Changes are often gradual and frequently result in the replacement of one tree species by another in the process of ecological succession. The vegetation and other organisms within the forest thus effect autogenic change. For example, shade-tolerant species growing beneath the main forest canopy may gradually replace dominant species of lesser shade-tolerance that are unable to regenerate under their own shade. In contrast, allogenic change occurs as a result of

Importance value

Black oak

13

changes in climate, defoliation by exotic insects and pathogens, the movement of soil by wind and water, or from other forces originating outside the forest. Autogenic and allogenic factors sometimes jointly affect the direction and rate of succession. Moreover, disturbances such as windthrow, insect and disease outbreaks, and timber harvesting can accelerate succession or alter its direction. Although the oaks are relatively intolerant of shade, species vary substantially in this attribute. In some habitats, oaks are vulnerable to successional replacement by more shade tolerant species. Compared to many of their competitors, oak seedlings grow more slowly during their first few years after initial establishment. When young oaks are overtopped and heavily shaded by other vegetation, few survive for very long. On the other hand, the oaks tend to be relatively drought tolerant, and often survive in habitats that limit the development of species of lesser drought tolerance. Oaks also can produce vigorous

Northern red oak White oak

Sugar maple American beech

Bur oak

Xeric

Mesic Compositional index

Fig. 1.4. Changes in the relative importance of six tree species in the upland forests of southern Wisconsin in relation to the regional soil moisture gradient. Species’ importance is quantitatively expressed by an importance value, which is an index of species’ importance based on its frequency of occurrence, density and basal area relative to other species within a stand. Although there is much overlap among species’ importance value curves, no two species behave exactly the same way with respect to the moisture gradient. The length of the gradient spanned by a species’ range of importance values together with the shape of its importance value curve reflects its niche with respect to the gradient. Importance value curves also define the ecological amplitude of a species, i.e. the range of conditions it can tolerate and the magnitude of its importance in relation to the gradient under the prevailing (i.e. relatively undisturbed) stand conditions. The moisture gradient shown is inferred from the species composition of a series of relatively undisturbed stands (see Curtis, 1959). (Adapted from Curtis, 1959, used with permission.)

14

Chapter 1

sprouts that often outgrow competitors. The balance of these factors thus determine the relative permanence of oaks within a given cover type. In the eastern half of the United States, oaks are often relatively permanent members of cover types on drier sites. In the absence of disturbance, many of the pine and oak–pine cover types occurring on dry habitats are successional to oaks because the oaks are somewhat more shade tolerant than the pines. This successional pattern creates silvicultural problems in maintaining pure pine stands in the south and other regions where oaks and pines co-occur (Burns and Barber, 1989). In bottomlands and mesic uplands, shadetolerant or faster growing species often successionally displace the oaks. Such displacement creates silvicultural problems in perpetuating oaks in these forests. The relative permanence of an oak species within a given cover type (i.e. its resistance to successional replacement by other species) is likely to be highly variable if the cover type spans a broad range of environments. For example, the white oak cover type occurs across a wide range of site conditions from dry to moist. Whereas the type tends to be relatively permanent on dry sites, it is successional to other types on the more mesic sites. Cover type designations, although useful, largely fail to consider these and other ecological factors that determine changes in species composition and how those changes vary spatially (e.g. in relation to climate and site quality), and temporally (e.g. in relation to plant succession and disturbance). Consequently, two or more stands representing a single cover type may represent quite different ecologies with respect to the successional status of oaks, physical environment, understorey vegetation, forest regeneration, fauna and other factors. Forest inventories and satellite imagery have been used to describe the geographic distribution of forest types in the United States (e.g. Fig. 1.5). These maps identify broad cover type groups that are aggregates of the stand cover types described above.

Four groupings widely used to delineate oak forests at the regional scale are: the oak–hickory group, the oak–pine forest group, the oak–gum–cypress group (bottomland forests), and the western hardwood group that includes the western oaks as a subset (Fig. 1.5). However, the names commonly applied to the resulting species aggregations can be misleading. For example, hickory is absent throughout much of the northern part of the range delineated as oak–hickory (Fig. 1.5). Moreover, other forest cover types dominated by oaks are also included within the delineated oak–hickory area. The term ‘oak–hickory’ nevertheless is widely used in reporting forest resource statistics at the regional level even though it is an ecologically imprecise term. Oaks also occur as ecologically and silviculturally important components of many nonoak forests (e.g. pine forests and maple–beech–birch forests). In the eastern United States the oak– hickory, oak–pine and the oak–gum–cypress cover type groups collectively covered 187 million acres or 52% of the timberland4 in 1997. That is an increase from 162 million acres and 45% of eastern timberland in 1953 (USDA Forest Service, 2000). At 124 million acres, the oak–hickory group is the largest cover type in the United States. The western oaks are also significant geographically and ecologically. Western hardwood forests (including oaks, tanoak, red alder and aspen) cover 43 million acres or 12% of western forestland. Oaks comprise about 23% of the cubic volume of growing stock trees in the eastern United States and about 1% in the western United States (USDA Forest Service, 2000).

Distribution of oaks by hierarchically classified ecoregions Climate and landform strongly influence the distribution of oaks. Locally, the distribution of oaks is influenced by factors such as physiography, soil moisture and geology.

4 Timberland is forest land that is producing, or is capable of producing, more than 20 feet3 acre–1 year–1 of industrial wood crops under natural conditions, and that is not withdrawn from timber use, and that is not associated with urban or rural development. Currently inaccessible and inoperable areas are included.


(a)

15

Long. 100

240

210

210 330 340 220

250

320 230 260

310

410

Oak–hickory (b)

240

210

210 330 340 220

250

320 230 260

310

410

Oak–gum–cypress (c)

240

210

210 330 340 250

220

320 230 260

310

Western hardwoods 410

Oak–pine

Fig. 1.5. The major areas of oak–hickory, oak–pine, oak–gum–cypress, and western hardwoods (shaded areas) by state and ecoregion Divisions. In the western US, the map shows the composite western hardwood group that includes oaks, tanoak, red alder, cottonwood and aspen. Numbered ecoregion boundaries on the map are from Bailey (1997) and are summarized in Tables 1.2 and 1.3. Generated from Advanced Very High Resolution Radiometer satellite images at a scale of 1 km2 (1990) and an associated system of land cover classification (USDA Forest Service, 1993; Powell et al., 1994). (Map compiled by W.D. Dijak, USDA Forest Service, North Central Research Station, Columbia, Missouri.)

16

Chapter 1

These and other factors have been used to structure a hierarchical ecological classification system (McNab and Avers, 1994; Bailey, 1995, 1997, 1998). This system recognizes the increasing detail necessary to explain the spatial arrangement of forests at increasingly smaller spatial scales (Table 1.1). It thus provides an objective basis for the regional delineation of ecosystems into successively smaller and more homogeneous units. The hierarchical ecological units range in size from continents to a few acres. The larger units are often referred to as ecoregions; the smallest units are often equivalent to forest stands. Domains, Divisions and Provinces form the larger ecoregions (Table 1.1). These are climatic and climatic–physiographic regions that cover millions to tens of thousands of square miles. Provinces are further subdivided into smaller units termed Sections, Subsections, Landtype Associations (LTAs), Ecological

Landtypes (ELTs) and Ecological Landtype Phases (ELTPs). These units range in size from thousands of square miles for Sections to less than 10 acres for some Ecological Landtype Phases. Ecological Landtypes and Ecological Landtype Phases are important silviculturally because they often correspond to individual stands, which are the objects of silviculture. The oaks occur in all three Domains (major climatic regions) of the 48 contiguous states: Humid Temperate, Dry and Humid Tropical (Bailey, 1997). The latter occurs only in the southern tip of Florida. The three Domains are further subdivided into 11 climatic Divisions. Within each Division, mountainous areas with elevational zonation of vegetation are also identified. Although oaks naturally occur in all 11 of the Divisions, within each Division the distribution of the four major oak forest types is closely related to Division boundaries (Fig. 1.5; Tables 1.2 and 1.3).

Table 1.1. Hierarchy of ecological units used to classify forest ecosystems in the United States.a Ecological unit

Scale (reference size)b

Delineating factorsc

Domain

Millions to tens of thousands of square miles (subcontinent)

Macroclimate, ocean temperature and currents, geomorphology

Division

Millions to tens of thousands of square miles (multi-state)

Geomorphology, climate

Province

Millions to tens of thousands of square miles (multi-state, state)

Geomorphology, climate

Section

1000s of square miles (state, multi-county, National Forest)

Geomorphology, climate, vegetation

Subsection

10s to 100s of square miles (multiple counties, National Forest Ranger District)

Geomorphology, climate, vegetation

Landtype association

10s to 1000s of acres (landscape, watershed)

Landforms, species composition of overstorey, soil associations

Ecological landtype

10s to 100s of acres (multiple stands)

Landform, natural vegetative communities, soils

Ecological landtype phase

1 to 10s of acres (stand)

Soils, landscape position, natural vegetative communities

a

Adapted from McNab and Avers (1994), Bailey (1995) and Cleland et al. (1993); see also Figs 1.2, 1.5 and 1.6. b Indicates a familiar unit of comparable size for reference purposes. This reference unit is not used to delineate the ecological unit. c Some of the factors used to distinguish among ecological units at a given level. Classification complexity typically increases with decreasing unit size.


17

Table 1.2. The ecoregion domains, divisions and provinces in the eastern conterminous United States where oaks are found and the principal species occurring in each. Ecoregions from Bailey (1995). Division and province boundaries are shown in Fig. 1.2. Division

Province

– – – – – – – – – – – – – – – – – – – 200 Humid Temperate Domain – – – – – – – – – – – – – – – – – – – 210 Warm Continental M210 Warm Continental Mountains

10 oak species:

220 Hot Continental M220 Hot Continental Mountains

22 oak species:

230 Subtropical M230 Subtropical Mountains

31 oak species:

250 Prairie

211 Mixed deciduous coniferous forests M211a Mixed forest–coniferous forest–tundra, medium M211b Mixed forest–coniferous forest–tundra, high

bear, black, bur, chestnut, chinkapin, n. pin, n. red, scarlet, swamp white, white 221a Broadleaved forests, oceanic 221b Broadleaved forests, continental M221 Deciduous or mixed forest–coniferous forest–meadow M222 Broadleaf forest–meadow

basket, bear, black, blackjack, bur, cherrybark, chestnut, chinkapin, n. pin, n. red, overcup, pin, post, scarlet, shingle, Shumard, s. red, swamp chestnut, swamp white, water, willow, white 231 Broadleaved–coniferous evergreen forests 232 Coniferous–broadleaved semi-evergreen forests M231 Mixed forest–meadow province

Arkansas, bear, black, blackjack, bluejack, bur, Chapman, cherrybark, chestnut, chinkapin, Durand, Georgia, laurel, live, myrtle, Ogelthorpe, n. red, Nuttall, overcup, pin, post, scarlet, shingle, Shumard, s. red, swamp chestnut, swamp white, turkey, water, white, willow 251 Forest-steppes and prairies province 252 Prairies and savannas province

20 oak species:

black, blackjack, bluejack, bur, chinkapin, Durand, live, n. pin, n. red, overcup, laurel, pin, post, s. red, shingle, Shumard, swamp chestnut, swamp white, water, white

– – – – – – – – – – – – – – – – – – – 400 Humid Tropical Domain – – – – – – – – – – – – – – – – – – – 410 Savanna

4 oak species:

411 Open woodlands, shrubs and savanna 412 Semi-evergreen forests 413 Deciduous forests province Chapman, live, laurel, myrtle

– – – – – – – – – – – – – – – – – – – Intrazonal Regions – – – – – – – – – – – – – – – – – – – R Riverine forest

The 11 ecoregion Divisions within the conterminous United States are further subdivided into 44 Provinces (Fig. 1.2). Provinces are delineated based on broad vegetation groups and related regional landforms. Oak forests and woodlands commonly occur in 23 Provinces (Tables 1.2 and 1.3). Province boundaries are useful in delineating oak distributions in some

parts of the United States. For example, Province boundaries correspond with the spatial distribution of the oak forests and woodlands encircling California’s Central Valley. Province boundaries also separate the oak–pine forests of the Piedmont (Province 232) from the wetter oak habitats of the Coastal Plain and the lower Mississippi flood plain (Province 231 and

18

Chapter 1

Table 1.3. The ecoregion domains, divisions and provinces in the western conterminous United States where oaks are found and the principal species occurring in each. Ecoregions from Bailey (1995). Division and province boundaries are shown in Fig. 1.2. Division

Province

– – – – – – – – – – – – – – – – – – 200 Humid Temperate Domain – – – – – – – – – – – – – – – – – – 240 Marine M240 Marine Mountains

3 oak species: 260 Mediterranean

M260 Mediterranean Mountains

13 oak species:

241 Mixed forests M241 Deciduous or mixed forest–coniferous forest–meadow M242a Forest–meadow, medium M242b Forest–meadow, high Oregon white, California black, canyon live 261 Dry steppe 262 Mediterranean hardleaved evergreen forests, open woodlands and shrub 263 Redwood forests M261 Mixed forest–coniferous forest–alpine meadow M262 Mediter. woodland or shrub–mixed or conif. forest–steppe or meadow M263 Shrub or woodland–steppe–meadow blue, California black, California scrub, canyon live, coast live, Dunn, Engelmann, interior live, island live, McDonald, Oregon white, turbinella, valley

– – – – – – – – – – – – – – – – – – 300 Dry Domain – – – – – – – – – – – – – – – – – – 310 Tropical/ Subtropical Steppe 311 Coniferous open woodland and semideserts 312 Steppes 313 Steppes and shrubs 314 Shortgrass steppes M310 Tropical/ Subtropical M311 Steppe or semidesert–mixed forest–alpine meadow or steppe Steppe Mountains 16 oak species: Arizona, canyon live, Chisos, Dunn, Emory, Gambel, Gray, Havard, Lacey, lateleaf, Mohr, sandpaper, silverleaf, Toumey, wavyleaf, turbinella 320 Tropical/ Subtropical Desert

M320 Tropical/ Subtropical Desert Mountains 22 oak species:

330 Temperate Steppe M330 Temperate Steppe Mountains

2 oak species: 340 Temperate Desert M340 Temperate Desert Mountains 3 oak species:

321 Semideserts 322 Oceanic semideserts 323 Deserts on sand M321 Semidesert–shrub–open woodland–steppe or alpine meadow M322 Desert or semidesert–open woodland or shrub–desert or steppe Arizona, chinkapin, Chisos, Dunn, Durand, Emory, Gambel, Graves, gray, Havard, Lacey, lateleaf, live, Mexican blue, Mohr, netleaf, post, sandpaper, silverleaf, Toumey, turbinella, wavyleaf 331 Steppes 332 Dry steppes M331 Forest-steppe–coniferous forest–meadow–tundra M332 Steppe–coniferous forest–tundra M333 Steppe–coniferous forest M334 Steppe–open woodland–coniferous forest–alpine meadow bur, Gambel 341 Semideserts 342 Semideserts and deserts M341 Semidesert–open woodland–coniferous forest–alpine meadow

Gambel, turbinella, wavyleaf


Riverine Forest). Province boundaries are also useful in separating the regions where oaks occur from those where they do not. In contrast to the coarser levels of the classification hierarchy (Domains through Subsections), which have been delineated nationally, classification of the ELT and ELTP levels is incomplete across much of the oak range. Even though classification systems down to the ELT or ELTP have been developed for millions of acres, they include only a small fraction of the total area of oak forests. ELTs or ELTPs are usually mapped in the field based on differences in soils, physiography and vegetation (including herbs and shrubs). The species composition of the herbaceous layer is often used to distinguish among different ELT or ELTP units because of the fidelity of some herbaceous species (‘indicator’ species) to specific biophysical conditions. Accordingly, the presence or absence of one or more indicator species can be used to differenti-

19

ate among otherwise similar ELTs or ELTPs. Shrubs are also sometimes used as indicator species. Compared to the herbaceous layer, the composition of the tree layer often recovers slowly from disturbances. Moreover, the tree component may not recover to its pre-disturbance composition. Joint consideration of physical and biological factors and their interactions provide a basis for identifying ecologically homogeneous land units that are silviculturally relevant and useful in delineating management units (Barnes et al., 1982). Ecological classification provides a broader ecological context for understanding why oaks occur where they do, and how those occurrences change with time, disturbance and other factors. At the broadest scale the oak forests of the United States can be divided into four eastern and two western groups based on species associations, ecological conditions and successional relations (Fig. 1.6).

Pacific Mediterranean–Marine Region

Coast Range

340 260 Sierra Nevada

Central Valley

320 310

Southwestern Desert–Steppe Region

Edwards Plateau Cross Timbers

Forest–Prairie Transition Region

330

240

Driftless Area 210 Northern Hardwood Region 210 Ohio Valley

250

220 Central Hardwood Region

Southern Pine–Hardwood Region 230

Lower Mississippi Ozark Highlands Valley Boston Mountains

Allegheny Plateau Appalachian Mountains Coastal Plain Piedmont Cumberland Plateau Highland Rim Piedmont Coastal Plain

410

Fig. 1.6. The six regions where oaks commonly occur: Northern Hardwood Region; Central Hardwood Region; Southern Hardwood–Pine Region; Forest–Prairie Transition Region, Southwest Desert–Steppe Region; and Pacific Mediterranean–Marine Region. Numbers correspond to Ecoregion Divisions (Figs 1.2 and 1.5) (Bailey, 1997). Not considered by the above regional groupings are the ranges of Gambel and bur oak, which extend into Division 330, and the ranges of Gambel, turbinella and wavyleaf oaks, which extend into Division 340. The shading shows the distribution of oaks from Fig.1.2.

20

Chapter 1

Boundaries between regions follow Division boundaries in the hierarchical ecological classification system. These regional groupings are useful ecologically and silviculturally because they identify areas with broadly similar macroclimates and species associations. Regional differences in the application of silvicultural methods are closely related to corresponding differences in species composition, environmental factors and other ecological conditions. The six forest regions are described below in relation to the Domains, Divisions and Provinces of the hierarchical ecological classification system described in the preceding section. However, the regional designations do not explicitly identify the lowland and riparian forests occurring within them. There, along the major rivers and streams within the Southern Pine–Hardwood Region, the oaks attain their greatest size and growth.

Eastern Oak Forests The Northern Hardwood Region Geographic extent The Northern Hardwood Region includes the northern halves of Minnesota, Wisconsin, Michigan and much of the northeastern United States including New Hampshire, Vermont and Maine in their entirety. It includes two ecoregion Provinces: Mixed Deciduous– Coniferous Forests Province (211) and Mixed Forest–Coniferous Forest–Tundra, High Province (M211b) within the Warm Continental Division (Figs 1.2 and 1.6; Table 1.2). The Region extends 1300 miles from west to east and covers 123 million acres, about three-quarters of which is forested. Braun (1972) called this area the Hemlock–White Pine–Northern Hardwood Region. She recognized two major subsections, the Great Lakes–St Lawrence and the Northern Appalachian Highlands. The western and eastern portions of the Northern Hardwood Region share many of the same species, but they differ ecologically and

silviculturally (Godman, 1985). Those differences are due in part to the influence of the Appalachian Mountains in the eastern part of the Northern Hardwood Region. More than 1.5 million non-industrial private forest owners own approximately half of the forests in the Northern Hardwood Region. Corporate and other private owners hold an additional 25% (Birch, 1996). There are 11 national forests in the region (primarily in the Lake States) that cover 6.5 million acres.

Climate, physiography and soil Precipitation typically ranges from 24 to 45 inches per year although as much as 70 inches occurs in some mountainous areas in the eastern part of the region. Snowfall of 60 to 100 inches per year is common throughout the region. More than 100 inches of snowfall occurs at some of the higher elevations, and snowfall exceeds 400 inches in some locales near the Great Lakes. Mean annual temperature ranges from 35° to 52°F (2 to 11°C) and the growing season lasts from 100 to 160 days (Fig. 1.7) (McNab and Avers, 1994). The region is characterized by low relief with numerous lakes, depressions, morainic hills, drumlins, eskers, outwash plains and other glacial landforms. Variation in the depth and type of glacial deposits and associated heights of water tables are important factors in the identification of silviculturally relevant ELTPs (Fig. 1.8). Elevations in the mountainous areas range from 1000 to 4000 feet with individual peaks exceeding 5000 feet. Valleys in the mountainous areas include outwash plains and lakes resulting from glaciation (Bailey, 1995). Soils have formed in diverse organic and mineral materials including peat, muck, marl, clay, silt, sand, gravel and boulders in various combinations. At lower elevations in New England and along the Great Lakes, Spodosols are common. Inceptisols and Alfisols dominate at lower elevations elsewhere. In the mountainous zones the soils are primarily Spodosols (Bailey, 1995).


Iron mountain, MI

Fort Wayne, IN

Boone, NC

42˚F 30 in.

50˚F 49 in.

52˚F 55 in.

21

Atlanta, GA 62˚F 49 in.

Fargo, ND

41˚F 19 in.

Key West, FL 78˚F 40 in.

Fig. 1.7. Representative climates for selected ecoregion Divisions in the eastern United States. Mean monthly precipitation is shown by the solid lines (right axis) and temperature by dashed lines (left axis). Mean annual values are given above each graph. Division boundaries are shown in Figs 1.2 and 1.5. (Ecoregion and climatic data from Bailey, 1995.)

22

Chapter 1

ELTPs on dry ice-contact and sand hills

ELTPs on outwash plains 1

10

11

12

20

21

ELTPs on mesic ice-contact and sand hills 34

35

24

25

WATER TABLE

WATER TABLE

32

22

ELTPs on herb-poor moraines 37

40

42

43

WATER WATER

Fig. 1.8. Ecological landtype phases (ELTPs) for the upland forests of the Huron–Manistee National Forests in the lower peninsula of Michigan (Province 211: Mixed Deciduous–Coniferous Forests Province). Site productivity generally increases with increasing ELTP value. ELTP 1: Northern pin oak/white oak – Deschampsia type; ELTP 10: Black oak/white oak – Vaccinium type; ELTP 11: Black oak/white oak – Vaccinium type with loamy sand to sandy loam bands in substrata; ELTP 12: Black oak/white oak – Vaccinium type with perched water table at 6–15 ft; ELTP 20: Mixed oak/red maple – Trientalis type; ELTP 21: Mixed oak/red maple – Trientalis type with loamy sand to sandy loam bands in substrata; ELTP 22: Mixed oak/red maple – Trientalis type with perched water table at 6–15 ft; ELTP 24: Mixed oak/red maple – Trientalis type with perched water table at 3.5–6 ft; ELTP 25: Mixed oak/red maple – Trientalis type with coarse loamy substrata; ELTP 32: Northern red oak/red maple – Viburnum type with perched water table at 6–15 ft; ELTP 34: Northern red oak/red maple – Viburnum type with perched water table at 3.5–6 ft; ELTP 35: Northern red oak/red maple – Viburnum type with fine loamy substrata; ELTP 37: Northern red oak/red maple – Desmodium type with sandy loam over fine loamy substrata; ELTP 40: Sugar maple/beech – Maianthemum type; ELTP 42: Sugar maple – Maianthemum type with perched water table at 6–15 ft; ELTP 43: Sugar maple/northern red oak – Maianthemum type with fined texture substrata. (Adapted from Cleland et al., 1993.)


Forest history The forests of the Northern Hardwood Region were strongly influenced by the aboriginal people who lived there. For thousands of years before the arrival of Europeans, Native Americans used fire and land clearing to shape the forest to meet their needs. Accounts of early European settlers indicate that Native Americans burned large portions of the landscape each year (Pyne, 1982). Dry fuels in the late spring before ‘greenup’ and again after leaf fall during ‘Indian summer’ provided favourable conditions for burning. Fires often eliminated forest understorey layers, which in turn encouraged the growth of edible berries and increased forage that attracted edible wildlife. A combination of fire and tree cutting or girdling also was used by Native Americans to create and maintain openings for cultivated crops (Pyne, 1982). Maintaining an open understorey condition by burning also helped defend Indian villages from surprise enemy attacks. Repeated burning helped maintain large areas of oak savannas and barrens. Topography greatly influenced the spatial distribution of fires, and the frequency and intensity of burning were lower on wet or mesic sites and at higher elevations. This produced a landscape mosaic of diverse species composition. The land clearing and burning practices used in New England were carried westward by settlers who immigrated to the Lake States. Oaks were known as indicators of fair to good conditions for agriculture. Oak forests therefore were often girdled, felled and burned in preparation for agriculture. Over time, enormous areas in the Northern Hardwood Region were cleared for agriculture. Ultimately, however, it was logging that had the greatest impact on the region’s forests. Logging gradually accelerated with the influx of Europeans to New England in the 17th century. Demands for forest products were initially modest in the developing agrarian society. Local timber harvesting supplied wood for homes, barns, fences, heating and cooking, and making potash and tannin. Forests were considered more

23

an impediment to agriculture than a valued resource. However, this changed with industrialization during the 19th century (Williams, 1989). The industrialization of America provided the capacity and economic incentive to exponentially increase lumber production from less than 1 to nearly 45 billion board feet between 1800 and 1900. That lumber, which was principally white pine and other softwoods, came primarily from the Northern Hardwood Region. In 1839, 30% of the nation’s lumber (on a value basis) came from New York. Combined, New York, Pennsylvania, New Jersey and New England produced two-thirds of the nation’s lumber. As supplies of white pine dwindled in the eastern part of the region, timber production moved westward. Although New York reached peak production in 1849, the northeastern states together had by then dropped to half of the national lumber output. The shift in lumber production from the Northeast to the Lake States occurred between 1840 and 1860. Lake States harvest reached a peak of 10 billion board feet annually in 1889. Relatively level terrain, easy access and high demand facilitated the rapid rise in timber harvesting across the Lake States. By 1940, Lake States production dropped below a billion board feet as the timber industry moved south (Williams, 1989). Although white pine was the preferred species, oaks and other hardwoods were utilized locally where they were abundant. Oaks were in less demand by the logging industry. Because of their high density, oak logs could not be floated down rivers as easily as white pine and required overland transportation to avoid losses (Williams, 1989). Over time, repeated timber harvesting removed the trees of greatest economic value leaving behind stands of inferior quality and composition. Farmers emigrating westward subsequently completed the land clearing. As a preferred fuelwood, the oak’s utilization for that purpose significantly affected the region’s forests during the early agricultural period. A colonial family used 20–60 cords of wood annually for

24

Chapter 1

heating and cooking. Although the per capita volume of wood used for fuel decreased over time, the total volume increased because of a growing population. In 1880, more than half of America’s energy needs were still met with fuelwood (Whitney, 1994). Iron furnaces in the region were fired with hardwood charcoal. Less than 1% of the fuelwood burned from 1800 to 1930 was used to produce charcoal for iron smelting, but large areas of forest surrounding iron furnaces were greatly affected (Whitney, 1994). A typical 18th-century smelting operation consumed 100 acres of forest annually to produce charcoal. Because forest regrowth could be repeatedly harvested for this purpose every 25 years, about 2500 acres of forest were required to sustain production of the ironworks. By the late 19th century, charcoal production for large ironworks required annual harvests of 1000– 2000 acres, and some large companies owned 100,000 acres of forest adjacent to their smelters for that purpose (Whitney, 1994). The large clearcuts surrounding the ironworks radically changed the age structure of the forests and also influenced their species composition. Harvesting hardwoods for fuel favoured the development of hardwood sprouts and increased the relative proportion of hardwood trees in areas that were not converted to agricultural land (Whitney, 1994). During the last half of the 19th century, many cleared acres marginally suited to agriculture were abandoned and subsequently reverted to forest. During the latter half of the 20th century, the combination of abandoned agricultural lands and natural regeneration of cutover lands resulted in large increases in timber volumes throughout the region. In New England, forest volumes increased by 16% between 1970 and 1982 (cubic foot basis) (Seymour, 1995). The increase was predominantly in hardwoods that as a group increased in volume by 24%. The increase in oak volume (16%) was low relative to other hardwoods. In 1992, net annual growth in the Northern Hardwood Region remained at more than twice the annual harvest (Powell et al., 1994).

Oaks as components of the region’s forests The forests of the Northern Hardwood Region today are dominated by more than half a dozen recognized northern hardwood forest types comprising various combinations of sugar maple, red maple, beech, paper birch, yellow birch and eastern hemlock. Although northern red oak typically occurs as a minor component within these types, it sometimes forms pure or nearly pure stands (Fig. 1.9A). Wherever it occurs, it is a valuable and desirable species for timber, acorn production and species diversity. White, black, northern red and chestnut oaks also occur in the southern portions of the region. Oaks are thus a relatively small component of northern hardwood forests. They are most abundant and attain their best development in the southern parts of the region including New York, Massachusetts, northern Pennsylvania, and central and southern Minnesota, Wisconsin and Michigan. There the oak and mixed-hardwood forests grade into the oak–hickory forests of the Central Hardwood Region. In both New England and the Lake States, about 11% of the forestland is classified as oak–hickory or oak–pine. These forest types include 17 billion cubic feet of growing stock (inclusive of oaks and associated species). Throughout the Northern Hardwood Region, sugar maple, red maple and aspen are the most abundant hardwoods. Conifer forests (red and eastern white pines, spruce and balsam fir) also exceed oak in acreage and volume (Powell et al., 1994). Oaks often reach their greatest density on sites that have been repeatedly disturbed by fire, timber harvesting and other events. After burning or timber harvesting, oaks originating from vigorous seedling sprouts and stump sprouts often dominate stands. However, in the absence of disturbance the oak forests of the Northern Hardwood Region are usually successional to other hardwoods on all but the poorest sites. On the poorer sites, oaks are often relatively permanent members of the forest.


25

A

B

Fig. 1.9. (A) A 130-year-old stand of northern red oak in the Northern Hardwood Region of northern Wisconsin (Province 211: Mixed Deciduous–Coniferous Forests Province; Southern Superior Uplands Section). The absence of oak reproduction and a sparse sub-canopy of shade tolerant red and sugar maples are indicators of what is likely to eventually replace the oaks in the absence of disturbance. (B) Xeric northern pin oak–white oak/Deschampsia type (see Fig. 1.8) on deep outwash sand in the northern lower peninsula of Michigan (Province 211: Mixed Deciduous–Coniferous Forests Province; Northern Great Lakes Section). This oak stand is mixed with jack pine; oak site index is ≤50 ft. (USDA Forest Service, North Central Research Station photographs.)

There, oaks frequently invade and successionally replace established pine stands (Seymour, 1995). The loss of the American chestnut to chestnut blight fungus in New England oak forests began in the early 1900s (Fig. 1.10). This increased the relative importance of oaks because oaks often captured the growing space vacated by dying American chestnuts.

Today, the single-tree selection method of silviculture is often applied to northern hardwood forests dominated by shade tolerant species such as sugar maple. This practice focuses on maintaining stands of high quality trees while largely relying on the natural regeneration of shade tolerant species to sustain the silvicultural system. Although this system favours the develop-

26

Chapter 1

pine, form relatively stable forest types of low productivity (Fig. 1.9B).

The Central Hardwood Region Geographic extent

Fig. 1.10 A standing dead American chestnut (minus bark). Chestnut was a common associate and dominant member of eastern oak forests throughout the Appalachians from Maine to Alabama and westward to Missouri. The chestnut blight, which decimated the species throughout its range, permanently altered the ecology of eastern oak forests. The blight was first identified in New York in 1904. Fifty years later it spanned the entire natural range of chestnut. Oaks and associated hardwoods quickly captured the growing space vacated by dead and dying chestnuts. (USDA Forest Service, North Central Research Station photograph.)

ment of high quality oaks in stands where oaks are already present, regenerating oaks beneath the relatively closed canopies of selection forests is usually difficult in this region. On the poorer sites, oaks may develop beneath a pine overstorey and eventually displace the less shade tolerant pine through natural succession or the exposure of oak reproduction in the understorey to full light after timber harvest. On deep sandy soils of the upper Lake States, stands of northern pin, black and white oaks, often mixed with jack

The Central Hardwood Region includes the predominantly deciduous broadleaf forests of Central United States. The region lies entirely within the Humid Temperate Domain. The region includes the two Hot Continental Divisions (Divisions 220 and M220), and intergrades with the eastern part of the Forest–Steppes and Prairies Province (251) of the Prairie Division (250) (Figs 1.2 and 1.6; Table 1.2). The Hot Continental Division is subdivided into two provinces: Broadleaf Forests, Oceanic (221a); and Broadleaf Forests, Continental (221b). The Hot Continental Mountains Division (M220) also is divided into two provinces: Deciduous or Mixed Forest–Coniferous Forest–Meadow (M221), and Broadleaf Forest–Meadow (M222). The Northern Hardwood Region, the Southern Pine–Hardwood Region, the Forest–Prairie Transition and the western edge of the Appalachian Mountains bound the Central Hardwood Region. The Central Hardwood Region extends 1200 miles from southwest to northeast and covers approximately 220 million acres; about half the region is forested. Approximately threequarters of the forest area in the Central Hardwood Region is in non-industrial private ownership. Within that ownership, most holdings are 50 acres or smaller (Birch, 1996). There are seven national forests in the region comprising about 4 million acres distributed across the southern half of the region in Arkansas, Missouri, Illinois, Indiana, Ohio, Kentucky and Tennessee.

Climate, physiography and soil The climate in the Central Hardwood Region is hot continental with warm summers and cold winters. Mean annual temperature ranges from 40 to 65°F (4–18°C), with the warmer temperatures in the south.


Annual precipitation ranges from 20 inches in the northwest to 65 inches in the southeast and reaches as much as 80 inches on some Appalachian peaks (Fig. 1.7). Precipitation occurs throughout the year, but tends to be somewhat greater in spring and summer. Droughts may occur during the summer when evapotranspiration is high. Frost-free periods range from 100 days in the northern Appalachians to 220 days in the southern part of the region (Bailey, 1995). Topography is diverse in this region. Elevations in the Appalachian Highlands (province M221) range from 300 to 6000 ft with as much as 3000 ft of local relief. Further west (province 221a), the hills and low mountains of the dissected and uplifted Appalachian Plateau (including the Allegheny and Cumberland Plateau) range from 1000 to 3000 ft in elevation. In the western half of the Central Hardwood region (province 221b), most of the land is rolling but varies from extensive, nearly level areas to areas like the Ozark Highlands where relief reaches 1000 ft. Most of the northern portions of this province were glaciated with the exception of the driftless area of southwestern Wisconsin and adjacent states. Major soils are Alfisols, Inceptisols, Mollisols and Ultisols (Bailey, 1995).

Forest history The utilization and exploitation of forests in the Central Hardwood Region has passed through various historical phases (Hicks, 1997). Even before the arrival of Europeans, humans influenced the nature and extent of the region’s forests (Whitney, 1994). The use of fire to control vegetation by Native Americans significantly influenced the extent and character of presettlement forests (Pyne, 1982; DeVivo, 1991; Olson, 1996). These human-caused alterations of the landscape continued for thousands of years before the arrival of Europeans (Hicks, 1997). After settlement by Europeans, human impacts on the forest expanded and intensified. Burning, grazing, exploitative timber harvesting and

27

clearing of forests for agriculture occurred on an unprecedented scale. These practices occurred about 200 years earlier in the eastern part of the Central Hardwood Region than in the western part. Historically, different human disturbances were further confounded by intrinsic ecological differences among oak forests within the various ecoregion provinces. Each subregion of the Central Hardwood Forest has its own unique combination of disturbance history, climate, physiography, soils, species associations and successional possibilities. This complicates generalizing the application of silvicultural methods to oak forests across the region. As in the Northern Hardwood Region, the loss of American chestnut to chestnut blight increased the relative proportion and importance of oaks throughout the region. Shortly after 1900, the disease became epidemic and within 40 years it had invaded the entire natural range of the chestnut (Kuhlman, 1978). The loss represents one of the greatest recorded changes in a natural population of plants caused by an introduced organism (Liebhold et al., 1995). The chestnut comprised 25% of the eastern hardwood forest that covered 200 million acres. In the Appalachians, it was the most ecologically and economically important tree species (Kuhlman, 1978). There and in other regions, it grew faster and taller than associated oaks. Before the blight, chestnut was especially important in moist upland forests where it sprouted vigorously and often increased in dominance after logging. In 1900, half the standing timber in Connecticut was chestnut, which was largely comprised of young stands of stump sprout (coppice) origin (Smith, 2000). Although American chestnut provided only about 1% of the nation’s hardwood lumber even at the peak of its importance, its loss (beginning in the early 1900s) had a significant impact on local economies in the Appalachians. There its nuts and bark (for tannin) provided scarce cash income, and its wood was valued for a variety of uses (Youngs, 2000). The practice of silviculture in the Central Hardwood Region dates back to the

28

Chapter 1

genesis of North American forestry in the late 19th century (Fernow, 1911; Pinchot, 1987). From then until the 1960s, the major emphasis was on uneven-aged silviculture (Roach, 1968). During the 1960s, the emphasis shifted to even-aged silviculture, especially clearcutting, and this emphasis persisted for about 20 years (Roach and Gingrich, 1968; Johnson, 1993a). Where applied, hardwood silviculture in the region usually follows the ‘ecological model’, which relies on the existing forest vegetation and its natural regeneration capacity. Silvicultural prescriptions are usually focused on controlling stand structure and species composition using cutting methods such as those recommended by Roach and Gingrich (1968). This approach contrasts with the more intensive ‘agronomic model’ of silviculture based on artificial regeneration, the introduction of improved genotypes, use of herbicides and fertilizer, prescribed burning, and other intensive cultural methods like those commonly used in the silviculture of pine monotypes in the south and elsewhere. Where applied, silviculture in the Central Hardwood Region has usually focused on growing high quality sawtimber. During the course of stand management (but before final harvest of even-aged stands), this requires ‘leaving the best’ and ‘cutting the worst’ at each harvest. In evenaged silviculture, each timber harvest concentrates on removing small, sub-canopy trees and poor quality trees in the main canopy, with concomitant attention to species composition. Similarly, in unevenaged silviculture, timber removals are concentrated on poor quality trees, but cutting occurs across a wide range of diameter classes in order to create and maintain the uneven-aged stand structure. In both systems the objective is the improvement of the quality and the economic value of the residual stand. Today, only a small fraction of the forests of the Central Hardwood Region receive systematic silvicultural treatment. This is largely due to the pattern of forest ownership, which is characterized by numerous small tracts owned by private

individuals. Many forest owners are uninterested in silviculture or lack information on its benefits (Bliss et al., 1994, 1997; English et al., 1997). Consequently, the systematic application of silviculture has largely been limited to industrial forests and public lands. The predominant methods of timber harvesting on private lands are probably commercial clearcutting and other forms of high-grading. Not to be confused with silviculturally prescribed methods, these methods consist of harvesting all trees with commercial value without regard to regeneration needs and future stand condition. Such malpractice typically leaves stands of highly variable residual stocking comprised of trees of poor vigour, low quality and undesirable species composition. These practices persist and continue to impact negatively on the quality of the region’s forests. Nevertheless, annual forest growth for the region exceeds annual harvest, and total standing volume of timber has increased steadily since the 1950s (Powell et al., 1994).

Oaks as components of the region’s forests The predominant oaks are black, white, scarlet, chestnut, post, northern red, southern red and bur oak (Fig. 1.11). These species typically occur in various combinations with hickories, sassafras, flowering dogwood, blackgum, black cherry, red maple, and other upland oaks and deciduous tree species. The Ozark Highlands Section of the region, which covers southern Missouri, and parts of northeastern Oklahoma, northern Arkansas and southwestern Illinois, comprises one of the largest contiguous areas dominated by the oak–hickory association in the Central Hardwood Region. Many oak–hickory forests of today may have originated from extensive fire-maintained oak savannas of the presettlement period; these formed closed canopy forests when fires were suppressed (Johnson, 1993a; Olson, 1996). The oak cover types of the Central Hardwood Region include various combinations of oaks, hickories and other tree


29

Fig. 1.11 A mature black–northern red–white oak stand on a good site in the Central Hardwood Region of southeastern Ohio (Province 211a: Broadleaved Forests, Oceanic Province). (USDA Forest Service, North Central Research Station photograph.)

species that vary geographically (Appendix 2). Although hickories are common and persistent members of this forest type, they seldom represent more than a small proportion of trees in the main canopy of a mature forest (Braun, 1972). Oak–hickory forests develop on relatively dry sites where oaks persist as dominant members of the forest through successive disturbance events. This persistence is facilitated by the oaks’ drought tolerance and by light intensities in dry ecosystems that are sufficient for the regeneration of the relatively shade-intolerant oaks (Bourdeau, 1954; Carvell and Tryon, 1961; Abrams, 1990). Oaks and hickories are found together on the drier sites throughout the region and comprise a commonly occurring species association. These forests dominate the landscape in the western part of the region. Elsewhere, oaks and hickories as a group commonly occur on dry ridges and southfacing slopes. On the more mesic sites, oaks are often interspersed with other hardwoods. Slope position and aspect strongly influence the spatial distribution of these forests and are thus useful in defining ELTPs in much of the region (Figs 1.12 and 1.13).

From southern Illinois eastward and in northern Arkansas, the more mesophytic forests of the Central Hardwood Region generally include more complex species mixtures than found in drier oak forests (Fig. 1.14). Although oaks commonly share dominance with non-oaks on these sites, in the absence of recurrent fire and grazing the oaks are often successionally displaced by more moisture-demanding and more shade tolerant non-oaks (Jokela and Sawtelle, 1985; Lorimer, 1985, 1989; Nowacki et al., 1990; Abrams, 1992). Understoreys of these stands are typically lacking in oak reproduction, especially large oak seedling sprouts. Over time, the dominance of oaks decreases while the proportion of non-oaks increases. The latter include various combinations of maples, American beech, black cherry, white ash, American basswood and yellowpoplar. Timber harvesting may accelerate the successional replacement of the oaks (Abrams and Nowacki, 1992). Diverse mixtures of hardwoods are common throughout the Ohio Valley, the Cumberland Plateau and Highland Rim areas of Tennessee and Kentucky, the Appalachian and Allegheny Plateau of

30

2.1, 2.2

1.3 sinkholes

Chapter 1

1.1, 1.2

1.1 Ultic rocky, 1.2 Silty RO Summits 2.1 Ultic, rocky RO/UG Shoulder (some 2.2)

3.1 Exposed, rocky RO/UG Ultic backslope (some 3.2 Alfic) 7.1 to 7.3 Cherty, non-cherty and glade, variable depth to dolomite 2.4 UG Alfic crypt reef bench

3.1, 3.2

Roubidoux (RO)

4.1 Protected RO/UG Ultic backslope (some 4.2 Alfic) 8.1 Cherty protected, variable depth to dolomite 2.3 UG Ultic crypt reef bench 4.1, 4.2

Upper Gasconade (UG)

Cryptozoan Reef

Upper Gasconade (UG) 10 Footslope

10 Footslope Exposed aspects 135–315˚

Protected aspects 315–135˚

Fig. 1.12. Ecological landtype phases (ELTPs) for the Ozark Highlands of Missouri (Province 221b: Broadleaved Forest, Continental Province; Ozark Highlands Section, upland ELTPs in the Current, Eleven Point and Black River Landtype Associations). Aspect, landform and bedrock geology are factors in the classification system. ELTP 1.1, 1.2, 1.3, 2.1, 2.3 and 3.1: pine–oak/Vaccinium dry ultic (chert) woodland; ELTP 2.2 and 3.2: mixed oak–pine/Desmodium, Vaccinium dry-mesic alfic (chert) woodland; ELTP 4.1: mixed oak–hickory/dogwood/Desmodium dry-mesic ultic (chert) forest; ELTP 4.2: mixed oak (white, red) dogwood dry-mesic alfic (chert) forest; ELTP 7.1: post oak (blackjack oak, pine) bluestem xeric chert woodland; ELTP 7.2: red cedar–hardwood/redbud dry dolomite woodland; ELTP 7.3: bluestem, Missouri coneflower dolomite glade; ELTP 8.1: mixed oak–sugar maple/redbud dry-mesic dolomite forest; ELTP 10: mixed oak (white)/dogwood drymesic alfic (chert) footslope forest. (From Nigh et al., 2000, used with permission.)


31

Brown County Hills Subsection

Crawford Upland and Escarpment Subsections

Fig. 1.13. Ecological landtype phases (ELTPs) for the forests of the Brown County Hills, Crawford Upland and Escarpment Subsections of southern Indiana (Province 221b: Broadleaved Forests, Continental Province; Interior Low Plateau, Shawnee Hills Section). Oaks and pines typically dominate the exposed (hotter) aspects whereas sugar maple, American beech, yellow-poplar and other shade tolerant hardwoods dominate the protected (cooler) aspects. (Adapted from Van Kley et al., undated.)

32

Chapter 1

Fig. 1.14. A large white oak (47 inches dbh) in Dysart Woods in southeastern Ohio (Province 221a: Broadleaved Forests, Oceanic Province). This 55-acre old-growth oak forest is dominated by white and northern red oaks and is the largest known remnant of the original mixed mesophytic forest of the Central Hardwood Region in southeastern Ohio. (USDA Forest Service, North Central Research Station photograph.)

western West Virginia and western Pennsylvania, the southern Lake States, and other parts of the region. Specific combinations of canopy dominants often form distinct geographic species groupings. Examples include the beech–maple forests of central Indiana and eastern Ohio, the maple–basswood–northern red oak forests of the driftless area of southwestern Wisconsin, and the black cherry–ash– yellow-poplar forests of the Allegheny Plateau of Pennsylvania. Toward the eastern end of the region, eastern white pine and eastern hemlock may be locally important members of mixed hardwood forests. Mixtures of oak and mesophytic species also occur in northern Arkansas in the Broadleaf Forest–Meadow Province (M222)

(Fig. 1.2; Table 1.2) (Braun, 1972). However, unlike the mixed mesophytic forests further to the east, yellow-poplar is absent. These are the most mesophytic forests in the western end of the region. Some of these combinations are formally recognized as cover types (Eyre, 1980); others form mixtures that are only locally distinguished silviculturally. It is within these mesic, mixed hardwood stands that northern red oak, one of the most commercially valuable tree species of the region, reaches its best development. It is also within these forests that the oaks are also the most difficult to regenerate silviculturally (Carvell and Tryon, 1961; Arend and Scholz, 1969; Trimble, 1973; Loftis, 1988; Johnson, 1993b, 1994a,b). Mesophytic mixed hardwood forests generally occur where oak site index (chapter 4) is ≥ 65 ft at an index age of 50 years. Oak–pine mixtures occur most frequently in the southern and eastern parts of the region and are closely correlated with fire and succession in old fields, heavily disturbed hardwood stands, and pine plantations. Oak–pine mixtures represent an early- to mid-stage in the succession toward oak–hickory or mixed hardwood forests. In the absence of fire or other disturbances, oak–pine forests may change successionally from predominantly shortleaf pine, pitch pine or Virginia pine to hardwoods as the more shade tolerant hardwoods replace the intolerant pines (Cunningham and Hauser, 1989; Sheffield et al., 1989; Smith et al., 1989; Orwig and Abrams, 1994). The oak–pine mixtures are important for maintaining biodiversity as well as economic timber production (Phillips and Abercrombie, 1987; Cooper, 1989; Kerpez and Stauffer, 1989; Leopold et al., 1989). Consequently, there is increasing interest in methods to create and maintain oak–pine forests (Waldrop, 1989). Specific combinations of oaks and pine vary with subregion and site quality. Because the pines tend to be associated with the driest (xeric) sites, the associated oaks often include species such as post oak and blackjack oak. On somewhat less xeric


sites, pines are commonly associated with black, white, scarlet, southern red or chestnut oaks. In the extreme northwestern part of the region in Minnesota and Wisconsin, jack pine and northern pin oak commonly occur together. Stands of eastern redcedar are closely affiliated with the oak–pine mixtures. Eastern redcedar is a common invader of old fields and glades (Lawson, 1990). It may eventually form dense pure stands if succession is allowed to progress unimpeded by disturbance. However, such stands are short-lived. As the redcedar matures and forms canopy gaps conducive to hardwood or pine regeneration, stands may succeed to oak–pine and oak–hickory mixtures.

The Southern Pine–Hardwood Region Geographic extent The Southern Pine–Hardwood Region includes broadleaved forests, conifer forests and various hardwood–pine mixtures. The region includes the two Subtropical Divisions (230 and M230) of the Humid Temperate Domain (Figs 1.2 and 1.6; Table 1.2). The region covers approximately 270 million acres of which 60% are forested. The extent of the Southern Pine–Hardwood Region is best illustrated by the joint ranges of the oak–pine and oak–gum cypress forest types (Fig. 1.5B and C). The region extends 1300 miles from eastern Texas to Virginia and occurs in a band extending 200–400 miles inland from the coast. At its northern boundary, the Southern Pine–Hardwood Region meets the Central Hardwood Region. Nearly 90% of the forest area in this region is privately owned. Four million non-industrial private forest owners control about 60% of all timberland. About 45% of this ownership is comprised of tracts smaller than 100 acres (Birch, 1996). The 25 national forests in the region comprise 9 million acres of timberland (Powell et al., 1994).

33

Climate, physiography and soil Annual precipitation in the region ranges from about 40 to 60 inches and is well distributed throughout the year. Mean annual temperature ranges from 60 to 70°F (16 to 21°C) and the growing season from 200 to 300 days (Bailey, 1995) (Fig. 1.7). The Southern Pine–Hardwood Region includes four major physiographic regions: the Piedmont (Province 232), the Coastal Plain (Province 231), the Interior Highlands (Province M231) and the lower Mississippi Valley (Riverine Intrazonal Province (R)) (Fig. 1.2). Gentle slopes characterize 50–80% of the area. Elevations range from sea level to 600 ft in the Coastal Plain, 300–1000 ft in the Piedmont, and up to 2600 ft in the Ouachita Mountains of the Interior Highlands. Numerous low-gradient streams, lakes, swamps and marshes characterize the flat Coastal Plain. The wet habitats along the Coastal Plain, the Mississippi Valley and other major rivers support bottomland forest types that are largely absent from the Piedmont and Interior Highlands. The principal soil groups are Ultisols, Spodosols, Vertisols and Entisols, all of which tend to be low in fertility. Exceptions are the Inceptisols, which occur in the alluvial bottoms of the Mississippi River (Bailey, 1995).

Forest history Here, as in other regions, fire greatly impacted the early forests. Fire was an essential tool for maintaining agricultural openings, eliminating brush and hardwood reproduction from pine forests, and increasing forage for grazing. Native Americans regularly burned the forests where they lived. Increased burning associated with European settlement increased the proportion of pine in the region relative to earlier periods (Pyne, 1982; Skeen et al., 1993). Fire was combined with land clearing to open the hardwood forests of the south for agriculture. Even today, forest burning is a prominent practice throughout the region.

34

Chapter 1

In the Piedmont and alluvial river bottoms, vast areas were cleared for agriculture before industrial logging peaked in the region (Sargent, 1884; Hodges, 1995). Logging and the production of naval stores began on a small scale in the 1600s. But by 1880, forests accessible by water and close to population centres were heavily cut over. Charles Mohr noted the rapidity at which the cypress swamps were being logged in some localities and the apparent lack of forest regeneration. He observed that ‘the large number of logs harvested shows clearly with what activity the destruction of these treasures of the forest is being pushed; and the reports, as of heavy thunder, caused by the fall of the mighty trees, resounding at short intervals from near and far, speak of its rapid progress’ (Sargent, 1884). However, Mohr also noted that immense areas of pine forest remained unaffected by logging and that many former hardwood forests that were earlier cleared for agriculture had reverted to pine after their abandonment. The South did not become the centre of the US logging industry until shortly after logging peaked in the Lake States in 1890. By 1900, lumber production in the South exceed that of the Lake States and by 1910 the South produced half of all US lumber. The movement of large lumber companies to the Southern Pine–Hardwood Region coincided with technological advances that increased the speed with which logs could be removed from the woods and transported to the mills (Williams, 1989). Steam powered stationary skidders and loaders were mounted on boats and railcars. As rail lines were extended into the southern forest, logging trains followed and systematically removed virtually all timber within the long reach of a cable skidder mounted on a rail car. The joint enterprise of rail construction and logging greatly accelerated the harvest of southern pines (Williams, 1989). By 1925 southern lumber production began to decline and western production increased. Following the Great Depression, timber production in the South never returned to the levels of 1910–1930, and

the bulk of US timber production moved west. The subsequent establishment of southern paper mills coupled with successful fire prevention and a reduction in openrange grazing accelerated the reforestation of one million cutover acres. This gave rise to the South’s ‘third forest’ which today again produces a greater volume of wood than any other region of the United States.

Oaks as a component of the region’s forests The oak forests in this region can be divided into upland and lowland types. Were it not for the complex spatial intermingling of upland and lowland forests, they could be treated as two ecologically distinct regions. The upland and bottomland oak forests of the region differ substantially in species composition, ecology and the application of silvicultural practices. The Southern Pine–Hardwood Region today includes about 172 million acres of timberland. Of the broadly defined oak forests recognized in national inventories, the oak–hickory group occurs on 55 million acres or one-third of the region’s timberland. Oak–gum–cypress and oak–pine each occur on an additional 16% of the timberland. Thus, oak forests collectively cover more than 60% of the region. Loblolly-shortleaf pine and longleaf-slash pine make up most of the remaining forest acreage. A more detailed cover type classification (Eyre, 1980) recognizes 63 cover types that occur within the region (Walker, 1995) – 15 of those include oaks as primary species, and several others include oaks as important associated species (Appendix 2). Southern silviculture has largely focused on pine, especially on industrial forestlands. There, intensive silviculture is commonly practised to maximize timber and wood fibre yields through site preparation, planting genetically improved seedlings, frequent thinning, prescribed burning, and the use of fertilizers, herbicides and pesticides. However, annual softwood removals are nearly equal to annual growth and may soon exceed annual growth (Walker, 1995).


The importance of pine in the Piedmont is related to the region’s history – the historical sequence of lumbering, land clearing and farming deforested large areas that were abandoned before 1930 and burned frequently. This disturbance favoured the establishment of pine forests, which greatly increased in acreage relative to other species. Oaks and other hardwoods occur in most natural southern pine stands, and on these sites they increase in importance through succession. Fire suppression, silvicultural thinnings and partial harvests often accelerate this trend (Skeen et al., 1993). The large oak–hickory acreage in the region, the increasing hardwood volumes in pine–hardwood mixtures, and the nearly full utilization of the annual pine growth in the region has recently shifted the utilization of the region’s forests towards the hardwoods. Much of this change has resulted from the utilization of hardwood chips for paper production and composite products. Chips can be made from low quality, small diameter (> 4 inches) hardwoods. This technology created new markets for the abundant low-quality trees that previously had been considered a silvicultural liability. However, this utilization capability has also raised concerns about the potential for overutilization of hardwoods, especially well formed, small hardwood trees that comprise the future hardwood growing stock for solid hardwood products. The region’s oak–hickory forests attain their best development along the border separating the Southern Pine–Hardwood Region and the Central Hardwood Region. Closely related to the oak–hickory forests are mixtures of oak and pine (Fig. 1.15). These mixed forests are increasingly recognized for their importance in maintaining forest biodiversity and their historical importance in the region. The oak–pine type, which occurs on 16% of the timberland of the region, rates high in aesthetic appeal and species richness compared to even-aged pine stands. However, relatively little is known about the longterm management and productivity of

35

Fig. 1.15. A black oak–white oak–shortleaf pine stand in the Ozark Highlands of Missouri (Province 221b: Broadleaved Forests, Continental Province; Ozark Highlands Section). (USDA Forest Service, North Central Research Station photograph.)

oak–pine mixtures for lumber, fibre or other values. In the absence of disturbance, the oaks tend to successionally displace the pines and harvesting the pine often accelerates the process. Southern bottomland hardwoods commonly include 11 species of oaks (cherrybark, Delta post, laurel, Nuttall, overcup, pin, Shumard, swamp chestnut, water, white and willow oaks) (Hodges, 1995). These oaks occur in mixture with other bottomland species along the major rivers of the Coastal Plain as well as the lower reaches of the Mississippi, Arkansas, Missouri, Ohio and Wabash Rivers (Fig. 1.16). In total, southern bottomland hardwoods cover more than 27 million acres (16% of the region’s timberland) and are physiographically and ecologically distinct from surrounding upland oak–hickory, oak–pine and pine forests.

36

Chapter 1

Hodges (1995) reduced the number to three types (cottonwood–willow, baldcypress–tupelo, mixed bottomland hardwoods), and the USDA Forest Service usually reports only two types (oak–gum–cypress, elm–ash–cottonwood) in regional summaries.

The Forest–Prairie Transition Region Geographic extent

Fig. 1.16. A cherrybark oak–sweetgum bottomland stand near the Tombigbee River in Alabama (Province 232: Coniferous-Broadleaved Semievergreen Forests Province). (USDA Forest Service, Southern Research Station photograph.)

Although bottomland forests are relatively flat, elevational differences of only a few feet alter soil formation processes, soil moisture regimes and species composition. Thus, changes in species composition are often associated with relatively minor differences in physiography (Fig. 1.17). Moreover, floodplain physiography can quickly and frequently change as a result of scouring and deposition of sediments. These factors, coupled with the high tree species diversity of bottomland forests, complicate classifying forest types and developing silvicultural prescriptions appropriate to each. Up to 70 tree species occur in southern bottomland forests (Putnam et al., 1960), and species mixtures often change over short distances within stands. Consequently, species associations are difficult to classify meaningfully into more than a few broad types. Although Eyre (1980) listed 14 bottomland cover types (six named for oaks),

Within the United States, the Forest–Prairie Transition Region extends from southern Texas northward to Minnesota and North Dakota (Fig. 1.6). The region coincides with two ecoregion provinces: Forest–Steppes and Prairies (251) and Prairies and Savannas (252) within the Prairie Division (250) (Fig. 1.2, Table 1.2). The region includes Braun’s (1972) Grassland or Prairie Region, Forest–Prairie Transition and Prairie Peninsula Sections, which fall within her Oak–Hickory Forest Region. As its name implies, the Forest–Prairie Transition Region is transitional between the eastern forests and the prairies and dry woodlands of the Dry Domain to the west. On its eastern border, the region adjoins the Northern Hardwood Region, the Central Hardwood Region and the Southern Pine–Hardwood Region. The Forest–Prairie Transition Region spans 1400 miles in latitude and varies in width from as little as 100 miles along the Canadian border to 600 miles between eastern Nebraska and western Indiana. The region includes approximately 191 million acres, about 7% of which are forested. Between Canada and Oklahoma, forests cover 5% of the landscape with most of the remainder devoted to tilled cropland or pasture. The forest cover increases to 13% in parts of Oklahoma and Texas. Most of the forestland in the Forest–Prairie Transition Region is privately owned. Although there are three national grasslands within the region, only 15,000 acres of national forest land (in central Missouri) are included.

River base level

R ID Sweetgum, hickory G Red oak, swamp chestnut oak E Winged elm, black tupelo gum FL SL A Sweetgum, water oak,willow oak O T U Green ash, nuttall oak G H Overcup oak Water hickory

SW AM P Cypress, water tupelo

River birch LE VE Beech, yellow poplar E Sycamore, sweetgum, spruce pine Oaks FL AT SL Sweetgum, oaks, hickories O Blackgum, winged elm U G H Cypress, swamp tupelo FL AT Overcup oak, willow oak Nuttall oak TE R R AC White oak, red oak E Hickories, sweetgum Yellow poplar Loblolly pine

Minor Bottom

37

BA R

River base level

BA R Willow FR O Cottonwood N T Elm, sycamore

Major Bottom

Pecan, sugarberry F SL LA T O Nutall oak, green ash U G H Sugarberry, elm, red maple R ID Willow G Overcup oak, water hickory E Sweetgum, wateroak Willow oak, green ash FL AT Overcup oak, water hickory


Fig. 1.17. The topographic distribution of southern bottomland oaks and associated species in major and minor stream valleys of the Southern Pine–Hardwood Region (Province 232: ConiferousBroadleaved Semi-Evergreen Forests Province). (Reprinted from Hodges and Switzer, 1979, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

Climate, physiography and soil Precipitation within this vast region varies from less than 20 inches per year in the north to 55 inches along the gulf coast of Texas. One-half to two-thirds of the precipitation typically falls during the growing season and snowfall is common north of Texas. From north to south, mean annual temperature ranges from 36° to 70°F (2 to 21°C) with corresponding growing seasons ranging from 111 to 320 days (McNab and Avers, 1994) (Fig. 1.7). Throughout the region precipitation is largely offset by evapotranspiration, creating soil moisture conditions in many localities that are marginal for tree growth.

Most of the region comprises gently rolling plains, although high rounded hills occur and steep bluffs border some river valleys. Elevations range from sea level to 2000 ft. Local relief is less than 165 ft throughout most of the region, but it reaches 500 ft in the Flint Hills of Kansas (McNab and Avers, 1994). Soils are predominantly Mollisols although Vertisols occur on the prairies, and Alfisols occur on savannas and within the Mississippi Valley (Bailey, 1995).

Forest history Native Americans who were largely nomadic inhabited the region for at least 10,000 years. Crops were cultivated as

38

Chapter 1

early as 1000 years ago. A few large Native American communities developed in the major river valleys. One of these was Cahokia (near present-day St Louis), which flourished between AD 1000 and 1400 with an estimated population of 25,000. The forests in the region were an important resource for both nomadic people and larger permanent communities. The demise of Cahokia may have been caused by the exhaustion of the surrounding forests that were used for fuel and for the construction and maintenance of a 2mile-long perimeter wall around the city (Lord, 1999). Frequent fires were essential to the maintenance of prairie and savanna vegetation in many parts of the Forest–Prairie Transition Region, and Native Americans burned the grasslands and woodlands where they lived. Grazing and trampling by herds of bison and other ungulates also were important in maintaining prairies and preventing the encroachment of forests and other woody vegetation. By mid-19th century, European settlers began farming the prairies and draining prairie wetlands. The latter produced some of the nation’s most productive agricultural lands. Trees were largely confined to riparian corridors, steep slopes and scattered savannas. With the exclusion of fire and the elimination of free-ranging ungulates, forests frequently encroach upon abandoned fields and pastures. In 1884, Sargent (1884) stated that, ‘Dakota, with the exception of its riverlands and the small territory between the north and south forks of the Cheyenne River, is practically destitute of timber. The bottoms of the principal streams contain extensive groves of hardwood.’ In Iowa he observed that ‘since the first settlement of the state the forest area has increased by the natural spread of trees over ground protected by fire, and by considerable plantations of cottonwood, maples, and other trees of rapid growth made by farmers to supply fuel and shelter’. Further south, in Texas, Charles Mohr noted, ‘The timber growth immediately west of the Brazos is stunted and scanty; large areas of grass land intervene between

the scrubby woods until all at once ligneous growth disappears and the seemingly boundless prairie, in gently undulating swells expands before the view on all sides’ (Sargent, 1884). Since that time, farms have been established on virtually all the lands suitable for row crops or forage production (McNab and Avers, 1994). Depending on the farm economy, the forested acreage has decreased or increased as forests and woodlands were cleared to create more farm land, or as marginal farm land reverted to forest through tree planting or abandonment.

Oaks as components of the region’s forests The best forest development in this region occurs on its eastern border where it abuts the Northern Hardwood Region, the Central Hardwood Region and the Southern Pine–Hardwood Region. Of the 7% of the Forest–Prairie Transition Region that is forested, three-quarters is classified as oak–hickory or oak–pine. Few of the savannas that formerly occupied the transition zone between forest and prairie exist today. The prairie fires that historically restricted the extent of the region’s forests have been replaced by agricultural practices that now limit most forests to riparian areas or to slopes unsuitable for forage or other crops (Fig. 1.18). Before the mid-19th century, fire was the primary regulator of the distribution of tree species in the region. Narrow bands of forest along streams and ravines, sometimes called gallery forests, provided refuges for trees from the frequent fires that burned across the prairie. Oaks dominated many of these forests. With the advent of farming in this region, the frequency of wildfires was greatly reduced. This allowed the gallery forests to expand into untilled areas that were formerly covered by native grasses (Abrams and Gibson, 1991). However, the invading woody species were generally species such as American elm, hackberry and eastern redcedar rather than oaks. The


39

A

B

Fig. 1.18. (A) Aerial view of the distribution of forests in the Forest–Prairie Transition Region (Province 251: Forest-Steppes and Prairies Province) of northwestern Missouri. Throughout much of this ecoregion, forests are largely restricted to narrow belts occupying steep slopes along rivers and drainages interspersed with agricultural lands. (B) Forested bluffs dominated by oaks (background) along the Missouri River in central Missouri fronted by cultivated bottomland fields. Before settlement by Europeans, these bottomlands were covered by lowland forests dominated by American elm, silver maple, green ash, eastern cottonwood, bur oak and pin oak. (USDA Forest Service, North Central Research Station photographs.)

reduction in wildfires also allowed those species to increase in abundance within existing forests that were formerly dominated by oaks, especially on the more mesic sites. In much of this region, frequent fires are required to prevent the displacement of the oaks by other species (Penfound, 1968; Abrams, 1988; Abrams and Gibson, 1991).

Oak–hickory forests extend from the Central Hardwood Region westward across eastern Oklahoma and into northern Texas (Fig. 1.5A). From east to west the forests become increasingly scrubby and open. An exception is the relatively dense oak forest of the Cross Timbers Region. In Texas, the Cross Timbers comprise two bands of scrubby oak woodland extending 175 miles

40

Chapter 1

southward from the Oklahoma border. These bands are 20–50 miles wide and separated by the Fort Worth Prairie. Forest cover occurs along outcrops of sandy soils of greater porosity than adjacent prairie soils (Braun, 1972). The Cross Timbers were prominent landmarks for westward travellers who otherwise traversed relatively open landscapes (Dyksterhuis, 1948). Although the heavier forest cover in the Cross Timbers area of Texas is somewhat evident from Fig. 1.6, the two distinct strips of woodland are not distinguishable at the resolution shown. Post oak and blackjack oak are the dominant tree species and account for 60% and 20% of the trees, respectively. Except in floodplains, these oaks seldom exceed 12 inches in diameter and 30–45 ft in height. At one time, the herbaceous vegetation in the Cross Timbers was probably similar to that of the surrounding prairie, but grazing during the last century has greatly altered the species composition of the herbaceous layer (Dyksterhuis, 1948). The Cross Timbers vegetation extends northward through Oklahoma and eventually disappears in southern Kansas. Except for the Cross Timbers Region, the upland woodlands of eastern Oklahoma were formerly post–blackjack oak savannas maintained by frequent fires. Grazing and a reduction in burning have since reduced grass cover and facilitated the establishment of dense tree reproduction in many areas; post and blackjack oaks dominate most stands. Although the average basal area of these forests historically has been relatively low, in the absence of burning it has increased from 49 ft2 acre1 in 1957 (Rice and Penfound, 1959) to 80 ft2 acre1 in 1993 (Rosson, 1994). From Kansas northward there are few forests, but where they do occur, oaks often dominate (Figs 1.2 and 1.6). Many of the oak forest types and conditions occurring in the Central Hardwood Region extend westward through the central portion of the Forest–Prairie Region. The central part of the region is capable of supporting forest vegetation and is successional to forest in

areas protected from cultivation. However, because agriculture is the dominant land use, forests are usually restricted to riparian corridors, wet areas, steep slopes and highly erodible lands, or other sites unsuited to agriculture. Nevertheless, oaks and other hardwoods often develop into commercially valuable stands in those parts of the region lying within Illinois, Iowa, northern Missouri and eastern Kansas. Bur oak is the dominant oak species in the northern reaches of the Forest–Prairie Transition Region. It is the only major oak species with a natural range that extends across western Minnesota and into the Dakotas. Bur oak is well adapted to this region because its deep taproot makes it resistant to drought and able to invade prairie grasslands (Johnson, 1990). Its thick bark makes it highly resistant to fires that eliminate most other woody species. Bur oak also thrives on moist alluvial bottoms that support dense hardwood forests in the northern portion of the Forest–Prairie Transition Region. Here, the bur oak type covers approximately 2% of the land area and is the principal forest type. Cottonwood, quaking aspen and American elm are other abundant hardwoods in the northern part of the Forest–Prairie Transition Region.

Western Oak Forests The Southwestern Desert–Steppe Region Geographic extent The Southwestern Desert–Steppe Region includes the scattered oak forests of Arizona, New Mexico, southern Utah, west Texas and southwest Oklahoma (Figs 1.2, 1.6; Table 1.3). Although the range of Gambel oak extends northward as far as southern Wyoming, the oaks there are a small component of the vegetation. Forests and woodlands cover about 20% of the area, but only 7% of this is considered productive forest. Soil moisture deficiencies limit the distribution of oaks and other plant life throughout the region. Oaks


occur as scattered trees and in open woodlands. Their distribution within the region is often limited to discontinuous elevational zones that provide the required regime of precipitation and temperature. The region lies entirely within the Dry Domain and comprises parts of three Divisions: Tropical/Subtropical Steppe (310), Tropical/Subtropical Steppe Mountains (M310) and Tropical/Subtropical Desert (320). Included are six ecoregion Provinces: Coniferous Open Woodland and Semideserts (311), Steppes and Shrubs (313), Shortgrass Steppes (314), Steppe or Semidesert–Mixed Forest–Alpine meadow or Steppe (M311), Semideserts (321), and Deserts on Sand (323) (Fig. 1.2; Table 1.3). The Southwestern Desert–Steppe Region extends 1200 miles from northwestern Arizona to the Gulf of Mexico in southern Texas. It varies from 300 to 700 miles in width, and encompasses roughly 250 million acres including the Mojave Desert, the Sonoran Desert, the Painted Desert, the Colorado Plateau, the southern Rocky Mountains, Texas High Plains and the Edwards Plateau. Within the region, oak forests are widely scattered and cover only a small fraction of the landscape (Fig. 1.2). The federal government or Native Americans own two-thirds of the forests and woodlands in Arizona and New Mexico, but in Texas and Oklahoma most are privately owned (Powell et al., 1994).

Climate, physiography and soil A defining characteristic of this region is a rate of surface evaporation that exceeds precipitation. The climate varies from dry to desert. Annual precipitation ranges from less than 10 inches to 30 inches (Bailey, 1995). Even in areas with greater precipitation, high rates of evaporation limit moisture availability. Average annual temperature ranges from 40 to 70°F (4 to 21°C). Although temperatures decrease with increasing elevation, mean monthly temperatures generally exceed 32°F (0°C) (Fig. 1.19). Elevation ranges from sea level along the southern Texas Gulf Coast to 7000 ft in the Colorado Plateau; some

41

mountain peaks are substantially higher. Soils are variable throughout the region and include Mollisols, Aridisols and dry Entisols (Bailey, 1995).

Forest history As in other regions of the United States, Native Americans customarily burned the forests and woodlands where they lived. Lightning was also a common cause of combustion. These fires maintained an open understorey in the extensive ponderosa pine forests of higher elevations. In 1880 alone, about 75,000 acres burned – which accounted for 0.1 to 1% of the woodland within the settled area (Sargent, 1884). Beginning in the mid-19th century, European settlers were drawn to the region by opportunities for mining and livestock production. Lands were not suitable for agriculture, and the great land clearing that decimated the eastern oak forests did not occur here. However, logging, grazing and changes in fire regimes changed the species composition of forests and woodlands. In recent decades, the suppression of fires has increased the amount of tree reproduction, especially conifers, and decreased grasses and forbs growing beneath forest canopies (Long, 1995). Throughout the region, oaks have historically had little commercial value. In 1884, Sargent (1884) described the forests in and around New Mexico: ‘The deciduous trees of this entire southwestern region, often of considerable size, are generally hollow, especially the oaks; they are of little value for any mechanical purpose, although affording abundant and excellent fuel.’ Then, as now, ponderosa pine was the principal timber species.

Oaks as components of the region’s forests Forests and woodlands cover only a small portion of the total area in the region, and the oaks comprise only a small percentage of that. In Arizona and New Mexico, only 15% of the land base is forested. Only 3% of the area of those two states can produce more than 20 ft3 of timber per acre per year,

42

Chapter 1

Astoria, OR

51˚F 76 in.

Pasadena, CA 62˚F 19 in.

Tahoe, CA 42˚F 31 in.

Abilene, TX

65˚F 25 in.

Brawley, CA 72˚F 2 in.

Colorado Springs, CO

48˚F 14 in.

Fig 1.19. Representative climates for selected ecoregion Divisions in the western United States. Mean monthly precipitation is shown by the solid lines (right axis) and temperature by dashed lines (left axis). Mean annual values are given above each graph. Periods of drought are indicated where the precipitation line falls below the temperature line (e.g. as in Division 260). Division boundaries are shown in Figs 1.2 and 1.5. (Ecoregion and climatic data from Bailey, 1995.)


and virtually none of that is oak forest. Commercial forests include ponderosa pine (75%), Douglas-fir (13%), spruce–fir (9%) and aspen (3%). The only recognized oak cover type here is western live oak (Appendix 3) (Eyre, 1980). It occurs at elevations from 4000 to 6000 ft in the foothills and lower mountain slopes of Arizona and New Mexico. At higher elevations, the western live oak cover type gives way to ponderosa pine and pinyon–juniper, with oak–conifer mixtures occurring in the transition. At lower elevations the western live oak type yields to an open growth of shrubby evergreen oaks. Mesquite and desert vegetation typically occurs below that. Characteristic species of the western live oak type include Emory, Arizona white, Mexican blue and silverleaf oaks (Eyre, 1980) (Fig. 1.20). Ajo oak, Dunn oak, grey oak and Havard oak also occur in Arizona and New Mexico. At the eastern end of the Southwestern Desert–Steppe Region towards the High Plains and Edwards Plateau of west-central Texas, precipitation increases and oaks become more prominent. The Mohr (shin) oak forest type covers more than 8 million acres in Texas where it develops best under

43

20–25 inches of precipitation annually (Eyre, 1980). However, that amount of precipitation represents the upper end of the range for the region (e.g. see Fig. 1.19, Division 310). Other oaks that occur in west-central Texas include Arizona white, blackjack, bur (marginally), chinkapin, Durand, Emory, Havard, Lacey, live, sandpaper, Texas and Texas live oaks.

The Pacific Mediterranean–Marine Region Geographic extent The Pacific Mediterranean–Marine Region includes the oak forests and woodlands of California, Oregon and Washington (Fig. 1.6). The region lies within the western portion of the Humid Temperate Domain and includes the Mixed Forest–Coniferous Forest–Alpine Meadow Province (M261) and the Mediterranean Woodland or Shrub–Mixed Coniferous Forest–Steppe or Meadow Province (M262) within the Mediterranean Mountains Division (M260) (Fig. 1.2). Oaks also occur within the Coast Ranges of California, which includes the Mediterranean Hardleaved Evergreen Forest,

Fig. 1.20. Emory oak woodland in the Peloncillo Mountains of southwestern New Mexico, Coronado National Forest, New Mexico (Province M321: Semideserts Province). (USDA Forest Service, Rocky Mountain Research Station photograph.)

44

Chapter 1

Open Woodlands and Shrubs Province (262) and the Redwood Province (263). At its northern extent, the Pacific Mediterranean– Marine Region also reaches the Mixed Forest Province (241) of the Marine Division (240) in Oregon and Washington. The Pacific– Mediterranean–Marine Region also includes California’s Central Valley (province 261) and the mountainous zones of Washington and northern Oregon (M261) where oaks are not abundant. The Pacific Mediterranean–Marine Region extends nearly 900 miles from Washington to southern California but less than 200 miles from the Pacific Ocean to the eastern slopes of the Sierra Nevada Mountains. Although the region covers about 75 million acres, the oaks are limited to relatively narrow elevational zones. In California, Oregon and Washington, slightly less than half the timberland is publicly owned (Powell et al., 1994). In contrast with the eastern United States, most of the privately owned timberland in this region is held by corporations rather than by nonindustrial private owners (Birch, 1996). However, the ownership of oak forests and woodlands does not follow this trend; about three-quarters of that acreage is in non-industrial private ownership (Thomas, 1997).

Climate, physiography and soil Climate is strongly influenced by the Pacific Ocean and by the Coast and Sierra Nevada Ranges, which dominate the physiography of the region. Elevations range from sea level to more than 14,000 ft. In the mountain ranges, increasing elevation is associated with decreasing temperatures and variation in precipitation. For a given elevation, precipitation is generally greater on western slopes than on eastern slopes. Latitude also influences climate so that a given climatic zone occurs, from north to south, at progressively higher elevations. However, mountainous topography creates climatic irregularities and discontinuities, and the distribution of oaks and associated tree species varies accordingly. Most of the precipitation occurs during the autumn, winter and spring. Annual pre-

cipitation generally ranges from 10 to more than 60 inches in the ecological provinces where oaks occur. Temperature extremes and moisture stress are reduced near the coast where fog supplements precipitation and the ocean reduces fluctuations in temperature. Elsewhere the region’s Mediterranean climate is characterized by 2 to 4 months of drought during the summer (Table 1.3, Fig. 1.19). Low precipitation generally occurs at lower elevations and on the east faces of mountain ranges. Soils include Ultisols, Alfisols, Mollisols, Entisols and Inceptisols (Bailey, 1995).

Forest history The historical importance of oaks is recorded in ancient bedrock mortars that were used by Native Americans to grind acorns into flour. Acorns were a staple food of Native Americans in this region, and Biswell (1989) suggests that oaks were so important to their diet that they burned oak woodlands to both encourage oak reproduction and to facilitate acorn gathering. Although human-caused fires have been historically associated with the oaks of the region for thousands of years, there is uncertainty about what proportion of the landscape was regularly affected by humans. During the post-settlement period of 1850–1950, the mean interval between fires in the oak–pine forests of the foothills of central California was 8 years (Stephens, 1997). Commercial logging in the region has largely focused on the conifers. In 1884, Sargent (1884) stated: The forests of California, unlike those of the Atlantic States, contain no great store of hardwoods. The oaks of the Pacific forests, of little value for general mechanical purposes, are unfit for cooperage stock. No hickory, gum, elm, or ash of large size is found in these forests, California produces no tree from which a good wine cask or wagon wheel can be made. The cooperage business of the state, rapidly increasing with the development of grape culture, is entirely dependent upon the forests of the Atlantic region for its supply of oak.


Sargent further noted that large quantities of chestnut oak (sic tanoak), once common in the northern Coast Range of California, are ‘now becoming scarce and in danger of speedy extermination’ due to utilization by the tanning industry. Sargent’s reference to the oaks of Oregon and Washington is slightly less disparaging. In the Willamette Valley, he noted that Oregon white oak woodlands were becoming re-established after reductions in fire frequency. Along the Yakima River in Washington, he noted that Oregon white oaks were limited to 15 ft in height and 6 inches in diameter. The logging industry on the Pacific Coast was established in the 18th century under Hispanic influence. Through the middle of the 19th century the relatively small industry served markets in South America, Australia and the Pacific Rim (Williams, 1989). The gold rush of 1849 and the completion of the transcontinental railroad opened additional markets, but the increase in lumber production in this region occurred gradually, beginning about 1900 when the large timber companies and railroads moved west after exhausting the ready supply of timber in the Lake States. Increases in timber production in the region continued into the Great Depression, but output eventually dropped by 75%. By 1950, however, annual timber production in the West exceeded 16 billion board feet annually, which was greater than that produced in other regions of the United States. Today lumber production in this region lags significantly behind that of the south. Harvest of hardwood growing stock has remained nearly constant since 1976, but the volume of hardwoods harvested annually is only about 5% of the region’s total. Historically, oak forests were little affected by commercial logging, but locally they were widely utilized for firewood and fence posts. Ranchers and farmers had the greatest influence on the oak woodlands of the foothills and lower slopes as a consequence of clearing them for agriculture and grazing. Sargent (1884) noted:

45

The permanence of the mountain forests of California is severely endangered, moreover, by the immense herd of sheep, cattle, and horses driven to the mountains every year, at the commencement of the dry season, to graze. From the foothills to the highest alpine meadows, every blade of herbage and every seedling shrub and tree is devoured.

In California, oak woodlands were reduced from an estimated 10–12 million acres to about 7 million acres today (Thomas, 1997). The oak woodlands are predominantly owned by farmers and ranchers, and between 1945 and 1970 the primary loss of woodland acreage resulted from conversion to rangeland. Invasion of non-native grasses and the suppression of fire have created problems in maintaining oak woodlands and savannas (see Chapter 9 for details of savanna restoration and management). More recently, the greatest losses of oak woodland have resulted from suburban residential development (Bolsinger, 1988). This has given rise to concern for property damage from the wildfires historically associated with the oak woodlands.

Oaks as components of the region’s forests Most of the region’s oak forests and woodlands occur in California where they account for approximately one-quarter of the wooded acres. Oaks surround California’s Central Valley in the foothills of the Sierra Nevada, Cascade and Klamath Ranges (Figs 1.5 and 1.6). Although oaks were formerly abundant within parts of California’s Central Valley (province 261), their distribution has been greatly reduced there (Griffin, 1977). Oaks also occur on the western slopes of the Coast Ranges in central and southern California. The range of Oregon white oak extends northward into central Oregon and Washington in the Willamette Valley and the Puget lowlands between the Cascade and Coast Ranges. Included are 18 species of oak trees and shrubs plus additional hybrids (Bolsinger, 1988; Thomas, 1997). Eight oak species that reach tree size are abundant: California black, blue, interior live, coast live, canyon live, valley, Oregon white and Engelmann oaks (Plumb and McDonald, 1981).

46

Chapter 1

Western oak forests are often categorized as either timberland (forests suitable for commercial wood production and capable of producing at least 20 ft3 acre1 year1 of merchantable volume), or woodlands (sites of lower productivity primarily utilized for forage and firewood). In California, only about 1 in 4 acres of hardwood forest qualifies as timberland. Oak woodlands are sparsely covered with trees compared to oak timberlands. The statewide volume of oaks in woodlands and in timberlands is nevertheless nearly equal because the acreage of woodlands is approximately three times that of timberlands (Table 1.4). Three-quarters of the oak woodlands are grazed and these account for about onethird of California’s total forage (Thomas, 1997). Only about 500,000 board feet of hardwood lumber was produced in California in 1992 (Ward, 1995). The combined effects of temperature and precipitation (which latitude, elevation, slope and aspect affect) regulate the distribution of oaks. In the Pacific Mediterranean–Marine Region, many oak forests and woodlands are restricted to elevational zones in the transition between grassland and chaparral at lower elevations and coniferous forest at higher elevations. Mean temperatures within the region increase with decreasing latitude, and the

oaks occur at higher elevations at lower latitudes. Due to the interaction of climate and mountainous topography, the distribution of oaks in this region is more geographically restricted than in the eastern United States. Several classification schemes have been proposed for the complex vegetation relationships that occur in the Pacific Mediterranean–Marine Region (e.g. Griffin, 1977; Paysen et al., 1980, 1982; Barbour, 1988; Allen, 1990) (Fig. 1.21). Eyre (1980) recognized five oak cover types and two additional types where oaks commonly occur in mixtures with other species (Appendix 3). The Oregon white oak type is found in the northern portion of the Pacific Mediterranean–Marine Region from northern California to Vancouver Island. This type occurs at lower elevations (0–3900 ft) and primarily in inland valleys or lower slopes between the Coast Ranges and the Cascade or Sierra Nevada Ranges (Eyre, 1980). The type makes its best development in the vicinity of the Willamette Valley where closed-canopy Oregon white oak stands developed from former oak savannas when periodic ground fires were excluded (Thilenius, 1968). The species also occurs in mixtures with other hardwoods and conifers including California black oak, canyon live oak, ponderosa pine and Douglas-fir (Appendix 3).

Table 1.4. Standing oak volumes in California timberlands and woodlands. Although oaks make up 63% of the total volume of California’s hardwoods, oak timberlands (commercial forest lands) comprise only 8% of the 50 billion cubic feet total volume (softwoods plus hardwoods on California timberlands).a

Species California black oak Canyon live oak Blue oak Coast live oak Oregon white oak Interior live oak California white (valley) oak Engelmann oak Total oak Total hardwoods (all species) a

Volume in timberlands (million ft3)

Volume in woodlands (million ft3)

Total (million ft3)

Species total as a proportion of all oaks (%)

2,254 1,302 1 126 211 45 34 0 3,973 7,661

277 731 1,112 755 389 508 164 10 3,946 4,855

2,531 2,033 1,113 881 600 553 198 10 7,919 12,516

32 26 14 11 8 7 3 0 100 –

Adapted from Shelly (1997) and Bolsinger (1980, 1988).


47

14,500 12,000 Alpine

Meadow

Subalpine forest Lodgepole pine forest

6000

4000

2000

Jeffrey pine forest

Red fir forest

8000

White fir forest

Upland live oak woodland

d an wl k Lo oa d e n liv dla o wo

d xe ral Mi par a ch

Juniper woodland

Elevation (ft)

10,000

Ponderosa pine forest Mixed evergreen forest and black oak woodland

Chamise chaparral

Blue oak woodland

Mesic

Xeric Topographic moisture gradient

Fig. 1.21. Relation of oak forests to elevation, moisture gradients and other forest types found in the Pacific–Mediterranean–Marine Region of California (Ecoregion Provinces 261, 262, 263 and M261). Oak forests and woodlands are usually found above the chaparral zone and below the ponderosa pine zone. (Redrawn from Barbour (1988) and Vankat (1982).) Reprinted with permission of Cambridge University Press.)

California black oak attains a greater volume (Table 1.4) and is distributed across a greater area than the other California oaks (Plumb and McDonald, 1981). The California black oak type occurs from central Oregon to the Mexican border across elevations ranging from 200 to 8000 feet with corresponding annual precipitation of 25–85 inches annually. Best development of the forest type occurs in the northern half of California in the Klamath and Cascade Mountains and the Coast and Sierra Nevada Ranges. There the forest type is found at elevations between 1500 and 3000 ft with corresponding annual precipitation between 30 and 50 inches (Eyre, 1980). After disturbance, this species maintains itself through sprouting to form even-aged stands. On suboptimal sites it is successional to other forest types. Associated species include other oaks, ponderosa pine, Douglas-fir and Pacific madrone (Appendix 3).

Canyon live oak occurs from the Willamette Valley to the Baja Peninsula and east into Arizona at elevations from near sea level in the north to 9000 feet in the south (Eyre, 1980). It comprises about one-quarter of California’s oak volume and is second only to California black oak in this regard (Table 1.4). Canyon live oak forms pure stands on very steep slopes and dry canyon bottoms. Elsewhere it occurs in mixture with Douglas-fir, ponderosa pine and other conifers. The species is shade tolerant when young and often maintains itself in relatively stable communities (Eyre, 1980). The blue oak–digger pine forest type surrounds California’s Central valley at elevations between 500 and 5000 ft, although blue oak occasionally extends to the valley floor (Fig. 1.22). This forest type occurs between the valley grasslands and the montane forests above, where it can

48

Chapter 1

Fig. 1.22. Blue oak woodland in the Sierra Nevada Range (Province M261: Dry Steppe Province). (USDA Forest Service, North Central Research Station photograph.)

endure a meagre 10 inches of annual precipitation (Eyre, 1980). Forest cover ranges from 30 to 80% with canopy heights between 15 and 50 ft. Associated species include California live oak, interior live oak, valley oak and California black oak (Barbour, 1988). At low elevations blue oak and valley oak mixtures develop savanna communities. Valley oak savannas extend into the Central Valley where they make their best development on alluvial soils (Griffin, 1977). The California coast live oak forest type (sometimes referred to as southern oak woodland) occurs on the west side of the Coast Range in the southern two-thirds of California. It extends inland on north-facing slopes of narrow valleys and other cool sites. This type occurs at elevations of up to 3000 ft in the northern part of its range and to 5000 ft in the southern portion. Although it can form pure, closed canopy stands, it is

considered a woodland type and commonly occurs in savannas comprised of scattered oaks or in mixture with conifers (Appendix 3). California coast live oak is long-lived, moderately shade tolerant, and forms relatively permanent woodlands. When trees reach about 8 inches dbh they are also highly resistant to fire (Eyre, 1980). The ecological importance of California’s oak woodlands and timberlands is receiving increased attention (Pillsbury et al., 1997). Although their value for commercial products is low, their importance to wildlife, water quality, aesthetics, soil protection, recreation and fuelwood is widely acknowledged (Helms and Tappeiner, 1996). A principal silvicultural problem related to the oak woodlands of the Pacific Mediterranean–Marine Region is ensuring that the regeneration of oaks is sufficient for replacing trees periodically lost to natural mortality and timber harvesting.

References Abrams, M.D. (1988) Effects of prescribed fire on woody vegetation in a gallery forest understory in northeastern Kansas. Transactions of the Kansas Academy of Science 91, 63–70. Abrams, M.D. (1990) Adaptations and responses to drought in Quercus species of North America. Tree Physiology 7, 227–238. Abrams, M.D. (1992) Fire and the development of oak forests. BioScience 42, 346–353.


49

Abrams, M.D. and Gibson, D.J. (1991) Effects of fire exclusion on tallgrass prairie and gallery forest communities in eastern Kansas. USDA Forest Service General Technical Report SE SE-69, pp. 3–10. Abrams, M.D. and Nowacki, G.J. (1992) Historical variation in fire, oak recruitment, and post-logging accelerated succession in central Pennsylvania. Bulletin of the Torrey Botanical Club 119, 19–28. Aizen, M.A. and Patterson, W.A. (1990) Acorn size and geographical range in the North American oaks (Quercus L.). Journal of Biogeography 17, 327–332. Allaby, M. (ed.) (1994) The Concise Oxford Dictionary of Ecology. Oxford University Press, New York. Allen, B.H. (1990) Classification of oak woodlands. Fremontia 18(3), 22–25. Arend, J.L. and Scholz, H.F. (1969) Oak forests of the Lake States and their management. USDA Forest Service Research Paper NC NC-31. Axelrod, D.I. (1983) Biogeography of oaks in the Arcto-Tertiary Province. Annals of the Missouri Botanical Gardens 70, 629–657. Bailey, R.G. (1995) Descriptions of the ecoregions of the United States. USDA Forest Service Miscellaneous Publication 1391. Bailey, R.G. (1997) Ecoregions of North America, 1:15,000,000 scale map (rev.). USDA Forest Service, Washington, DC. Bailey, R.G. (1998) Ecoregions. Springer-Verlag, New York. Barbour, M.G. (1988) Californian upland forests and woodlands. In: Barbour, M.G. and Billings, W.D. (eds) North American Terrestrial Vegetation. Cambridge University Press, New York, pp. 131–164. Barnes, B.V., Pregitzer, K.S., Spies, T.A. and Spooner, V.H. (1982) Ecological forest site classification. Journal of Forestry 80, 493–498. Birch, T.W. (1996) Private forest-land owners of the United States, 1994. USDA Forest Service Resource Bulletin NE NE-134. Biswell, H.H. (1989) Prescribed Burning in California Wildlands Vegetation Management. University of California Press, Berkeley. Bliss, J.C., Nepal, S.K., Brooks, R.T., Jr and Larsen, M.D. (1994) Forestry community or granfalloon. Journal of Forestry 92(9), 6–10. Bliss, J.C., Nepal, S.K., Brooks, R.T., Jr and Larsen, M.D. (1997) In the mainstream: environmental attitudes of Mid-South forest owners. Southern Journal of Applied Forestry 21, 37–43. Bolsinger, C.L. (1980) California forests: trends, problems, and opportunities. USDA Forest Service Resource Bulletin PNW PNW-89. Bolsinger, C.L. (1988) The hardwoods of California’s timberlands, woodlands, and savannas. USDA Forest Service Resource Bulletin PSW PSW-148. Bourdeau, P. (1954) Oak seedling ecology determining segregation of species in Piedmont oak–hickory forests. Ecological Monographs 24, 297–320. Braun, E.L. (1972). Deciduous Forests of Eastern North America. Hafner, New York. Burns, R.M. and Barber, J.C. (1989) Silviculture of southern pines. USDA Forest Service General Technical Report WO WO-55, pp. 31–38. Burns, R.M. and Honkala, B.H. (1990) Silvics of North America. USDA Forest Service Agriculture Handbook 654, Vol. 2. Carvell, K.L. and Tryon, E.H. (1961) The effect of environmental factors on the abundance of oak regeneration beneath mature oak stands. Forest Science 7, 98–105. Chase, A. (1995) In a Dark Wood: the Fight Over Forests and the Rising Tyranny of Ecology. Houghton Miflin, Boston, Massachusetts. Cleland, D.T., Hart, J.B., Host, G.E., Pregitzer K.S. and Ramm, C.W. (1993) Ecological Classification and Inventory System of the Huron-Manistee National Forests. USDA Forest Service, Washington, DC. Cooper, A.W. (1989). Ecology of the pine–hardwood type. USDA Forest Service General Technical Report SE SE-58, pp. 3–8. Cunningham, R.J. and Hauser, C. (1989) The decline of the Missouri Ozark forest between 1880 and 1920. USDA Forest Service General Technical Report SE SE-58, pp. 34–37. Curtis, J.T. (1959) The Vegetation of Wisconsin. University of Wisconsin Press, Madison. DeVivo, M.S. (1991) Indian use of fire and land clearance in the southern Appalachians. USDA Forest Service General Technical Report SE SE-69, pp. 306–310.

50

Chapter 1

Dyksterhuis, E.J. (1948) The vegetation of the western Cross Timbers. Ecological Monographs 18, 325–376. English, B.C., Bell, C.D., Wells, G.R. and Roberts, R.K. (1997) Stewardship incentives in forestry: participation factors in Tennessee. Journal of Forestry 21(1), 5–10. Eyre, F.H. (ed.) (1980) Forest Cover Types of the United States and Canada. Society of American Foresters, Washington, DC. Fernow, B.E. (1911) A Brief History of Forestry. University Press, Toronto. Flora of North America Editorial Committee. (1997) Flora of North America North of Mexico, Vol. 3. Oxford University Press, New York. Godman, R.M. (1985) What are northern hardwoods. In: Hutchinson, J.G. (ed.) Northern Hardwood Notes. USDA Forest Service, North Central Research Station, St Paul, MN, pp. 1.1–1.3. Griffin, J.R. (1977) Oak woodland. In: Barbour, M.G. and Major, J. (eds) Terrestrial vegetation of California. John Wiley & Sons, New York, pp. 383–415. Helms, J.A. (ed.) (1998) The Dictionary of Forestry. Society of American Foresters, Bethesda, Maryland. Helms, J.A. and Tappeiner. (1996) Silviculture in the Sierra. In: Status of the Sierra, Vol. II, Assessments and Scientific Basis for Management Options. Sierra Nevada Ecosystem Project final report to Congress. University of California, Davis, pp. 439–476. Hicks, R.R., Jr (1997) A resource at the crossroads: a history of the central hardwoods. USDA Forest Service General Technical Report NC NC-188, pp. 1–22. Hodges, J.D. (1995) The southern bottomland hardwood region and the brown loam bluffs subregion. In: Barrett, J.W. (ed.) Regional Silviculture of the United States. John Wiley & Sons, New York, pp. 227–270. Hodges, J.D. and Switzer, G.L. (1979) Some aspects of the ecology of southern bottomland hardwoods. In: Gateway to Opportunity: Proceedings of the Society of American Foresters Convention. Society of American Foresters, Washington, DC, pp. 360–365. Jensen, R.J. (1997) Quercus Linnaeus sect. LOBATAE Loudon, red or black oaks. In: Flora of North America North of Mexico (Flora of North America Editorial Committee). Oxford University Press, New York, pp. 447–468. Johnson, P.S. (1990) Quercus macrocarpa Michx. Bur oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 686–692. Johnson, P.S. (1993a) Perspectives on the ecology and silviculture of oak-dominated forests in the central and eastern states. USDA Forest Service General Technical Report NC NC-153. Johnson, P.S. (1993b) Sources of oak reproduction. USDA Forest Service General Technical Report SE SE-84, pp. 112–131. Johnson, P.S. (1994a) The silviculture of northern red oak. USDA Forest Service General Technical Report NC NC-173, pp. 33–68. Johnson, P.S. (1994b) La sylviculture du chene rouge aux USA. In: Timbal, A., Kremer, A., LeGoff, N. and Nepveu, G. (eds) Le chene rouge d’Amerique. Institut National de la Recherche Agronomique, Paris, pp. 272–283. Jokela, J.J. and Sawtelle, R.A. (1985) Origin of oak stands on the Springfield Plain: a lesson on oak regeneration. Proceedings of the Central Hardwood Forest Conference V. University of Illinois, Urbana-Champaign, pp. 181–188. Kerpez, T.A. and Stauffer, D.F. (1989) Avian communities of pine–hardwood forests in the Southeast: characteristics, management, and modeling. USDA Forest Service General Technical Report SE SE-58, pp. 156–169. Kroeber, A.L. (1925) Handbook of the Indians of California. Bureau of American Ethnology of the Smithsonian Institution Bulletin 78. Kuhlman, E.G. (1978) The devastation of American chestnut by blight. In: MacDonald, W.L., Cech, F.C., Luchok, J. and Smith, H.C. (eds) Proceedings of the American Chestnut Symposium. West Virginia Books, Morgantown, pp. 1–3. Lawson, E.R. (1990) Juniperus virginiana L. Eastern redcedar. USDA Forest Service Agriculture Handbook 654, Vol. 1, pp. 131–140. Leopold, B.D., Weaver, G.H., Cutler, J.D. and Warren, R.C. (1989) Pine–hardwood forests in northcentral Mississippi: an ecological perspective. USDA Forest Service General Technical Report SE SE-58, pp. 211–222.


51

Liebhold, A.M., MacDonald, W.L., Bergdahl, D. and Mastro, V.C. (1995a) Invasion by exotic forest pests: a threat to forest ecosystems. Forest Science Monograph 30. Little, E.L., Jr (1971) Atlas of United States Trees, Volume 1, Conifers and Important Hardwoods. USDA Forest Service Miscellaneous Publication 1146. Little, E.L., Jr (1977) Atlas of United States Trees, Volume 4, Minor Eastern Hardwoods. USDA Forest Service Miscellaneous Publication 1342. Little, E.L., Jr. (1979) Checklist of United States trees. USDA Agriculture Handbook 541. Loftis, D.L. (1988) Regenerating red oak in the southern Appalachians: predictive models and practical applications. PhD dissertation, North Carolina State University, Raleigh. Long, J.N. (1995) The middle and southern Rocky Mountain region. In: Barrett, J.W. (ed.) Regional Silviculture of the United States. Wiley & Sons, New York, pp. 3335–3386. Lord, L. (1999) The Americas. US News and World Report (16–23 August), 84–87. Lorimer, C.G. (1985) The role of fire in the perpetuation of oak forests. Proceedings of Challenges in Oak Management and Utilization. University of Wisconsin, Madison, pp. 8–25. Lorimer, C.G. (1989) The oak regeneration problem: new evidence on causes and possible solutions. University of Wisconsin Forestry Research Analyses 8. McMinn, H.E. (1964) An Illustrated Manual of California Shrubs. University of California Press, Berkeley. McNab, W.H. and Avers, P.E. (1994) Ecological subregions of the United States: section descriptions. USDA Forest Service Administrative Publication WO-WSA WO-WSA-5. Manos, P.S. (1997) Quercus Linnaeus sect. PROTOBALANUS (Trelease) A. Camus, Intermediate oaks. In: Flora of North America North of Mexico, Vol. 3. Oxford University Press, New York, pp. 468–471. Miller, H.A. and Lamb, S.H. (1985) Oaks of North America. Naturegraph Publishers, Happy Camp, California. Nigh, T.A., Grabner, J.K., Becker, C. and Kabrick, J.M. (2000) Ecological Classification of the Current River Hills Subsection: Draft Manual. Missouri Department of Conservation, Jefferson City. Nixon, K.C. (1997) Quercus Linnaeus, oak. In: Flora of North America North of Mexico, Vol. 3. Oxford University Press, New York, pp. 445–447. Nixon, K.C. and Muller, C.H. (1997) Quercus Linnaeus sect. QUERCUS, White oaks. In: Flora of North America North of Mexico, Vol. 3. Oxford University Press, New York, pp. 471–506. Nowacki, G.J., Abrams, M.D. and Lorimer, C.G. (1990). Composition, structure, and historical development of northern red oak stands along an edaphic gradient in north–central Wisconsin. Forest Science 36, 276–292. Olson, S.D. (1996) The historical occurrence of fire in the Central Hardwoods, with emphasis on southcentral Indiana. Natural Areas Journal 16, 248–256. Orwig, D.A. and Abrams, M.D. (1994) Land-use history (1720–1992), composition, and dynamics of oak–pine forests within the Piedmont and Coastal Plain of northern Virginia. Canadian Journal of Forest Research 24, 1216–1225. Pavlik, B.M., Muick, P.C., Johnson, S. and Popper, M. (1991) Oaks of California. Cachuma Press, Los Olivos, California. Paysen, T.E., Derby, J.A., Balck, H., Jr, Bleich, V.C., and Mincks, J.W. (1980) A vegetation classification system applied to southern California. USDA Forest Services General Technical Report PSW PSW-45. Paysen, T.E., Derby, J.A. and Conrad, C.E. (1982). A vegetation classification system for use in California. USDA Forest Service General Technical Report PSW PSW-63. Penfound, W.T. (1968) Influence of a wildfire in the Wichita Mountains Wildlife Refuge, Oklahoma. Ecology 49, 1003–1006. Petrides, G.A. (1988) A Field Guide to Eastern Trees. Houghton Mifflin, Boston, Massachusetts. Petrides, G.A. and Petrides, O. (1992) A Field Guide to Western Trees. Houghton Mifflin, Boston, Massachusetts. Phillips, D.R. and Abercrombie, J.A., Jr (1987) Pine–hardwood mixtures – a new concept in regeneration. Southern Journal of Applied Forestry 11, 192–197. Pillsbury, N.H., Verner, J. and Tietje, W.D. (1997) Proceedings of symposium on oak woodlands: ecol-

52

Chapter 1

ogy, management, and urban interface issues. USDA Forest Service General Technical Report PSW PSW-160. Pinchot, G. (1987) Breaking New Ground. Island Press, Washington, DC. Plumb, T.R. and McDonald, P.M. (1981) Oak management in California. USDA Forest Services General Technical Report PSW PSW-54. Powell, D.S., Faulkner, J.L., Darr, D.R., Shu, Z. and MacCleery, D.W. (1994) Forest resources of the United States, 1992. USDA Forest Service General Technical Report RM RM-234 (rev.). Putnam, J.A., Furnival, G.M. and McKnight, J.S. (1960) Management and inventory of southern hardwoods. USDA Forest Service Agriculture Handbook 181. Pyne, S.J. (1982) Fire in America. Princeton University Press, Princeton, New Jersey. Rice, E.L. and Penfound, W.T. (1959) The upland forests of Oklahoma. Ecology 40, 593–608. Roach, B.A. (1968) Is clear cutting good or bad? Keep Tennessee Green Journal 8, 4–5, 12–14. Roach, B.A. and Gingrich, S.F. (1968) Even-aged silviculture for upland central hardwoods. USDA Forest Service Agriculture Handbook 355. Rosson, J.F., Jr (1994) Quercus stellata growth and stand characteristics in the Quercus stellata–Quercus marilandica forest type in the Cross Timbers region of central Oklahoma. In: Proceedings North American Conference on Barrens and Savannas. US Environmental Protection Agency, Great Lakes National Program Office, Chicago, Illinois, pp. 329–333. Sargent, C.S. (1884) Report on the Forests of North America (exclusive of Mexico). Government Printing Office, Washington, DC. Seymour, R.S. (1995) The northeastern region. In: Barrett, J.W. (ed.) Regional Silviculture of the United States. Wiley & Sons, New York, pp. 31–80. Sheffield, R.M., Birch, T.W., Leatherberry, E.C. and McWilliams, W.H. (1989) The pine–hardwood resource in the Eastern United States. USDA Forest Service General Technical Report SE SE-58, pp. 9–19. Shelly, J.R. (1997) An examination of the oak woodland as a potential resource for higher-valued wood products. USDA Forest Service General Technical Report PSW PSW-160, pp. 445–455. Skeen, J.N., Doerr, P.D. and Van Lear, D.H. (1993) Oak–hickory–pine forests. In: Martin, W.H., Boyce, S.G. and Echternacht, A.C. (eds) Biodiversity of the Southeastern United States. John Wiley & Sons, New York, pp. 1–34. Smith, D.M. (2000) American chestnut: ill-fated monarch of the eastern hardwood forest. Journal of Forestry 98(2), 12–15. Smith, H.C., Lamson, N.I. and Miller, G.W. (1989) An esthetic alternative to clearcutting? Journal of Forestry 87(3), 14–18. Stephens, S.L. (1997) Fire history of a mixed oak–pine forest in the foothills of the Sierra Nevada, El Dorado County, California. USDA Forest Service General Technical Report PSW PSW-160, pp. 191–198. Thilenius, J.F. (1968) The Quercus garryana forests of the Willamette Valley, Oregon. Ecology 49, 1124–1133. Thirgood, J.V. (1971) The historical significance of oak. Proceedings of Oak Symposium (USDA Forest Service Northeastern Forestry Experimental Station), pp. 1–18. Thomas, J.W. (1997) California’s oak woodlands: where we have been, where we are, where we need to go. USDA Forest Service General Technical Report PSW PSW-160, pp. 3–9. Trimble, G.R., Jr (1973) The regeneration of Central Appalachian hardwoods with emphasis on the effects of site quality and harvesting practice. USDA Forest Service Research Paper NE NE282. Tucker, J.M. (1980) Taxonomy of California oaks. USDA Forest Service General Technical Report PSW PSW-44, pp. 19–29. USDA Forest Service (1993) Forest Type Groups of the United States (map). USDA Forest Service, Washington, DC. USDA Forest Service (2000) Resources Planning Act (RPA) Statistical Tables, 1997. USDA Forest Service, Washington, DC. Vankat, J.L. (1982) A gradient perspective on the vegetation of Sequoia National Park, California. Madroño 29, 200–214. Van Kley, J.E., Parker, G.R., Franzmeier, D.P. and Randolph, J.C. (undated). Field Guide: Ecological Classification of the Hoosier National Forest and Surrounding Areas of Indiana. USDA Forest Service, Hoosier National Forest, Beford, Indiana.


53

Waldrop, T.A. (ed.) (1989) Proceedings Pine–Hardwood Mixtures: a Symposium on Management and Ecology of the Type. USDA Forest Service General Technical Report SE SE-58. Walker, L.C. (1995) The southern pine region. In: Barrett, J.W. (ed.) Regional Silviculture of the United States. John Wiley & Sons, New York, pp. 271–334. Ward, F.R. (1995) California’s forest products industry: 1992. USDA Forest Service Resource Bulletin PNW PNW-206. Whitney, G.G. (1994) From Coastal Wilderness to Fruited Plain. Cambridge University Press, Cambridge, UK. Williams, T. (1989) Incineration of Yellowstone. Audubon 1989(1), 38–89. Youngs, R.L. (2000) A right smart little jolt: loss of the chestnut and a way of life. Journal of Forestry 98(2), 17–21.

2 Regeneration Ecology I: Flowering, Fruiting and Reproduction Characteristics

Introduction Ecologically, the terms ‘regeneration’ and ‘reproduction’ are closely associated. In the narrow biological sense, regeneration refers to the regrowth of lost or destroyed parts of organs. However, regeneration also can be used in a population context to refer to ‘rebirth’, e.g. the rebirth of a forest after its destruction by natural or human causes. The latter meaning is useful in ecology and silviculture because it connotes population process, and has been so used in the literature of those fields, albeit inconsistently (cf. Harper, 1977; Keeley, 1981; Bartolome et al., 1987; Muick and Bartolome, 1987; Helms, 1998). We herein use regeneration to refer to the ecological processes involving the establishment, growth and population changes of juvenile trees, co-occurring plants, and their propagules rather than to a physiological process (e.g. see Grubb, 1977). In this context, juvenile trees are those not yet capable of flowering and producing seed, which for most of the oaks takes 15–25 years.1 Similarly, the term ‘reproduction’ has a narrow biological definition referring to sexual or asexual mechanisms by which organisms generate others of their own 1Among

kind.2 Like ‘regeneration’, ‘reproduction’ also can more generally imply the process of reproducing something. We have chosen to use the term reproduction to refer to individual or populations of juvenile trees (or other plants) already reproduced. Unlike regeneration, reproduction in this context does not connote process. Although the two terms have been treated as synonymous in the silvicultural literature, with reproduction relegated to obsolescence by Helms (1998), we believe it is conceptually important to distinguish between objects (young trees) and process (forest renewal). The regeneration of oak forests accordingly can be defined as a multifaceted ecological process. It includes the flowering, fruiting and seed dispersal of mature trees, as well as the germination, seedling establishment, growth and population changes of oak reproduction and associated plants. Forest regeneration thus involves time frames related to stages of stand development. In even-aged silviculture (Chapter 7), the regeneration period for oaks (at the stand scale) may span the last two and the first two decades of a rotation,3 during which reproduction becomes established and develops into the new stand. In con-

the oaks of eastern United States, exceptions include sprout-origin bear oaks and chestnut oaks, both of which can produce acorns as early as the end of the third growing season. Sawtooth oak, an Asian species, can produce acorns by age 3 on trees that are only 1 m tall (Nakashizuka et al., 1997). Certain seedling propagation methods combined with selecting progenies with early flowering can be used to obtain early acorn production in some species. 2The American Heritage Dictionary, 3rd edn. Houghton-Mifflin, Boston, Massachusetts. 3A rotation is the period between the establishment and final harvest of an even-aged stand. 54

Regeneration Ecology I: Reproduction

trast, uneven-aged silviculture (Chapter 8) and old-growth forests (Chapter 9) are characterized by a regeneration period that is essentially continuous or frequently periodic. Within any time frame, each step of the regeneration process is beset with uncertainties and unknowns that have contributed to our inability, in some cases, to successfully manage and sustain oak forests. Yet, oaks have thrived throughout North America for millennia seemingly without the help of humans. Much of this apparent enigma can be untangled by considering what we know about the regeneration process, which begins with the oak flower.

Flowering Oaks of the United States fall into one of three species groups. Species in the white oak group require one growing season to complete their reproductive cycle, while most of the species in the red oak group require two growing seasons. The five species in the intermediate group, which are found only in western North American, also require two growing seasons (Tucker, 1980). Oaks are monoecious, i.e. they produce male and female flowers on the same tree. All trees that have attained flowering age and size usually bear both kinds of flowers in abundance each year. In many species, flowering does not occur until trees are 15–25 years old. The annual regularity of flower production in oak contrasts with its irregular acorn production, which may range from bumper acorn crops in some years to poor or no crops in others. In a sexually mature oak, buds of several types occur along the outer branches comprising current-, 1-year-old and 2-year-old shoots. Some buds are strictly vegetative, i.e. they produce only leaves and other vegetative structures. Other buds are the progenitors of male and female flowers. Some produce only male flowers (catkins) whereas others produce both male flowers and vegetative structures. Female flowers arise from tissues located in leaf axils (Cecich and Larsen, 1997).

55

In white oak, strictly vegetative buds account for about half of all buds and the remaining half represent approximately equal numbers of male and female buds (Cecich and Larsen, 1997). The distribution of buds also varies with position along the branch. About 72% of all buds occur on current-year shoots, 26% on 1-year-old shoots and 2% on 2-year-old shoots. The latter are all vegetative buds occurring in leaf axils. Among the buds that produce flowers, female flower buds occur only on current year shoots whereas male flowers are approximately equally represented on 1year-old and current shoots. However, in the red oak group, female flowers in different states of development may be present on both current year and 1-year-old shoots at the same time. In both the white oak and red oak groups, female flowers occur throughout the crown but are most abundant in the upper crown. In contrast, male flowers occur mainly in the top and middle portions of the crown (Cecich and Larsen, 1997). Of the species native to the United States, flowering is most completely described for white oak (Turkel et al., 1955; Sharp and Chisman, 1961; Stairs, 1964; Mogensen, 1965, 1975; Sharp and Sprague, 1967; Merkle et al., 1980; Feret et al., 1982), although our knowledge of flowering in other oaks is growing (Cecich, 1997). Accordingly, the following account of the ontogeny of oak flowering focuses on white oak. Discussion of other oaks is limited to illustrating differences among species and species groups.

Male flowers The male (staminate) flowers of oaks develop in the axils of scale leaves of the current vegetative buds, or separate male buds bear the staminate flowers (Fig. 2.1A). Species in the white oak group initiate catkin primordia in the buds of shoots produced the year before acorn maturation. Species in the red oak group initiate catkin primordia in the buds of shoots formed two years before acorn maturation. In white oak, catkin initiation occurred in late May

56

Chapter 2

(A)

Bud scales

Catkin

Staminate flowers

Bract

(B)

Inflorescence Perianth Filament

Pollen sacs

(C)

Pollen sac

Filament

Pollen grains

Fig. 2.1. The male (staminate) oak flower. (A) Catkins bearing numerous flowers originate from the axils of bud scales. (B) A single male flower with pollen sacs. (C) Pollen sacs split open to release pollen grains when environmental conditions are favourable. (From Cecich, 1994.)

in Virginia (Merkle et al., 1980), and about a month later in Pennsylvania (Sharp and Chisman, 1961). In Virginia trees, the catkin primordia were macroscopically distinguishable by early August. Individual staminate flowers formed within the primordial catkins in late June or early July and were structurally complete before the onset of dormancy in October. These struc-

tures resumed development in mid-March and the catkins emerged from the bud in early April; further north in Pennsylvania they emerged in late April. In eastern forests, swelling and opening of male flower buds usually is later in the white oak group than in the red oak group. When catkins in the white oak group are emerging or slightly drooping (semi-


pendent), pollen dispersal is already complete. Emergence of white oak catkins occurs when daily temperature minima are 50°F (10°C) or higher for 10 days (Sharp and Chisman, 1961). They emerge from the base of the inner scales of buds clustered at the ends of the woody twigs of the previous year’s growth. Numbers of catkins range from three to ten or more per twig tip and are erect shortly after emerging. Within a few days they become semipendent and are usually fully pendent and in full bloom, or anthesis, within 5–12 days (Fig. 2.2). The pollen matures while catkins are elongating. The catkins appear first in the topmost branches and emergence progresses downward. Meiosis, the process that reduces chromosome numbers by one-half, begins in the male flowers of white oak when catkins have elongated slightly beyond the bud scales. In the same geographic area, scarlet oak and bear oak (both in the red oak group) begin similar activity about 2 weeks earlier at the time of early bud swell (Stairs, 1964). The diploid (2n) chromosome number, i.e. the number before reduction division in cells takes place, is 24 for all oak species investigated. In white oak, the individual male flowers (Fig. 2.1B) within the catkins reach anthesis 11–16 days after emergence from the bud (Sharp and Chisman, 1961). The basal portions of the catkins mature first. The time from catkin emergence to the completion of pollen shedding (Fig. 2.1C) ranges from about 11 to 19 days. The topmost branches shed pollen first in forestgrown trees, but the reverse may occur in open-grown trees. Pollen shedding occurs when catkins are 3–6 inches long. Individual trees shed most of their pollen within a 48-hour period if weather is favourable. In New York, pollen shedding in white oak lags that of scarlet oak and bear oak by 2 weeks (Stairs, 1964). About the time white oak catkins are semipendent, new unfolded leaves appear. However, there is a lag in further leaf expansion if catkins contain pollen. At the time of pollen shedding, usually during the last 2 weeks of May in central Pennsylvania, leaf length averages about 2 inches in white oak

57

Fig. 2.2. Northern red oak catkins in full bloom in mid-May (northern Wisconsin). Leaves are not yet fully expanded. (Authors’ photo.)

and 3 inches in chestnut oak (Sharp and Chisman, 1961). The day after pollen dispersal, leaf area increases by about 50%. It is thus possible to identify trees that have shed pollen by the state of leaf development. The period of arrested leaf expansion may favour pollen dispersal by minimizing canopy interference. Dispersal of oak pollen is by wind and usually occurs before there is significant insect activity. This contrasts with many associated hardwoods, which flower later and are insect pollinated. Pollen dispersal in white oak occurs only on days when relative humidity is less than 45% for several hours (Sharp and Chisman, 1961). To complete pollen shedding, both white oak and chestnut oak require 2 to 3 such days with air temperatures of 63 to 69°F (17 to

58

Chapter 2

21°C). In contrast, dwarf chinkapin oak (a shrubby member of the white oak group) only needs a few hours of such weather to disperse pollen. A growth chamber study confirmed the adverse effects of high humidity on bear oak (a shrubby member of the red oak group), which produced no immature (first-year) acorns when humidity during the flowering period was 61–70% (Wolgast, 1972). When humidity was reduced to 38–50%, a significantly larger proportion of the flowers (10.5% of 467) produced immature acorns. A positive association between high relative humidity and acorn production was reported for white oak in Missouri when relative humidity was averaged over the 1week pollination period (Cecich and Sullivan, 1999). This apparent discrepancy with other studies points out the importance of the observed time interval used to express the effect of relative humidity (or other weather factors). If only a few hours of low humidity occurring over a few days during the pollination period are sufficient for effective pollen shedding (Sharp and Chisman, 1961), measurements of relative humidity should then represent this time-dependent sensitivity if the intent is to directly relate relative humidity to pollination and acorn production. Otherwise relative humidity measurements and their apparent effects become potentially confounded with other factors. Such confounding arises from correlated but indirect relations between humidity and other weather variables such as air temperature. A light wind of 5–8 miles per hour is further conducive to pollen shedding. However, in closed-canopy forests, the canopy may act as a windbreak that minimizes the influence of light winds on the ripening catkins. The canopy nevertheless offers little protection against strong winds during dry rainless periods. In open fields, oaks shed pollen on the windward side several hours before shedding occurs on the leeward size when weather conditions are otherwise favourable. High winds on humid, cool days do not induce pollen shedding and prolonged cool, wet weather

can cause over-ripened catkins to drop intact with anthers filled with pollen. Although male flowers can tolerate light frosts, temperatures below 25°F (4°C) are lethal (Sharp and Chisman, 1961). In northern Utah, a 5 May freeze averaging 27°F (3°C) across 37 Gambel oak sites killed semipendent male flowers (Neilson and Wullstein, 1980). Subsequent acorn production was less than in nearby areas unaffected by the freeze. In bear oak in Massachusetts, the number of male flowers per branch at or near the time of anthesis decreased from the top to the bottom of a slope representing a 66 ft (20 m) vertical gradient (Aizen and Kenigsten, 1990). Within a given topographic position, the number of male flowers also decreased with decreasing tree height (i.e. taller trees produced more male flowers). A higher incidence of low temperatures and frost occurred at lower points along the combined vertical topographic and tree-height gradient. Thus frosts, prolonged desiccating winds, and prolonged cool humid weather can be detrimental to male flowering and pollen dispersal. Any of these events may result in reduced or failed acorn production.

Female flowers In white oak in Virginia, the inflorescences bearing female (pistillate) flowers are first identifiable, but only microscopically, in early August of the year before acorn maturation (Merkle et al., 1980). These primordial inflorescences usually arise from the axils of the top three or four leaves within the developing bud. There they remain invisible to the unaided eye and continue to develop until late September or early October. Resumption of development occurs the following spring when dramatic changes occur during the last week of March and the first week of April. Each inflorescence then elongates rapidly, usually producing two to three functional flowers, but sometimes up to five. At this stage, flowers remain within the swollen buds.


Like male flowers, the time of appearance and development of female flowers varies with the weather and thus year. In central Pennsylvania, the leaves of white oak usually emerge from the buds during the second week of April (Sharp and Sprague, 1967). As the buds expand they carry with them the female inflorescence. A single elongating stalk then pushes the one to five pistillate structures upward. The individual female flowers become visible within 5–10 days after emergence of the male catkins (Merkle et al., 1980). At this time, the pendent catkins are about 50% of their final length, new vegetative shoots are about 2–3 inches long, and new leaves are 1–2 inches long (Sharp and Sprague, 1967). At this stage of development, the flowers resemble miniature acorns that envelope all the requisite, preformed floral structures including the cupule (which develops into the ‘cup’ of the mature acorn). The stalks of pistillate white oak flowers continue to elongate until the flowers are mature. Maturation occurs when the three stalk-like styles bearing the pollen-receptive organs, the stigmas, extend beyond the surrounding floral structures (the perianth) (Fig. 2.3a). Pollen grains released into the air by the male flowers land on the stigmas, germinate, and produce pollen tubes (Fig. 2.3b). A series of events involving callose plugs then occur, which isolate the contents of each pollen tube from other pollen grains (Fig. 2.3c–d). Pollen tubes then stop growing in 2–3 weeks (Fig. 2.3e). There is then a pause in pollen tube growth, which is resumed in 2 weeks in the white oak group and after about 13 months in the red oak group. Pollen tubes then enter the locules (Fig. 2.3f), and the egg of one ovule is fertilized; the other five eggs and ovules soon die. Approximately 1 month after pollination, or about 1 July in Pennsylvania, meiosis and fertilization have taken place (Turkel et al., 1955; Sharp and Sprague, 1967). Fertilization of Gambel oak (a white oak) in Arizona also occurs then (Brown and Mogensen, 1972). In central Missouri, fertilization of white oak flowers occurs during mid-June (Cecich, 1997). Fertilization of black and northern red oak

59

flowers, which require two growing seasons for ovule maturation, occurs during mid-June of the second year for black oak and 2 weeks later for northern red oak (Fig. 2.3g). Although not all eggs of the six ovules in each ovary may become fertilized, fertilization occurs at nearly the same time on those that do (Mogensen, 1965). Of 30 ovaries studied in three species (Gambel, white and black oak), nearly half the 180 ovules observed aborted because they were not fertilized (Mogensen, 1975). Thus, on average, 2.8 ovules per ovary were fertilized. The remaining aborted ovules were approximately equally divided between those with zygote or embryo failure (28%) and those without an embryo sac (26%). However, all but one fertilized embryo normally aborts very soon after fertilization to produce a one-seeded fruit. Although the reason for the abortion of fertilized ovules is not well understood, the functional ovule may be fertilized before the others. This, in turn, may suppress the growth of the other ovules in a process analogous to apical dominance in stems (Mogensen, 1975). Occasionally, more than one ovule develops to seed maturity within the same ovary to produce a multiple-seeded acorn. Some trees are especially prone to producing these (Coker, 1904; Buchholz, 1941). In four of five species observed in Illinois, from 1 to 2% of acorns were multiseeded, which suggests that multiseeds may be common in most oaks (Hosner, 1959). By mid-summer of the year of acorn maturation, acorn enlargement and embryo development begin in most species. The generalized sequence of acorn maturation events in central Missouri is illustrated in Fig. 2.3h–l. However, rates of acorn maturation may vary among species, climates, weather conditions and other factors. In white oak and chestnut oak in Pennsylvania, the acorn and cupule begin rapid growth in late July. By the first week in August the acorn begins to emerge from the cup. By mid-August, the acorns are about one-third filled out. By the last week in August they are full size (Fig. 2.4). By mid-September mature acorns are usually

60

Chapter 2

a – the flower

b – pollen grains

c – 1st callose plug

d – 3rd plug

e – pollen tubes cease growing

f – locules are entered

g – fertilization

h – embryo enlarging

i – 1 July

j – 15 July

k – 29 July

l – mature acorn


dropping from the cups. In years of heavy acorn production, all acorns on a stalk may mature, but often only one develops and the remainder die (Sharp and Sprague, 1967).

Factors Affecting Acorn Production Most oak species produce good acorn crops one year in three or four (Olson, 1974). Although consistent annual production of male and female flowers is an inherent characteristic of oaks, the large annual variation in acorn production is at least partially controlled by environmental factors (Sharp and Sprague, 1967; Sork and Bramble, 1993). Weather-related factors directly influence the early flowering process, as discussed in the previous section. Those and other factors also may impact the later stages of acorn development. The total number of female flowers produced and their per cent survival together explain about 90% of the variation in acorn production in black, northern red and white oaks in Missouri (Sork and Bramble, 1993). In contrast, factors related to site productivity such as soil nutrients, topographic position and site index (Chapter 4) appear to have little or no influence on acorn production (Tryon and

61

Carvell, 1962; Wolgast, 1972). Yet few studies have rigorously examined site productivity as a factor in acorn production in most species. Some trees appear to be under complete genetic control. Some never produce acorns even when they grow in favourable environments and occupy superior crown positions (Wood, 1934; Downs and McQuilkin, 1944; Sharp, 1958; Sharp and Sprague, 1967). Thus environment, genetics and the interactive effects of those factors are all potential determinants of acorn production.

Weather Good white oak acorn crops in Pennsylvania occurred when temperatures in late April were above normal but followed by a sudden drop in temperature in early May (Sharp and Sprague, 1967). A 10-day period in late April with warm nights followed by a 2-week period with cool nights produced good acorn set. Poor acorn production occurred when there was an even progression in mean daily April and May temperatures. In a Missouri study, warm spring temperatures during the year of acorn maturation were positively correlated with the size of acorn

Fig. 2.3. (Opposite) The development of a pollinated pistillate flower to a mature acorn. (a) The flower contains six ovules (two shown) in each of three locules (chambers) of the ovary. Three styles with stigmatic tips contain the transmitting tissue through which the pollen tubes reach the locules. (b) Pollen grains, released into the air by male (staminate) flowers, land on the stigmas, germinate, and produce pollen tubes. (c) A callose plug isolates the contents of each pollen tube from other pollen grains, which soon fall from the stigma. (d) Successively formed plugs isolate the growing tips of pollen tubes from earlier-formed remnants, which maintains turgor pressure in the tip. The third plug appears in the pollen tube after about 1 week. (e) Pollen tubes stop growing in 2–3 weeks. After a 2-week pause, the pollen tubes of species in the white oak subgenus resume growth towards the locules. In the red oaks, the pause lasts about 13 months. Thus, the red oaks require two growing seasons for pollinated flowers to mature into acorns whereas white oaks require only one growing season. (f) Pollen tubes enter the locules at an open space called the paracarpous region. This space allows pollen tubes from any style to randomly enter any of the six ovules. (g) Fertilization of the egg in one ovule, which occurs several days after the pollen tube enters the locule, produces an embryo. The other five eggs and ovules soon die. (h) The embryo enlarges and other parts of the flower differentiate into the cup and shell of the acorn. (i) By early July, the flower is recognizable as a young acorn. The bright green shell begins to protrude from the cup in response to the growing embryo. (j) By mid-July, the embryo is heart-shaped. The two lobes of the ‘heart’ represent the early stages of cotyledon development. (k) By late July, the embryonic axis (root tip, shoot tip and hypocotyl) appear between the two cotyledons. (l) Mature acorns begin to drop from their branches in September. The acorns are filled with starch, lipids and proteins. (Drawings and text contributed by Dr Robert A. Cecich, USDA Forest Service, North Central Research Station, Columbia, Missouri.)

62

Chapter 2

Fig. 2.4. A cluster of nearly mature northern red oak acorns (left) in late August of their second year of development. Acorns in their first year of development are visible in the leaf axils of the current-year’s shoot (right); flower styles are still intact and visible. (Authors’ photo.)

crops in white, black and northern red oaks (Sork and Bramble, 1993). In contrast, summer drought and late spring frost the previous year were negatively associated with acorn production. In another Missouri study, high daily maximum temperatures were negatively associated with white oak acorn production and flower survival and black oak flower survival (Cecich and Sullivan, 1999). Growth chamber studies of immature bear oak acorns revealed that the acorns did not develop under conditions of high humidity (Wolgast, 1972). This effect may have been directly related to a failure in pollen shedding (Sharp and Chisman, 1961). Other investigators have reported little or no evidence that relative humidity, precipitation, drought, site quality and wind velocity affect acorn yield (Downs and McQuilkin, 1944; Sharp and Sprague, 1967; Feret et al., 1982). These disparate findings may be related to inherent differences among species. But even for a given species, discrepancies in findings may arise from differences in how the duration, intensity and frequency of potential predictors of acorn production are measured and quantitatively expressed. Foresters and physiologists have long speculated about frost as a cause of failure in

acorn production. However, the frequency of killing frosts in central Pennsylvania indicates that during any given 50-year period, frost accounts for only 4 years of complete acorn crop failures and 4 years of severely reduced crops (Sharp and Sprague, 1967). In white oaks in Missouri, a late spring freeze reduced the number of white oak flowers to about 20% of non-frost years and emergence of surviving flowers was about 2 weeks later than normal (Cecich and Sullivan, 1999). In immature bear oak acorns in New Jersey, spring temperatures of 29 to 32°F (2°C to 0°C) for 3 consecutive days in late May reduced acorn yields (Wolgast and Trout, 1979). Acorn yield increased as the proportion of the crown not damaged by frost increased (Fig. 2.5).

Premature abscission Most female flowers of oaks succumb to premature separation (abscission) from their stalks (peduncles) before they can develop into acorns. Although premature abscission occurs primarily during the pollination–fertilization period, it continues through the acorn maturation period (Turkel et al., 1955; Williamson, 1966) (Fig. 2.6). About 52% of acorns of several species caught in acorn traps from mid-


63

30

300

25

250 Pounds 200

20

150

15

100

10

50

5

0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

Pounds of acorns per acre

Number of acorns per acre

Number

0 0.7

Proportion of crown area undamaged Fig. 2.5. Bear oak acorn production (fresh weight) in New Jersey in relation to crown frost damage. (From Wolgast and Trout, 1979, used with permission.)

summer on in Missouri were immature (Christisen and Kearby, 1984). These acorns fell continuously from early July through August. Observed losses from premature abscission in white oak range from 68 to 99% of female flowers observed early in the growing season (Williamson, 1966; Feret et al., 1982). Yet the lower abscission rate represented a ‘good’ acorn crop. Substantial premature abscission thus can occur even in years of high acorn production. Premature abscission is common not only to oaks but to many other hardwoods including willow, poplar, basswood, black locust and elm (Kramer and Kozlowski, 1979; Kozlowski and Pallardy, 1997). Fruitset can be diminished by premature flower abscission, limited pollination or fertilization, inadequate nutrition, embryo abortion, premature abscission of young fruits and other factors (Greulach, 1973). Premature abscission of fruits also may involve competition for photosynthates, which can occur among fruits or between fruits and vegetative growth (Abbott, 1960; Luckwill, 1977).

In the oaks, tree hoppers and other insects have been implicated in premature abscission (Cecich et al., 1991; Cecich, 1994). Hail storms also may damage pistillate oak flowers (Cecich, 1997; Cecich and Sullivan, 1999). The long slender peduncles (stems) of pistillate white oak flowers may be especially susceptible to hail damage. However, the long styles of black oak flowers may be more susceptible to hail damage than the sessile styles of white oak flowers. If the styles of oak flowers are broken off before pollen tubes cease growing, flowers will not be fertilized and will abort. Hail also can dislodge fragile catkins and thus eliminate the pollen source if hail occurs before pollen is shed. Failed or limited pollination in any species can result from inadequate production of pollen caused by unfavourable weather conditions or a limited period of stigma receptivity to pollen. Also, fertilization failure may result from pollen sterility, failure of ovary development, slow pollen tube growth and failure of meiosis. If pollen tube growth is slow, the tube may not reach the embryo sac or sperm viability may be lost before the tube reaches the sac.

64

Chapter 2

fertilized flowers, only 26–29% (5–11% of all flowers) developed into acorns during the 4 study years. However, among individual clones, that percentage ranged from 0 to 42%. Statistical analysis showed that clones explained a significant proportion of variation in flowering, fertilization and fruiting. But acorn yield was most strongly correlated with clonal differences in percentage of flowers fertilized and only weakly correlated with flower abundance. Though there was significant variation among clones in flower abundance, the greater limitation to acorn production was low fertilization rate.

140 A

Number of live flowers

120 100 80 60 40 20 0 May

July

Sep

May

July

Sep

Month 90 B

Variation in acorn production

80 Number of live flowers

70 60 50 40 30 20 10 0 Apr

May

June

July

Aug

Sep

Date

Fig. 2.6. Survival of female black oak and white oak flowers in central Missouri trees. (A) Black oak flowers. The line connected by circles is the average of five flower-bearing branches in the upper crowns of eight trees in 1991/92; unconnected circles show the range. The line connected by triangles is the average of 11 trees in 1992/93; the unconnected triangles show the range. (B) White oak flowers in nine trees in 1992; the mean and range are shown as in (A). (Data courtesy of Robert A. Cecich, USDA Forest Service, North Central Forest Experiment Station.)

A white oak clonal seed orchard in Tennessee demonstrated the strong genetic control over premature abscission (Farmer, 1981). Among 111 trees representing 31 clones, the percentage of fertilized flowers averaged over all clones ranged from 19 to 38% 7–11 years after orchard establishment. But among individual clones, that percentage ranged from 0 to 75%. Among

The most prominent and consistent attributes of acorn production are large year-toyear and tree-to-tree variation (e.g. see Koenig et al., 1994; Nakashizuka et al., 1997). This variation occurs within species and locales. In white oak and northern red oak in West Virginia, tree-to-tree variation was significantly greater than year-to-year variation in trees (Tryon and Carvell, 1962). In bear oak, tree-to-tree variation in acorn production was under strong genetic control (Wolgast, 1972). Forest-grown trees of three acorn-producing ranks (low, medium and high) transplanted to a homogeneous nursery environment conformed to their original rankings. In contrast, soil factors related to the forest environment did not explain a significant proportion of the observed variation in acorn production before transplanting. White oak similarly demonstrated strong genetic control over flowering and acorn production in a clonal seed orchard (Farmer, 1981). The large variation in time of acorn fall among individual willow oaks and water oaks during the same year suggests that factor also may be largely under genetic control (Cypert and Webster, 1948). Although there are a few reported studies on the allocation of net annual biomass production to acorns, on average the proportion apparently is relatively small but highly variable among trees. During a


5-year study of 29 open-grown sawtooth oaks in Japan, acorns accounted for an estimated 2.8% of net annual biomass production when averaged across years (Nakashizuka et al., 1997). However, among trees this proportion varied from 0 to 20%. The proportion of production allocated to acorns was independent of tree size. The absolute allocation to acorn production increased in proportion to whole-tree production but only after whole-tree production reached 0.76 kg (1.7 lb) per year. Above that threshold, acorn production was closely proportionate to leaf production. Other factors being equal, large trees generally produce more acorns than small trees (Downs and McQuilkin, 1944; Goodrum et al., 1971), and open-grown trees produce more acorns than trees growing in a closed-canopy forest (Sharp, 1958; Sharp and Sprague, 1967). Because both tree size and stand density can be measured and controlled silviculturally, these relations have practical application in managing oak forests for acorn production. There also are large differences among species in acorn-producing potential. Methods for assessing and predicting acorn crops based on these and other relations together with silvicultural guidelines for sustaining acorn production are presented in Chapter 9.

Periodicity Year-to-year variation in acorn production within an individual tree or population of trees can result from the single or joint effects of variation in inherited factors that control the timing and frequency of flowering and fruiting, and environmental factors. Environmental factors that can reduce acorn production below a tree’s inherent potential often occur as essentially random events, which may obscure the expression of any inherently regular pattern, or periodicity, in acorn production. Obscuring factors include weather events unfavourable to flowering and acorn production (Sharp and Sprague, 1967; Sork and Bramble, 1993; Koenig et al., 1996; Cecich, 1997), and insect damage to leaves (Crawley and Akhteruzzaman, 1988), flowers (Cecich et

65

al., 1991) and acorns (Myers, 1978; Christisen and Kearby, 1984). Because of confounding environmental factors, determining the existence of periodicity in natural environments usually requires the use of statistical methods that can reveal inherent cycles, or alternatively, experimental studies in controlled environments designed to eliminate or separate the confounding factors. Periodicity in acorn production thus refers to an inherent potential of a tree to recurrently produce acorn crops at a fixed time interval. Periodicity may characterize an individual tree or a population of trees of the same species. Even though an individual tree may be inherently able to produce acorns at a regular interval, the population as a whole may not if it includes trees with no discernible pattern of acorn production, trees that produce at regular intervals but of different lengths, or trees that produce at the same interval length but which are out of phase with other trees in the population. Periodicity in the occurrence of large or above average acorn crops, or ‘masting’, is also a possibility. The occurrence of periodicity and various patterns of periodicity is usually determined by observing trees known to be consistent, albeit not necessarily prolific, acorn producers (e.g. see Koenig et al., 1994). The occurrence of periodicity and periodicity characteristics in acorn production vary among oak species. In some species in the white oak group, acorn production is noticeably periodic. In the absence of adverse weather, acorn crops tend to occur every other year and good crops about once every 4 years (Downs and McQuilkin, 1944; Goodrum et al., 1971; Beck, 1977; Myers, 1978; Christisen and Kearby, 1984; Koenig et al., 1994). This pattern suggests that white oaks require at least 1 year to recover after a large investment in acorn production. Periodicity in the white oak group also is apparently synchronized, with all fruitful trees within a species, locale and year producing at their inherent individual, but often environmentally constrained, capacity (Sharp, 1958; Myers, 1978).

66

Chapter 2

For four California oak species, selected weather factors (temperature and precipitation) explained 38–78% of the variation in mean annual acorn production (Koenig et al., 1996). In valley and blue oaks, which mature in one growing season, mean annual April temperature (which includes the fertilization period) was positively associated with acorn production. In canyon live and coast live oaks, which mature in two growing seasons, total precipitation from September through the next acorn fall was positively associated with acorn production. The influence of weather factors thus may obscure any tendency in periodicity as confirmed by the predominance of reported irregularities in acorn production (Sharp, 1958; Grisez, 1975; Godman and Mattson, 1976; Sork et al., 1993). Periodicity tends to be less consistently expressed in the red oaks than in the white oaks (Sharp, 1958; Tryon and Carvell, 1962; Beck, 1977). For example, in northeastern Wisconsin northern red oak bumper crops (defined as 91% or more of maximum potential) occurred at 7-year intervals (Godman and Mattson, 1976). Yet, such crops occurred as few as 2 years apart over the 21-year study period. Good or better crops (60% or more of potential) occurred at 3-year intervals. Other species in the red oak group behave similarly (Sharp, 1958; Christisen and Kearby, 1984). In the red oak group, the combined effects of apparently asynchronous acorn production and unpredictable weather often obscure any tendency towards periodicity. This suggests that the apparent asynchrony of acorn production among individual red oaks may result in lower year-to-year variation in acorn production at the population level than in the white oaks (Tryon and Carvell, 1962). Acorn production among species is generally asynchronous (Sork and Bramble, 1993; Koenig et al., 1994). However, among five species of California oaks observed for 12 years, acorn production was relatively synchronous within species despite large variation among trees in numbers of acorns produced (Koenig et al., 1994). The same study found no evidence of periodicity in

the occurrence of large crops (‘masting cycles’) at the population level but found that such cycles did occur in individual trees. Individual trees of some species also produced large crops in successive years, but within a population such occurrences occurred no more often than expected by chance. There was no evidence of masting cycles at the population level in any of the five species studied, which included two species in the white oak group, two in the red oak group (including one evergreen oak), and one evergreen oak in the intermediate group. The investigators noted that regular masting cycles have yet to be demonstrated for any oak species. One method of detecting inherent periodicity in acorn production in individual trees is based on the correlation between the acorn production of a tree in one year with its production in other years. The correlation coefficients for various time intervals or ‘lag times’ (e.g. 1, 2, 3 and 4 years before and after a specified time) thus can be used to statistically identify time intervals associated with acorn production (positive correlations), time intervals associated with the absence of production (negative correlation), or the absence of correlation (0 correlation). For example, the individual-tree correlation coefficients for a 1-year time lag during a 10-year study period can be derived from the nine pairs representing year 1 vs. year 2, year 2 vs. year 3, . . . year 9 vs. year 10. Similarly, the correlations for 2-, 3- and 4-year or longer time lags can be calculated. The method can be used to derive, for each tree, a series of correlation coefficients each representing a given time lag. If this were done for several trees, the correlation coefficients for each tree then could be evaluated. A general consistency among individuals in the signs of the correlation coefficients (positive or negative) and their statistical significance (i.e. the likelihood that they differ from 0) provides a basis for accepting or rejecting the hypothesis of inherent periodicity in acorn production. For a given species, the average correlation coefficient of several trees for a given time lag can be used as the basis for testing the hypothesis. Accordingly, if


prior acorn production does not affect current acorn production, the mean correlation coefficient should not differ significantly from zero for any given time lag. A positive correlation would indicate a significant positive relation between a given time lag and acorn production, and a significant negative correlation would indicate a negative relation, i.e. inhibition of acorn production. This method of assessing periodicity also has the advantage of removing differences related to possible asynchrony in acorn production among individual trees among years because correlations are derived for each tree by lag interval rather than year by itself. Such statistical tests provided empirical evidence of an inherent 3-year cycle in white oak acorn production, i.e. years of high production tended to be followed by two successive years of little or no production (Sork and Bramble, 1993). There also is evidence of inherent 2- and 4-year cycles for black oak and northern red oak, respectively (Fig. 2.7). Significant negative correlations tended to occur in intervening years. Because all three species were observed in the same locale and yet had different cycles, variation in weather does not provide a likely explanation for variation in apparent periodicity. Evidence for periodicity was strongest for white oak because none of the numerous, measured weather variables were correlated with its apparent 3-year periodicity cycle. Among California oaks that have been similarly studied, a 3-year cycle was statistically evident in blue oak (a member of the white oak group) and canyon live oak (a member of the red oak group with a 1-year acorn maturation period) (Koenig et al., 1991) (Fig. 2.8A). Evidence for periodicity was less conclusive in valley oak (a member of the white oak group), canyon oak (intermediate group) and California black oak (red oak group) (Fig. 2.8B). Plausible explanations for inconclusive results include: (i) the ‘real’ absence of periodicity; (ii) real but weakly expressed periodicity; (iii) the overriding influence of one or more environmental factors during a given study

67

period; (iv) errors in measuring acorn crop size; (v) a sample size that is too small (i.e. too few trees observed) to substantiate that a ‘real’ association as expressed by a correlation coefficient of a given value is statistically significant; and (vi) possible genetic variation in periodicity within a species. Statistical relations (Figs 2.7 and 2.8) suggest that periodicity may be of two types: manifest and latent, i.e. readily expressed and seldom expressed. Accordingly, evidence of manifest periodicity would express itself as a consistent temporal pattern of positive and negative correlations such as that observed in blue oak (Fig. 2.8A). In contrast, latent periodicity would occur in species or populations of trees with an inherent periodicity that is seldom or never expressed because of acute sensitivity to adverse environmental events or frequent exposure to such events. Demonstrating the existence of latent periodicity would require experiments that eliminate or control the factors obscuring periodicity. In the absence of such experiments we can only speculate about the nature of possible genetic variation in periodicity. The occurrence of populations that include trees with both high and low lagtime consistencies (e.g. as expressed by the correlation patterns in Figs 2.7 and 2.8) nevertheless would be consistent with a population that possesses high genetic variation in periodicity. An alternative hypothesis is that, among trees that produce acorns, some individuals are inherently periodic whereas others are inherently non-periodic.

Spatial variation within tree crowns Acorns are usually unevenly distributed throughout the tree crown. However, acorns in open-grown trees were more evenly distributed than acorns in trees in closed stands. In closed stands, most acorns occurred on branches exposed to light (Sharp and Sprague, 1967). Post (1998) found that even in open-grown northern red oaks, production was greatest in the lower section of the south-facing side of crowns. Shading on one or more

Chapter 2

Number of years prior to current year

68

4

3

2

1

–0.8

–0.6

–0.4

–0.2

0.0

0.2

0.4

0.6

0.8

Correlation of current with prior year’s acorn production

Black oak Northern red oak White oak Fig. 2.7. Correlation of current acorn production (mature acorns per tree) with prior years’ production for three species in Missouri. The significant positive correlations (right side of graph) suggest 2-, 3- and 4year acorn production cycles (time lags) for black, white and northern red oaks, respectively, based on 13, 15 and 12 trees (n), respectively, observed for 8 years. Each correlation coefficient shown is the average of the individual-tree coefficients within a species/time-lag interval (see text for discussion). Asterisks (*, **, ***) indicate significance at P < 0.05, < 0.01, and < 0.001, respectively, based on t tests with n 1 degrees of freedom. (Redrawn from Sork et al., 1993, used with permission.)

sides of the crown resulted in a non-discernible pattern of production, although branches exposed to light tended to produce more acorns than those that were not. In a relatively open-grown stand (6 oaks acre1) of coast live oak on a south-facing slope in California, acorn numbers were about two times greater on south-facing sides of crowns than northeast- and northwest-facing sides (Lewis, 1992). Although numbers of acorns in lower and midcrowns were about three times greater than in upper crowns, the opposite trend occurred in trees in another stand that were pruned and irrigated (Lewis, 1989). Such inconsistencies suggest that environmental factors may influence the withincrown distribution of acorns.

Acorn destroying insects and other organisms may prefer different parts of the crown and thereby influence the distribution of sound acorns. The infestation of coast live oak acorns by filbert weevils and filbert worms was 1.5–3 times greater on northeasterly side of crowns than on south and northwesterly sides (Lewis, 1992). Preferences for cooler crown aspects may reflect insect avoidance of overheating. Moreover, the occurrence of acorns that were split open (possibly from bursting from within) occurred on south and northwesterly sides of crowns, making them vulnerable to invasion by ants, wasps, microbial pathogens and other parasites (Lewis, 1992). The numerous confounded biotic and physical factors that apparently can influence the spa-



6

A

5 4 3 Valley oak Blue oak Coast live oak

2 1 –0.4


69

6

–0.2

0.0

0.2

0.4

B

5 4 3 2 Canyon live oak California black oak

1 –0.4

–0.3

–0.2

–0.1

0.0

0.1

0.2

0.3

Correlation of current with prior year’s acorn production Fig. 2.8. Correlation of current acorn production with prior years’ production for five species in California. (A) Correlations for three species requiring 1 year for acorns to mature based on 87, 57 and 63 valley, coast live and blue oaks, respectively. For blue and coast live oaks, the patterns of significant correlations suggest a 3-year cycle. For valley oak, the pattern of correlations suggest that it is not strongly cyclic. (B) Correlations for two species requiring 2 years for acorns to mature based on 21 trees of both species. For both species in graph B, the correlations are inconclusive in relation to cyclic patterns because of the absence of significant positive correlations. For all species, correlations are based on the number of acorns counted in a 30-second sampling period on each tree in each of 10 consecutive years. Asterisks (*, **, ***) indicate significance at P < 0.05, < 0.01 and < 0.001 respectively, based on binomial tests. (Redrawn from Koenig et al., 1991.)

tial distribution of acorns within crowns (especially sound acorns) thus complicates generalizing about such distributions other than that they vary greatly among trees.

Acorn Predation and Dispersal Many organisms consume acorns including insects, millipedes, fungi, birds and mammals. Consumption by one or more biotic agents is

often so complete that in any given year few acorns remain to germinate and become seedlings. Although this consumption contributes to oak regeneration failures (Marquis et al., 1976; Galford et al., 1991c), acorns are valuable food for many birds and mammals because of their high caloric content, nutritiousness and availability during seasons when other food is often scarce. Oak forests thus can be managed specifically for acorn production to benefit wildlife (Chapter 9).

70

Chapter 2

The term dispersal is used to refer to the transport of acorns from their place of origin (i.e. tree crowns or the ground directly beneath them) to some other location irrespective of the distance, mode of transport or subsequent fate of the acorns. Because of their relatively large size and mass, acorns are not dispersed by wind. Although gravity and water may affect localized dispersion, birds and mammals are the most important dispersers of acorns. Dispersal by animals may carry acorns to places that are more favourable or less favourable for germination and seedling establishment than their site of origin. But even under the most favourable circumstances, the oaks pay a price for animal-mediated dispersal because at least some of the dispersed acorns are consumed by the dispersers. Whether an oak benefits from dispersal by a given disperser depends on: • Where the acorns are dispersed • The number dispersed • The pattern of dispersal (scattered or concentrated) • The proportion of acorns consumed by the disperser (including acorns consumed at their place of origin before dispersal has occurred). With respect to the oak, dispersal is ineffective if acorns are largely dispersed to habitats unfavourable for maintaining acorn viability and seedling establishment. If the proportion of total numbers of acorns dispersed is very small, the impact of the dispersing agent (whether positive or negative) on oak regeneration may be negligible. Numbers of dispersed acorns being equal, dispersers that scatter individual acorns or small groups of acorns are likely to benefit oak regeneration more than dispersers that cache large quantities of acorns in one or a few locations. Dispersers that consume virtually all the acorns they disperse clearly represent some loss to potential seedling establishment. Dispersers also may consume significant quantities of acorns at their place of origin. If proportionately few acorns are consumed by a disperser, dispersal may be advantageous to the oak if acorns are dispersed in significant numbers to habitats favourable for seedling establishment.

Insects: destroyers of acorns Insects may begin destroying acorn crops during flowering. However, the focus of this section is on insects as destroyers of already formed acorns (from immature to mature) through the germination period. During these stages of development, there is a predictable sequence of damaging events and agents involving not only insects but other organisms including bacteria, fungi, algae, protozoa, nematodes, mites and other organisms (Winston, 1956). Insects often initiate damage before acorns mature. For example, the adult female Curculio weevil can chew through the immature acorn’s shell to deposit her eggs inside the acorn. The developing larvae then consume part or all of the acorn’s interior. Invading insects also can carry pathogenic fungi and bacteria into the acorn, which may kill the embryo even though the insect itself may not. Direct insect damage occurs during the several weeks required for the larvae to develop. The developing larvae usually consume most acorns by mid- to late-autumn while they remain on the tree or shortly after they fall. The larvae eventually cut exit holes in the shell of the acorn, making it accessible to other organisms. One of these is the acorn moth, which lays its eggs on or in the exit hole. Their larvae feed on the remainder of the embryo and on the faeces of the previous occupant. After acorns fall, fungi and other decomposers take over the final stages of destruction until the acorn is thoroughly decayed and eventually incorporated into the soil (Winston, 1956). Only the acorns that escape this fate are available to wildlife and forest regeneration. Insects typically destroy 50% or more of acorns annually (Gibson, 1982; Kearby et al., 1986). In some years, destruction approaches 100%. Although many kinds of insects destroy acorns (Winston, 1956), most of the damage through the germination period is caused by a relatively small number of insect groups (Gibson, 1972, 1982; Kearby et al., 1986). These include:


• Acorn weevils (Curculio spp. and Conotrachelus spp.) • Moths (the filbertworm moth and the acorn moth) • Acorn gall wasps (cynipids) • Nitidulid sap beetles. Acorn insects are sometimes studied using acorn traps and ground emergence traps from which the insects and acorns can be identified and counted (Fig. 2.9). The rela-

71

tive importance of the various acorn insects expressed as a percentage of insectinfested acorns is indicated by state and regional studies (Table 2.1). Curculio weevils in any given year typically account for 50–80% of infested acorns, and Conotrachelus weevils for about 5–40%. The moths and gall wasps each typically account for about 0–20% of infestations. However, these percentages can vary

Fig. 2.9. Acorn collection traps (on stakes) and insect emergence traps (on ground) used to study acorn production and acorn insects in a Missouri forest. (Photo courtesy of Dr William H. Kearby, Regional Entomologist (Retired), Wisconsin Department of Natural Resources.) Table 2.1. The relative importance of the four major acorn-insect groups expressed as a percentage of insect-infested acorns. Species, region and years observed White oak

Insect group

Curculio weevils Conotrachelus weevils Filbertworm and acorn mothe Acorn gall wasps a

Ohio 1961a 79.1 15.1 5.7 0.1

Range-wide 1962b 1963c 59.5 38.0 2.5 post oak > scarlet oak > chestnut oak > black oak > white oak (Bourdeau, 1954; Mowbray and Oosting, 1968; Racine, 1971; Wuenscher and Kozlowski, 1971; Seidel, 1972; Fralish et al., 1978; Eyre, 1980; Parker et al., 1982). Oak seedlings are also relatively intolerant of shade, although species differ somewhat. For example, among five common upland oaks in the eastern United States, their shade tolerances have been ranked as follows: white oak > chestnut oak > northern red oak > black oak > scarlet oak (Burns and Honkala, 1990). Rates of photosynthesis in oak seedlings increase with increasing light intensity up to about onethird of full sunlight, but increase little with further increases in light intensity (Kramer and Decker, 1944; Bourdeau, 1954; Loach, 1967; McGee, 1968; Musselman and Gatherum, 1969; Phares, 1971; Shafer, 1971; Wuenscher and Kozlowski, 1971). Light levels under dense forest canopies often fall below 2% of that in the open (Hanson et al., 1986). Under those conditions, oak seedlings cannot live for long (Hanson et al., 1986; Crunkilton et al., 1992; Vivin et al., 1993). Low light levels also proportionately reduce the allocation of carbon to roots and increase the proportion allocated to aboveground parts of the seedling (Gottschalk, 1985; Kolb and Steiner, 1990). In turn, this impedes the development of a large root:shoot ratio even if the seedling survives. Shade tolerance in the oaks may not be as fixed as species’ shade tolerance ratings might infer. For example, oaks have adapted to variation in light conditions by adjusting their time of spring budbreak according to their exposure to light the previous year. Seedlings or saplings growing beneath a forest canopy begin flushing about 1 week earlier in the spring than open-grown oaks, and before the overstorey begins to leaf out (McGee, 1976, 1988, 1997). Understorey oaks therefore begin spring growth when both light and soil moisture conditions are favourable. This pattern of spring budbreak was common to all six of the oak species

92

Chapter 2

observed (white, black, scarlet, post, chestnut and northern red oaks) (McGee, 1997). Other understorey species including hickories, white ash, red maple and sugar maple also began budbreak and growth earlier under shade than in the open. However, some species began growth sooner in the open (serviceberry) or showed no consistent pattern between budbreak and the presence of a forest canopy (sassafras, yellow-poplar and blackgum) (McGee, 1986). In northern red oak seedlings, time of budbreak is also associated with elevation of acorn source. Across the elevational range of 1400–4600 ft above sea level in western North Carolina, acorns from the lowest elevation source flushed 11 days earlier, on average, than the highest elevation source (McGee, 1974, 1997). This difference in flush date was consistent among the four planting sites, which ranged from 1500 to 5400 ft in elevation. These apparent inherited differences in flushing time thus represent genetic adjustments to the timing of budbreak that coincide with ‘safe’ periods of growth initiation. A later onset of growth with increasing elevation provides a frost avoidance mechanism, whereas the earlier onset of budbreak at lower elevations is consistent with spring frost risks and the competitive advantage of the early onset of growth at lower elevations. Collectively, these relations in flushing pattern in oaks have practical silvicultural implications with respect to potential frost damage associated with early flushing related to seed source/planting site relations, and the timing of overstorey removal (McGee, 1997).

Seedling Sprouts Shoot dieback and root : shoot ratio Seedling sprouts are seedlings whose shoots have died back and resprouted one or more times. They are often the predominant form of oak reproduction growing beneath the forest canopy (Fig. 2.19). Sprouts can originate from dormant buds located anywhere along the stem between the root collar and

A

B

Fig. 2.19. Shoot dieback in oak seedling sprouts growing beneath a closed forest canopy in central Missouri. (A) Partial dieback (leafless portion of stem) in a northern red oak seedling sprout. New shoots from dormant buds were initiated 10 inches above ground and near ground line. (B) Complete dieback in a black oak seedling sprout to ground line (leafless stem on right) with two basal sprouts. Both seedling sprouts are approximately 30 inches tall (to tips of tallest dead stems), and both had previously died back several times. (USDA Forest Service, North Central Research Station photograph.)


the terminal bud cluster. Dieback and resprouting appear to be important processes in the life of oak reproduction. Those processes are facilitated by a ‘bud bank’ comprised of a reservoir of visible and (to the naked eye) invisible buds that are continually being formed on new branches. In addition, older buds persist at the root collar and on older stems and branches (Wilson, 1993). Shoot dieback is common in all oaks growing under a forest canopy, but is especially prominent in Mediterranean and semidesert climates and elsewhere in droughty uplands. In those environments, oak stands are dominated by species with morphological and physiological adaptations to surviving repeated burning and water stress (Wuenscher and Kozlowski, 1971; Grimm, 1984; Abrams, 1990). Oak reproduction growing beneath a forest canopy is subject to stresses that periodically reduce shoot mass and leaf area through the process of recurrent shoot dieback. Surviving seedling sprouts thus

93

tend to develop larger root: shoot ratios as they age. In turn, high root: shoot ratio and large root mass enable oak reproduction to opportunistically respond to forest disturbances. The growth potential of oak reproduction is not expressed until it is released from the growth-inhibiting shade of the parent stand. This occurs when the forest canopy is partially or completely destroyed by disturbances resulting from fire, windthrow, insects, disease, timber harvesting or other events. Then if the requisite size of the established oak reproduction has been obtained, it can produce two or more long flushes of shoot growth each growing season after disturbance (Johnson, 1979a; Dickson, 1991). Variation in the root : shoot ratio within populations of oak reproduction produces great variation in shoot growth patterns among individual trees of different basal diameters (which in turn are correlated with root mass) (Fig. 2.20). Whether oaks become an important part of the succeed-

Net cumulative shoot growth (ft)

6 5

a

4

b

3

c d

2 e 1 0 1

3

5

7

9

11

13

15

17

Growth week Apr

May

June

July

Aug

Month Fig. 2.20. Patterns of shoot growth in oak sprouts for a range of parent-stem basal diameters (and thus root : shoot ratios). The growth shown occurred the second year after clearcutting in a stand in the Ozark Highlands of Missouri. (a) Black oak sprout originating from a 7-inch diameter stump; (b) black oak sprout originating from an 8-inch diameter stump (the sprout died back from frost during weeks 4 and 5 and subsequently flushed twice); (c) white oak sprout originating from a 1-inch diameter parent stem; (d) black oak sprout originating from a 4-inch stump; (e) white oak sprout originating from a parent stem less than 0.6 inches in diameter. (Adapted from Johnson, 1979a.)

94

Chapter 2

ing stand thus depends on the pre-disturbance development of significant numbers of seedling sprouts with large roots and high root : shoot ratios. Oak reproduction is otherwise usually at a competitive disadvantage. This is especially true of the reproduction of the drought tolerant upland species which, even under optimal conditions, grow slowly until roots are large enough to support rapid shoot growth. Recurrent shoot dieback thus appears to be an important aspect of the evolutionary development and adaptive strategy of oaks.

Occurrence of shoot dieback Recurrent shoot dieback in oak reproduction differs from the ‘dieback and decline’ of older trees, natural pruning or twig abscission. The latter is a normal, nonpathological abscission process occurring in current twigs of mature oaks (Millington and Chaney, 1973). The phenomenon known as dieback and decline is common to older (non-juvenile) oaks. It is believed to be initiated by drought stress or insect defoliation, which may set up conditions conducive to disease and further insect attack (Staley, 1965; Nichols, 1968; Ammon et al., 1989; McCracken et al., 1991; Wargo and Haack, 1991; Dwyer et al., 1995; Jenkins and Pallardy, 1995). Dieback and decline of older trees usually result in the complete death of the tree (see Chapter 7). Shoot dieback also can be directly caused by insect damage to shoots of both juvenile and older oaks. But in juvenile oaks, resprouting regardless of the cause of dieback often follows shoot dieback. Whenever oak reproduction grows under a forest canopy, recurrent shoot dieback is a normal part of its life cycle. Dieback is especially prevalent on dry sites where reproduction survives long enough to accumulate over several acorn crops. There, oak seedlings and seedling sprouts must endure low light intensities together 4The

with high water stresses. Under those conditions, oak reproduction is likely to subsist near its compensation point 4 and thus on the edge of death (Hanson et al., 1986; Crunkilton et al., 1992; Vivin et al., 1993). This process reinforces the oak’s strategy of translocating to the root system most of the photosynthate produced during its juvenile growth period (Fig. 2.17). Through periodic reduction of shoot mass, the shoot can be minimized as a sink for the modest amount of photosynthate it produces when growing in shade (Hanson et al., 1986; Crunkilton et al., 1992; Dey and Parker, 1996). Seedling sprouts that survive recurrent shoot dieback thus eventually develop large root systems and large root : shoot ratios. The consequences of recurrent dieback enable oak reproduction to respond opportunistically to natural or silvicultural events that eliminate the shade-imposing overstorey. Shoot dieback in oak reproduction is linked to several factors including growing season water stress and spring frost (Bourdeau, 1954; Johnson 1979a; McGee, 1988). When soil moisture is ample such as in irrigated nursery beds, shoot dieback is uncommon. In contrast, recurrent dieback is characteristic of oak reproduction growing beneath a forest canopy (Liming and Johnston, 1944; Merz and Boyce, 1956; Tryon and Carvell, 1958; Sander, 1971; Abrams, 1990; McClaran and Bartolome, 1990; Crow, 1992). Experiments with seedlings of five upland oak species grown in pots showed that shoot dieback occurred in some seedlings when they were deprived of moisture (Bourdeau, 1954). After shoot dieback, new shoots developed from surviving dormant buds located below the point of dieback in some individuals of all but one of the species observed; some seedlings died. Water stress thus can directly result in seedling mortality or in shoot dieback followed by the growth of new shoots from surviving buds located below the dieback (Vivin et al., 1993). Succulent spring shoot growth is also susceptible to late spring

compensation point occurs at the light level where carbohydrate breakdown through respiration balances carbohydrate gain through photosynthesis.


frosts. Frost-induced diebacks in the spring are usually shortly followed by the growth of new shoots from dormant buds below the dieback. Although shoot dieback in oak reproduction commonly occurs during the growing season as a result of spring frosts and summer water stress, it also occurs during the dormant season. Winter dieback and mortality of northern red oak seedlings occurring between November and spring was observed at the onset of spring growth in an Ohio greenhouse experiment of potted seedlings grown from seed (Wright et al., 1989). After overwintering out of doors, both mortality and dieback were greater in seedlings experimentally subjected to high water-stress and artificially imposed root injury (roots partially removed by severing) the previous summer than in seedlings that were only moderately stressed or that had uninjured roots. In a Missouri study, shoot dieback of English oak seedlings planted in a clearcut occurred between time of planting in late October and the completion of the first flush the following May (Johnson, 1981). Moreover, the average length of dieback was greater for bare root seedlings than for container grown seedlings with intact roots. Similarly, in autumn-planted northern red oak, the frequency of winter dieback of 10 cm or more was greater in bare root than in container grown seedlings. Frequency of dieback also increased with increasing shoot length, decreased with increasing basal diameter of stems for both types of seedlings, and for a given seedling size was more frequent in clearcuts than in shelterwoods (Fig. 2.21). Dormant season dieback thus was influenced by seedling characteristics (size and morphology) and the presence or absence of a forest canopy. Collectively, these observations confirm that shoot dieback in oaks does occur during the winter. They further show that at least some of the observed variation in the frequency and amount of it is associated with factors that increase or decrease a seedling’s vulnerability to winter desiccation including:

95

• Large numbers of functional small roots at the start of the dormant season (greater in container grown than in bare root or root-injured seedlings), which are associated with a low frequency or small amount of dieback. • Large shoot diameters (positively correlated with root mass and root surface area), which are negatively associated with frequency of dieback. • Long stems, which are positively associated with frequency of dieback. • The presence of a forest canopy (potentially reducing frequency and/or intensity of freeze–thaw cycles), which is associated with reduced frequency of dieback. The observed occurrence of shoot dieback in oak reproduction is consistent with the principle of plant segmentation. According to this principle, the design of the plant’s water transport system (i.e. its hydraulic architecture) favours the preservation of the lower stem over the more vulnerable and less essential shoot tips (Zimmermann, 1983). By extension, shoot dieback in oak reproduction may be related to seasonal cycles of water stress involving: • Loss of fine roots to desiccation (root shedding) or injury in late summer (Head, 1973; Joslin and Henderson, 1987; Wright et al., 1989; Yin et al., 1989), leading to reduced absorption of water during late summer and the following dormant season. • Xylem dysfunction (the formation of xylem embolisms, i.e. air blockages, and tyloses) occurring during both dormant and growing seasons, leading to excessively reduced hydraulic conductivity especially in the terminal sections of stems (Zimmermann, 1983; Tyree, 1989; Cochard et al., 1992; Sperry and Sullivan, 1992; Tyree and Cochard, 1996). • Bud desiccation and mortality and the failure of the stem to initiate cambial growth in the spring between dead buds and the next lower live bud.

Chapter 2

0.6 Bare-root 0.4 Containergrown 0.2

80

ho

60

ls

12 Initia 10 l sho ot dia 8 mete r (mm )

ot

14

itia

0.0

len gt h

(c m )

100

6

In

Probability of shoot dieback >10 cm

96

Fig. 2.21. Estimated probabilities of occurrence of winter shoot dieback (10 cm or more) in northern red oaks autumn-planted under shelterwoods in the Ozark Highlands of Missouri. Estimates are shown in relation to initial seedling size and type of nursery stock. Dieback was observed shortly after the completion of spring flushing of the first field growing season. The response surface shown is for trees planted under shelterwoods thinned to 60% stocking based on Gingrich’s (1967) stocking equation. The model is based on 2950 planted trees and is given by the logistic regression equation: P = 1/(1 + exp(–(–6.1098 + 0.0483•H – 0.1510•D – 0.4251•S – 1.1932•T))), where P is the estimated probability of shoot dieback ≥ 10 cm between the time of planting (late October) and spring flushing (mid-May), H is initial tree shoot length (cm), D is initial basal diameter 2 cm above the root collar (mm), S is presence of shelterwood (shelterwood absent = 0, shelterwood present = 1), and T is type of nursery stock (bare-root = 0; container-growing = 1). For all parameter estimates, P < 0.001 except S (P = 0.01). Based on the Hosmer–Lemeshow goodness-of-fit test, differences between observed and estimated probabilities did not differ significantly (P = 0.0001). Estimates for bare-root stock represent the average of 2+0 seedlings and 1+1 transplants, which did not differ significantly ( = 0.05). Estimated dieback probabilities for trees planted in clearcuts are 0.01 to 0.11 larger than for trees planted under shelterwoods, depending on initial shoot diameter and length. (From authors’ data.)

Survival of buds through summer droughts and the dormant season is essential to initiating cambial growth the next spring (Romberger, 1963). This growth is initiated by the downward translocation of growth regulators originating in buds (Wareing, 1951). If buds in the terminal cluster die from winter desiccation (or any other cause), subsequent cambial initiation and thus a new annual ring will fail to develop between the terminal cluster and the next lower bud (Zasada and Zahner, 1969). If a new growth ring fails to develop, there can be little or no upward conduction

of water from the roots because older annual rings in oaks are largely nonfunctional (Zimmerman, 1983). Shoots may then die back to the next lower living bud (Fig. 2.19). The downward progression of shoot dieback in oak reproduction frequently can be observed in the spring when new flushes of shoot growth from terminal buds fail to develop and one or more of the normally suppressed lateral buds produce new shoots. In upland oak forests, there is much variation in the amount and frequency of seasonal shoot dieback (Crow, 1992). In dry


upland oak forests, this is reflected in the large variation in differences between root and shoot ages among individuals in the same population (Merz and Boyce, 1956; Powell, 1976). Some of this variation may be associated with differences in root : shoot ratios, absolute root size and absorptive capacity, and thus the ability of roots of oak reproduction to replace water losses in stems and buds before and during the dormant season. Despite its role in reinforcing the apparent root-centred growth and survival strategy of oak reproduction, recurrent shoot dieback exacts a price. Loss of shoots during the dormant season can negatively affect spring root growth. In northern red oak seedlings, the artificial removal of seedling shoots 6 inches above the root collar in late autumn reduced spring root growth of seedlings transplanted to greenhouse root observation chambers by up to 59% (Fig. 2.16C). Similar results have been observed in other oak species (Lee et al., 1974; Farmer, 1975, 1979). The collective evidence suggests that loss of shoots during the dormant season, whether through artificial removal or natural dieback, can reduce regulatory root growth promoters that originate in shoots and buds (Vogt and Cox, 1970; Carlson, 1974; Larson, 1975; Farmer, 1979). In turn, reduced root growth in the spring may predispose reproduction to water stress and thus shoot dieback later in the growing season (Johnson, 1979b). Also, during the shoot regrowth period after shoot dieback, there is a greater allocation of growth to shoots than to roots (Cobb et al., 1985; Kruger and Reich, 1989). This consequently lengthens the time required for roots to reach the requisite size for supporting, under field conditions, multiple flushing and thus rapid height growth. The physiological effects of shoot losses vary among oak species (Lee et al., 1974) and with other factors. In a young clearcut in Wisconsin, rates of transpiration and net photosynthesis (per unit leaf area) in northern red oak seedlings with pruned shoots (cut off 1 inch above the root collar to simulate dieback) were 30% greater than in unpruned seedlings (Kruger and Reich,

97

1989). However, leaf areas of pruned seedlings were 25% lower than those of unpruned seedlings. The joint effects of increased photosynthetic rate and reduced leaf area thus were largely compensatory. The higher photosynthesis and transpiration rates of the pruned seedlings were attributed to possible (but unmeasured) increases in water availability or other factors associated with increased root : shoot ratio resulting from shoot pruning (Kruger and Reich, 1989). This conclusion is consistent with the behaviour of northern red oak seedlings grown in a greenhouse (Parker, 1949). Transpiration rate (and thus photosynthetic rate) increased as root surface area: leaf area ratio increased (Fig. 2.18B). Shoot pruning of cherrybark oaks growing under forest canopies in Mississippi, combined with mid-storey and understorey competition control, significantly increased the height growth of seedlings and seedling sprouts based on 2-year observations (Lockhart et al., 1991). However, pruned seedlings in most cases did not completely regain the height lost to pruning. Survival also was lower in the pruned population of seedlings. The investigators nevertheless concluded, based on physiological and morphological measurements, that pruning-simulated dieback and subsequent resprouting confers a growth advantage to cherrybark oak reproduction. Low to moderate losses of leaves and shoots to defoliating insects and browsing animals can similarly result in growth and survival advantages to oak seedlings (Wright et al., 1989; Welker and Menke, 1990). But defoliation also can cause depletion of starch reserves in oak roots, which is augmented by drought (Parker and Patton, 1975). In blue oak seedlings in a California savanna, rapidly induced water stress combined with severe defoliation caused 100% seedling mortality after 2 years, whereas survival under slower rates of induced water stress combined with complete defoliation was associated with 80% survival (Welker and Menke, 1990). Defoliation of oak seedlings in the autumn also may reduce their growth the following spring (Larson, 1975).

98

Chapter 2

Despite some apparent negative effects, the widespread occurrence of shoot dieback in oak reproduction at the population level may be consistent with a growth and survival strategy that sacrifices large numbers of individuals for a high probability of obtaining a few individuals with large roots and attendantly rapid height growth potentials. Whatever the physiological explanation, shoot dieback in oak reproduction occurs within populations that are highly variable physiologically, morphologically and genetically. Moreover, individual seedlings and seedling sprouts occupy relatively small, homogeneous microenvironments within which there is a high probability of occurrence of insufficient soil moisture, nutrients or light. Spatial variability of those resources at small scales thus may explain much of the observed variation in the frequency and amount of oak shoot dieback.

Stump Sprouts and Related Growth Forms Definitions and origins Once a seedling has died back and resprouted, its continued survival can lead to other growth forms. Various names have been applied to these growth forms, depending on how and where they originate on the parent stem or root system. One growth form is the stump sprout. These sprouts originate from dormant buds at or near the base of the stump of a cut tree (Fig. 2.22A and B). However, they also can arise from the bases of trees topkilled by fire (Fig. 2.22C). In the Central Hardwood Region, stump sprouts are often defined as those originating from a cut tree 2 inches dbh and larger (Roach and Gingrich, 1967). However, that definition has not been universally adopted. Biologically, the distinction between a stump sprout and a seedling sprout is arbitrary because all oaks, from small seedlings to large standing trees, have some potential to produce basal sprouts when the parent stem is cut. Moreover,

when wind, fire or other factors destroy an oak stand, sprouts may develop from the bases of trees that have broken off or from standing trees with dead tops. Sprouts also can arise from dormant buds on the crowns of large, mature root systems having no aboveground stumps (Fig. 2.22D). This growth form, sometimes called a ‘grub’ (Curtis, 1959), is common in savannas, old fields and other disturbanceperpetuated communities where recurrent fire or decay destroys stumps and where, simultaneously, low overstorey density maintains high light levels sufficient for sprout survival over many decades. Such disturbances favour the accumulation of this form of reproduction, which like seedling sprouts, recurrently die back and resprout. The dieback process extends over a longer period than would be possible under the low light conditions of a closed canopy forest. Where favourable light conditions are maintained, oak grubs may attain ages of several hundreds of years (Curtis, 1959). Like seedling sprouts, stump sprouts and grubs originate from dormant buds at or near the root collar. These buds, connected to the pith of the tree by elements called bud traces, remain just beneath the bark by annually elongating the width of the annual ring (Liming, 1942). They do not develop further unless their vascular connections to the crown are severed by cutting or are otherwise interrupted. As long as the crown of the parent tree is alive, living buds under the bark usually remain in a dormant state imposed by growth-suppressing regulators translocated from the crown (Vogt and Cox, 1970). When buds fail to elongate each year, they are lost as a potential source of sprouts. Sometimes these losses are offset by buds that multiply by branching (Kramer and Kozlowski, 1979). The rate of bud branching and bud mortality changes with the age and size of the tree. The balance of these processes partially determines the spatial distribution and number of sprouts per stump that develop after cutting the parent tree. Also,


99

A

B

C

D

Fig. 2.22. Basal sprouting in oaks. (A) A newly initiated black oak stump sprout originating from a basal bud on a recently cut tree. (B) A ten-stem clump of northern red oak stump sprouts 10 years after the parent tree was cut; the dominant stem is 27 ft tall. (C) Basal sprouts of scarlet oak originating after the top of the parent tree was killed by fire. (D) An excavated white oak seedling sprout with a large root (or ‘grub’) many times the mass of the shoots (from a sandy outwash plain in northern Lower Michigan). This grub probably originated after the parent tree was top-killed by fire and persisted through recurrent sprouting in the understorey long after the top and stump disappeared. The root section shown is about 6 ft long; scale is shown by hard hat near the root tip (centre foreground). (USDA Forest Service, North Central Research Station photographs.)

100

Chapter 2

some buds fail to produce shoots after the parent tree is cut because of the physical resistance of the bark to shoot emergence or inhibition by sprouts already emerged (Wilson, 1968). Physical resistance increases with increasing bark thickness and thus tree diameter. As trees become older and larger, their ability to sprout consequently decreases (Johnson, 1977; Weigel and Johnson, 1998). Stool sprouts and root sprouts differ from the other forms of vegetative reproduction because they originate from buds formed from callous tissue around wounds or other tissues (Kramer and Kozlowski, 1979). These are called adventitious buds and unlike dormant buds do not have bud traces extending to the pith of the tree. Stool sprouts develop adventitiously from the cut or wounded surfaces of stumps. They are usually weakly attached to the stump and are therefore often short-lived. Root sprouts, as the name implies, similarly originate from roots. In many oaks, stool sprouts and root sprouts are unimportant ecologically and silviculturally because they tend to be either short-lived or rarely occur. California black oak often produces stool sprouts, but not sprouts from dormant buds, when sprouts originate from high stumps of large trees (McDonald, 1990b). Other oaks reported to produce stool and/or root sprouts include canyon live oak (Thornburgh, 1990), blue oak (McDonald, 1990a), Oregon white oak (Stein, 1990), water oak (Adams, 1983) and swamp chestnut oak (Edwards, 1990a). Some arid-region oaks have evolved special belowground structures that produce sprouts. In Gambel oak, there are three distinct root-like belowground structures: lignotubers, rhizomes and true roots (Tiedemann et al., 1987). Lignotubers are burl-like structures with adventitious buds. These buds are the primary source of new shoots when tree crowns are killed. Several lignotubers may be connected by rhizomes, which have fewer buds. Rhizomes facilitate the development of wide spreading clones (Muller, 1951) that quickly develop from

the seedling state (Christensen, 1955) (Fig. 2.23). Later, the physical separation of individual lignotubers resulting from the death or destruction of connected lignotubers within a clone represents a form of plant multiplication (and thus population growth) that is important in the regeneration of rhizomatous oaks. Lignotubers and rhizomes are anatomically similar to stem wood in that both possess a pith, buds and bud traces. In contrast, roots are devoid of those features. Rhizomes in oaks are associated with arid and semi-arid climates (Muller, 1951), where environment is unfavourable for seedling establishment. In addition to Gambel oak, other rhizomatous oaks in the United States include Havard, sandpaper, live, Vasey, Texas live, Brewer, turbinella, Mohr, Ajo and huckleberry oaks (Muller, 1951). Except for live oak, all are shrubby species confined to the arid southwest.

Frequency of sprouting The proportion of oak stumps that produce sprouts after trees are cut has been related to parent tree diameter and age, and site quality (Johnson, 1977; Weigel and Johnson, 1998). In general, sprouting decreases with increasing tree diameter, age and site quality (Fig. 2.24). But other factors such as season of cutting and shading also can affect stump sprouting in hardwoods. For some species of oaks, stump sprouting is greater for trees cut or killed during the dormant season than during the growing season (Clark and Liming, 1953). However, some of the live oaks of the western United States sprout prolifically regardless of season of cutting (Longhurst, 1956). Few studies have evaluated the effect of shading on the stump sprouting of oaks or other hardwoods. However, all stumps sprouted in a 28-yearold water oak plantation in Louisiana regardless of thinning intensity (Gardiner and Helmig, 1997). In contrast, sugar maple stumps exposed to full light sprouted more frequently than shaded stumps (Church, 1960)


101

(A)

(B)

(C)

Fig. 2.23. Regeneration by rhizomes in live oak. (A) Aerial shoot originating from a rhizome; (B) the distribution of rhizomes in a single clone (black dots indicate locations of aerial shoots); (C) a clump of trees in a single clone. (From Muller, 1951, used with permission.)

Sprout growth and mortality Young stump sprouts that arise from polesize and larger parent trees are, in effect, mature root systems connected to juvenile shoots. Even though stump sprouts start out as small as a new seedling (Fig. 2.22A), their large root systems buffer them from many of the adverse competition and site effects associated with the more limited site resources available to smaller reproduction. Stump sprouts thus have the potential to grow rapidly. During

their first decade, open-grown stump sprouts in eastern United States may produce four or more flushes of shoot growth per year totaling 3 ft or more even under droughty conditions (Johnson, 1979a; Reich et al., 1980; Cobb et al., 1985) (Fig. 2.20). Even when shaded by an overstorey of 34 ft2 acre1 of basal area, water oak stump sprouts grew at a rate of 1.7 ft year1 for the first 5 years. However, this rate slowed to 0.75 ft year1 by age 7 when overstorey density had increased to 52 ft2 acre1 (Gardiner and Helmig, 1997).

102

Chapter 2

1.0

0.8

B 70

0.8

70

110

90

70 0.6

0.6

0.4

90

Parent tree age 50

0.4

70

0.2

0.2 White oak 0.0

90

110

110

110

Chestnut oak 0.0

0

4 8 12 16 20 Parent tree dbh (inches)

1.0

24

0

50

50 70

50 70

0.6

70

0.6

90 90

70 0.4

0.4

90

0.2

90 110

110 0.2

110

110

Scarlet and northern red oaks

Black oak 0.0

0.0 0

4

24

D

0.8 50


1.0

C

0.8 Sprouting probability

50 90

Site index (ft) 60 75

50 Sprouting probability

1.0 50

A


28

0

4


28

Fig. 2.24. Estimated sprouting probabilities (P) for oak stumps in relation to site index and parent tree diameter and age. P is the probability that a stump will have at least one living sprout at the end of the first growing season after cutting. (A) White oak: P = {1/[1 + exp(–(–7.6179 – 1.5760•ln(dbh) + 2.7069•ln(SI) – 0.00667•ln(dbh)•age))]}. (B) Chestnut oak: P = {1/[1 + exp(–(4.5719 – 1.5760•ln(dbh) + 2.7069•SI – 0.00667•ln(dbh)•age))]}. (C) Black oak: P = {1/[1 + exp(–(–1.7718 – 0.0014•dbh•age + 0.0469•SI))]}. (D) Scarlet and northern red oaks: P = {1/[1 + exp(–(–1.1012 – 0.0014•age•dbh + 0.0469•SI))]}, where dbh is dbh of the parent tree (inches), SI is black oak site index in feet at base age 50 (from Carmean et al., 1989), and age is age of the parent tree in years. Based on trees in southern Indiana. (From Weigel and Johnson, 1998.)


The large root mass of stump sprouts and their large carbohydrate storage and absorptive capacity, coupled with other physiological factors, facilitate multiple flushing in oaks. In contrast most other growth forms do not produce multiple flushes. Single-flush trees include mature oaks, shaded seedlings and seedling sprouts, and small seedlings and seedling sprouts under water stress (Cook, 1941; Johnston, 1941; Kienholz, 1941; Longman and Coutts, 1974; Borchert, 1976; Buech, 1976). The number of shoot flushes in oak stump sprouts declines with sprout age. In scarlet oak stump sprouts, numbers of flushes decreased from an average of about two per growing season the first year to about one by the fourth growing season (Cobb et al., 1985). By the fifth year, numbers of flushes approached that of the single-flush mature tree. Borchert (1976) hypothesized that the progression from multiple to single flushes as trees grow larger may be attributable to a declining root : shoot ratio that results in increasingly longer periods for roots and shoot to restore ‘functional balance’ following periods of shoot elongation and leaf expansion. The number and spatial distribution of sprouts around the stump also influence sprout growth. Excavation of English oak stump sprouts in Russia showed that the development of numerous sprouts that are well distributed around the stump can maintain the complete parent-tree root system (Kharitonovich, 1937). Such sprouts grew rapidly in comparison to sprouts with stems on only one side of the stump. The latter often withered and died unless the new sprouts developed an independent root system. Even then, sprouts were less vigorous than those connected to the living parent root system. An even distribution of sprouts around the stump increases the likelihood that the vascular connections of sprouts to roots facilitate the translocation of photosynthates to all sides of the

103

parent root system (Kharitonovich, 1937; Wilson, 1968; Kramer and Kozlowski, 1979). Numerous sprouts per stump were correlated with rapid early height growth in five species of oaks in the Ozark Highlands of Missouri (Johnson, 1977). Northern red oak stump sprouts in Wisconsin behaved similarly (P.S. Johnson, 1975). However, the initial advantage of high clump density may be quickly lost as competition between stems within clumps intensifies with age. In northern red oak, 20 or more stems per clump are commonly initiated. The resulting crowding of stems induces a rapid decrease in clump density. Through this self-thinning process, numbers of live stems per clump typically decline to about four by age 15 (Fig. 2.25). Multiple stems may persist for 50 years or longer (Fig. 2.26). Similar rates of self-thinning were observed in water oak stump sprouts (Gardiner and Helmig, 1997). The negative effect of persistent multiple stems on the growth rate of individual stems has been demonstrated by clump thinning studies (Haney, 1962; Wendel, 1975; Lamson, 1983; Johnson and Rogers, 1984; Lowell et al., 1989). The diameter of the parent tree, and correlatively the size of the root system, affects the growth of oak stump sprouts. For five species of oaks in the Ozark Highlands, the correlation between stump diameter and the heights of the dominant stem within young sprout clumps in clearcuts was negative for all species (Johnson, 1977). Whether the relation is positive or negative, however, may depend on the range of stump diameters observed. For example, average 4year shoot elongation of the tallest stem in black oak and white oak clumps ranging from small seedling sprouts up to large stump sprouts increased as stump diameter increased up to a threshold diameter of 6 inches (Johnson, 1979b). For larger stumps, sprout height growth

104

Chapter 2

Number of living stems per clump

24

Parent tree dbh (inches) 10 15 20

20

16

12

8

4

0 4

8

12

16

20

24

Clump age (years)

Fig. 2.25. Number of living stems in 382 northern red oak sprout clumps in relation to sprout age and parent tree dbh. Each data point represents 1 to 14 clumps in clearcut stands in southwestern Wisconsin. Lines are mean trends based on a regression model that accounts for 33% of the variation in number of stems. (From P.S. Johnson, 1975, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

Fig. 2.26. Four stems persisted in this 45-year-old clump of northern red oak stump sprouts in southeastern Minnesota. The largest stem is 14 inches dbh. (USDA Forest Service, North Central Research Station photograph.)

decreased with increasing stump diameter (Fig. 2.27A). The stump diameter associated with maximum sprout growth becomes more evident as the observed range of stump diameters increases. Although this general pattern of height growth in oak sprouts is common to five species in the Ozark Highlands of Missouri, species vary in the diameter (6–12 inches) associated with maximum growth (Fig. 2.27B). The maximum growth rate of reproduction along the continuum of parent stem diameters thus may represent the diameter where the optimum root : shoot ratio occurs most frequently in a genetically, physiologically and morphologically heterogeneous population of trees in a highly variable environment. Other factors also may be significant sources of variation in the height growth of oak sprouts. These include variation in site quality (Fig. 2.28), stand density and thus competition from surrounding trees, insect and frost damage, animal browsing and season of cutting. The effects of these factors may vary greatly among oak species and regions.


105

16 White oak

A

Black oak

4-year net shoot growth (ft)

14 12 10 8 6 Stump sprouts 4 2 0

2

4 6 8 10 Stump diameter (inches)

12

16 B

5th-year sprout height (ft)

14

White oak Scarlet oak

12

10

8

Black oak

6

Blackjack oak 5

10

15

20

Post oak 25

Stump diameter (inches)

Fig. 2.27. Height growth of oak stump sprouts (the tallest stem per clump) in relation to stump diameter in the Ozark Highlands of Missouri. (A) Predicted and observed 4-year net shoot growth of 43 black oak and 41 white oak sprouts during the first 4 years after clearcutting. Sprouts from stumps ≥ 2.4 inches in diameter represent ‘stump sprouts’ originating from the parent stand (‘overstorey’ trees). These stumps were measured 12 inches above ground. Sprouts from stumps < 2.4 inches in diameter (measured 4 to 6 inches above ground) represent ‘advance reproduction’, i.e. reproduction growing beneath the overstorey at the time of clearcutting. The estimate (curved line) illustrates the continuous non-linear relation between shoot growth and stump diameter across the two arbitrarily defined growth forms (i.e. advance reproduction and stump sprouts). The unexplained variation in shoot growth (R 2 = 0.39) may largely reflect the imperfect correlation between stump diameter and root size – the presumed (but unmeasured) primary ‘causal’ factor. Growth of the two species did not differ ( = 0.05). (From Johnson, 1979a.) (B) Predicted sprout heights of five species in 5-year-old clearcuts. Heights (H5) are based on the regression model: H5 = b0 [exp[–(b1D + b2•1/D)]] where D is stump diameter measured 6 inches above ground, and b1 and b2 are parameters estimated by regression analysis. (From Dey, 1991.)

106

Chapter 2

TSC

60

1.0

Sprout height (ft)

50

0.55 0.1

40 30 20 10 0 4

8

12

16

20

24

Sprout age (years) Fig. 2.28. Estimated and observed height of the tallest stem within young clumps of northern red oak stump sprouts in southwestern Wisconsin in relation to site quality expressed by the topographic site coefficient (TSC). TSC ranges from 0.1 (poorest sites) to 1.0 (best sites) and is based on depth of soil, aspect and slope position (see Chapter 4). (From P.S. Johnson, 1975, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

References Abbott, D.L. (1960) The bourse shoot as a factor in the growth of apple fruits. Annals of Applied Biology 48, 434–438. Abrams, M.D. (1990) Adaptations and responses to drought in Quercus species of North America. Tree Physiology 7, 227–238. Abrams, M.D. and Kubiske, M.E. (1990) Leaf structural characteristics of 31 hardwood and conifer tree species in central Wisconsin: influence of light regime and shade-tolerance rank. Forest Ecology and Management 31, 245–253. Adams, J.C. (1983) Water oak regeneration in the South’s upland bottomland. USDA Forest Service General Technical Report SE SE-24, pp. 177–179. Ahlgren, I.F. (1974) The effect of fire on soil organisms. In: Kozlowski, T.T. and Ahlgren, E.E. (eds) Fire and Ecosystems. Academic Press, New York, pp. 67–72. Aikman, J.M. (1934) The effect of low temperature on the germination and survival of native oaks. Iowa Academy of Science 41, 89–93. Aizen, M.A. and Kenigsten, A. (1990) Floral sex ratios in scrub oak (Quercus ilicifolia) vary with microtopography and stem height. Canadian Journal of Forest Research 68, 1364–1368. Ammon, V., Nebeker, T.E., Filer, T.H., McCracken, F.I., Solomon, J.D. and Kennedy, H.E., Jr (1989) Oak decline. Mississippi Agriculture & Forestry Experiment Station Technical Bulletin 151. Auchmoody, L.R., Smith, H.C. and Walters, R.S. (1994) Planting northern red oak acorns: Is size and planting depth important? USDA Forest Service Research Paper NE NE-693. Barnett, R.J. (1977) The effect of burial by squirrels on germination and survival of oak and hickory nuts. American Midland Naturalist 98, 319–330. Barrett, L.I. (1931) Influence of forest litter on the germination and early survival of chestnut oak, Quercus montana Willd. Ecology 12, 476–484. Bartolome, J.W., Muick, M.C. and McClaran, P. (1987) Natural regeneration of California hardwoods. USDA Forest Service General Technical Report PSW PSW-100, pp. 26–31. Beal, J.A. (1952) The more important insects of Duke Forest and the Piedmont Plateau. Duke University School of Forestry Bulletin 44, pp. 18–24.


107

Beck, D.E. (1977) Twelve-year acorn yield in southern Appalachian oaks. USDA Forest Service Research Note SE SE-244. Bent, A.C. (1964a) Life Histories of North American Jays, Crows, and Titmice, Part II. Dover, New York. Bent, A.C. (1964b) Life Histories of North American Woodpeckers. Dover, New York. Bent, A.C. (1964c) Life Histories of North American Nuthatches, Wrens, Thrashers, and Their Allies. Dover, New York. Bock, C.E. and Bock, J.H. (1974) Geographical ecology of the acorn woodpecker: diversity versus abundance of resources. American Naturalist 108, 694–698. Bonfil, C. (1998) The effects of seed size, cotyledon reserves, and herbivory on seedling survival and growth in Quercus rugosa and Q. laurina (Fagaceae). American Journal of Botany 85, 79–87. Bonner, F.T. and Vozzo, J.A. (1987) Seed biology and technology of Quercus. USDA Forest Service General Technical Report SO SO-66. Borchert, M.I., Davis, F.W., Michaelsen, J. and Oyler, L.D. (1989) Interactions of factors affecting seedling recruitment of blue oak (Quercus douglasii) in California. Ecology 70, 389–404. Borchert, R. (1976) Differences in shoot growth patterns between juvenile and adult trees and their interpretation based on systems analysis of trees. Acta Horticulturae 56, 123–130. Bourdeau, P. (1954) Oak seedling ecology determining segregation of species in Piedmont oak– hickory forests. Ecological Monographs 24, 297–320. Brezner, J. (1960) Biology, ecology, and taxonomy of insects infesting acorns. University of Missouri Agriculture Experiment Station Research Bulletin 726. Bragg, W.K., Knapp, A.K. and Briggs, J.M. (1993) Comparative water relations of seedling and adult Quercus species during gallery forest expansion in tallgrass prairie. Forest Ecology and Management 56, 29–41. Brooks, F.E. (1910) Snout beetles that injure nuts. West Virginia Agriculture Experiment Station Bulletin 128, pp. 143–185. Brooks, F.E. and Cotton, R.T. (1929) The chestnut Curculios, Curculio proboscideus Fab. and C. auriger Casey. USDA Technical Bulletin 130. Brown, L.G. and Yeager, L.E. (1945) Fox squirrels and gray squirrels in Illinois. Bulletin of Illinois Natural History Survey 23, 449–536. Brown, L.R. (1980) Insects feeding on California oak trees. USDA Forest Service General Technical Report PSW PSW-44, pp. 184–194. Brown, R.C. and Mogensen, H.L. (1972) Late ovule and early embryo development in Quercus gambelii. American Journal of Botany 59, 311–316. Buchholz, J.T. (1941) Multi-seeded acorns. Transactions of the Illinois Academy of Science 34, 99–101. Buech, R.R. (1976) Tree shoot elongation in northern Wisconsin and relationships with temperature and precipitation. Canadian Journal of Forest Research 6, 487–498. Burns, P.Y., Christisen, D.M. and Nichols, J.M. (1954) Acorn production in the Missouri Ozarks. University of Missouri Agriculture Experiment Station Bulletin 611. Burns, R.M. and Honkala, B.H. (1990) Silvics of North America. USDA Forest Service Agriculture Handbook 654, Vol. 2. Cahalane, V.H. (1942) Caching and recovery of food by the western fox squirrel. Journal of Wildlife Management 6, 338–352. Campbell, R.W. (1981) Population dynamics. In: Doane, C.C. and McManus, M.L. (eds) The Gypsy Moth: Research Toward Integrated Pest Management. USDA Forest Service Technical Bulletin 1584, Washington, DC, pp. 65–214. Carlson, W.C. (1974) Root initiation induced by root pruning in northern red oak. OARDC Research Summary. Ohio Agricultural Research and Development Center 74, pp. 14–16. Carmen, W.J., Koenig, W.D. and Mumme, R.L. (1987) Acorn production by five species of oaks over a seven year period at the Hastings Reservation, Carmel Valley, California. USDA Forest Service General Technical Report PSW PSW-100, pp. 429–434. Carpenter, I.W. and Guard, A.T. (1954) Anatomy and morphology of the seedling roots of four species of the genus Quercus. Journal of Forestry 52, 269–274. Cecich, R.A. (1994) The reproductive biology of Quercus with an emphasis on Q. rubra. USDA Forest Service General Technical Report NC NC-173, pp. 11–19. Cecich, R.A. (1997) Influence of weather on pollination and acorn production in two species of Missouri oaks. USDA Forest Service General Technical Report NC NC-188, pp. 252–261.

108

Chapter 2

Cecich, R.A. and Larsen, D.R. (1997) Bud demography and bifurcation in Quercus alba L. and Quercus velutina Lam. Diversity and Adaptation in Oak Species, Proceedings of Working Party 2.08.05, Oct. 12–17, 1997, Genetics of Quercus, of the International Union of Forest Research Organizations (Pennsylvania State University, University Park) 2, pp. 68–75. Cecich, R.A. and Sullivan, N.H. (1999) Influence of weather at time of pollination on acorn production of Quercus alba and Q. velutina. Canadian Journal of Forest Research 29, 1817–1823. Cecich, R.A., Brown, G.L. and Piotter, B.K. (1991) Pistillate flower abortion in three species of oak (abstract). USDA Forest Service General Technical Report NE NE-148, p. 578. Christensen, E.M. (1955) Ecological notes on the mountain brush in Utah. Proceedings of Utah Academy of Science, Arts and Letters 32, 107–111. Christisen, D.M. and Kearby, W.H. (1984) Mast measurement and production in Missouri (with special reference to acorns). Missouri Department of Conservation Terrestrial Series 13. Church, T.W. Jr (1960) Factors affecting the development and survival of sugar maple sprouts. Proceedings of 1960 Society of American Foresters National Convention, pp. 32–35. Clark, F.B. and Liming, F.G. (1953) Sprouting of blackjack oak in the Missouri Ozarks. USDA Forest Service Central States Forest Experiment Station Technical Paper 137. Cobb, S.W., Miller, A.E. and Zahner, R. (1985) Recurrent shoot flushes in scarlet oak stump sprouts. Forest Science 31, 725–730. Cochard, H., Granier, N.B. and Aussenac G. (1992) Vulnerability to air embolism of three European oak species (Quercus petraea (Matt.) Liebl., Q. pubescens Willd., Q. robur L.). Annales des Sciences Forestieres 49, 225–233. Coker, W.C. (1904) Multi-seeded acorns. Botanical Gazette (Chicago) 37, 61–62. Cook, D.B. (1941) The period of growth in some northeastern trees. Journal of Forestry 39, 956–959. Crawley, M.J. and Akhteruzzaman, M. (1988) Individual variation in the phenology of oak trees and its consequences for herbivorous insects. Functional Ecology 2, 409–415. Crow, T.R. (1992) Population dynamics and growth patterns for a cohort of northern red oak (Quercus rubra) seedlings. Oecologia 91, 192–200. Crunkilton, D.D., Pallardy, S.G. and Garrett, H.E. (1992) Water relations and gas exchange of northern red oak seedlings planted in a central Missouri clearcut and shelterwood. Forest Ecology and Management 53, 117–129. Curtis, J.T. (1959) The Vegetation of Wisconsin. University of Wisconsin Press, Madison. Cypert, E. and Webster, B.S. (1948) Yield and use by wildlife of acorns of water and willow oaks. Journal of Wildlife Management 12, 227–231. Dalke, P.D. (1953) Yields of seeds and mast in second growth hardwood forest, south central Missouri. Journal of Wildlife Management 17, 378–381. Darley-Hill, S. and Johnson, W.C. (1981) Acorn dispersal by the blue jay (Cyanocitta cristata). Oecologia 50, 231–232. Davis, F.W., Borchert, M., Harvey, L.E. and Michaelsen, J.C. (1991) Factors affecting seedling survivorship of blue oak (Quercus douglasii H. & A.) in central California. USDA Forest Service General Technical Report PSW PSW-126, pp. 81–86. Deen, R.T. and Hodges, J.D. (1991) Oak regeneration in abandoned fields: presumed role of the blue jay. USDA Forest Service General Technical Report SE SE-70, Vol. 1, pp. 84–93. DeGange, A.R., Fitzpatrick, J.W. and Layne, J.N. (1989) Acorn harvesting by Florida scrub jays. Ecology 70, 348–356. Delong, C.A. and Yahner, R.H. (1991) Impact of small mammals on regeneration of northern red oak. USDA Forest Service General Technical Report NE NE-148, p. 581. Dey, D.C. (1991) A comprehensive Ozark regenerator. PhD dissertation, University of Missouri, Columbia. Dey, D.C. and Parker, W.C. (1996) Regeneration of red oak (Quercus rubra L.) using shelterwood systems: ecophysiology, silviculture and management recommendations. Ontario Forest Research Institute Forest Research Information Paper 126. Dickson, R.E. (1991) Episodic growth and carbon physiology in northern red oak. Proceedings of the Oak Resource in the Upper Midwest Conference. University of Minnesota, pp. 117–124. Downs, A.A. and McQuilkin, W.E. (1944) Seed production of southern Appalachian oaks. Journal of Forestry 42, 913–920. Dwyer, J.P., Cutter, B.E. and Wetteroff, J.J. (1995) A dendrochronological study of black and scarlet oak decline in the Missouri Ozarks. Forest Ecology and Management 75, 60–75.


109

Edwards, M.B. (1990a) Quercus michauxii Nutt. Swamp chestnut oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 693–696. Eyre, F.H. (ed.) (1980) Forest Cover Types of the United States and Canada. Society of American Foresters, Washington, DC. Farmer, R.E., Jr (1975) Dormancy and root regeneration of northern red oak. Canadian Journal of Forest Research 5, 176–185. Farmer, R.E., Jr (1977) Epicotyl dormancy in white and chestnut oaks. Forest Science 23, 329–332. Farmer, R.E., Jr (1979) Dormancy and root growth capacity of white and sawtooth oaks. Forest Science 25, 491–494. Farmer, R.E., Jr (1981) Early growth of black cherry, oaks, and yellow-poplar in southern Appalachian plantings. Tree Planters’ Notes 32(3), 12–14. Feret, P.P., Kreh, R.E., Merkle, S.A. and Oderwald, R.G. (1982) Flower abundance, premature acorn abscission, and acorn production in Quercus alba L. Botanical Gazette (Chicago) 143, 216–218. Fox, J.F. (1982) Adaptation of gray squirrel behavior to autumn germination by white oak acorns. Evolution 36, 800–809. Fralish, J.S., Jones, S.M., O’Dell, R.K. and Chambers, J.L. (1978) The effect of soil moisture on site productivity and forest composition in the Shawnee Hills of southern Illinois. In: Balmer, W.E. (ed.) Proceedings of Soil Moisture–Site Productivity Symposium. USDA Forest Service, State and Private Forestry, Southeastern Area, Asheville, North Carolina, pp. 263–283. Fuchs, M.A., Krannitz, P.G., Harestad, A.S. and Bunnell, F.L. (1997) Seeds that fly on feathered wings: acorn dispersal by Steller’s jays. USDA Forest Service General Technical Report PSW PSW-GTR-160, pp. 648–650. Galford, J.R. (1986) Primary infestation of sprouting chestnut, red, and white oak acorns by Valentinia glandulella (Lepidoptera: Blastobasidae). Entomological News 97, 109–112. Galford, J.R. (1987) Effect of Stelidota octomaculata (Coleoptera: Nitidulidae) on germinating acorns under laboratory conditions. Proceedings of the Central Hardwood Forestry Conference VI. University of Tennessee, Knoxville, pp. 419–422. Galford, J.R. and Weiss-Cottrill, D. (1991) Response of insects to damaged and undamaged germinating acorns. USDA Forest Service Research Paper NE NE-656. Galford, J.R., Peacock, J.W. and Wright, S.L. (1988) Insects and other pests affecting oak regeneration. Society of American Foresters Publication 88-03, pp. 219–225. Galford, J.R., Williams, R.N. and Beacom, M. (1991a) Notes on the biology and hosts of Stelidota ferruginea (Coleoptera: Nitidulidae). USDA Forest Service Research Paper NE NE-654. Galford, J.R., Williams, R.N. and Daugherty, A. (1991b) Life history and notes on the biology of Stelidota octomaculata (Coleoptera: Nitidulidae). USDA Forest Service Research Paper NE NE-644. Galford, J.R., Auchmoody, L.R., Smith, H.C. and Walters, R.S. (1991c) Insects affecting establishment of northern red oak seedlings in central Pennsylvania. USDA Forest Service General Technical Report NE NE-148, pp. 271–280. Galford, J.R., Auchmoody, L.R., Walters, R.S. and Smith, H.C. (1992) Millipede damage to germinating acorns of northern red oak. USDA Forest Service Research Paper NE NE-667. Gardiner, E.S. and Helmig, L.M. (1997) Development of water oak stump sprouts under a partial overstory. New Forests 14, 55–62. Gibson, L.P. (1962) Acorn weevils and associated insects of Ohio: their biology and ecology. MS thesis, Ohio State University, Columbus. Gibson, L.P. (1964) Biology and life history of acorn-infesting weevils of the genus Conotrachelus (Coleoptera: Curculionidae). Annals of the Entomological Society of America 57, 521–526. Gibson, L.P. (1969) Monograph of the genus Curculio in the New World (Coleoptera: Curculionidae), Part I. United States and Canada. Miscellaneous Publication of the Entomological Society of America 6, 239–285. Gibson, L.P. (1971) Insects of bur oak acorns. Annals of the Entomological Society of America 64, 232–234. Gibson, L.P. (1972) Insects that damage white oak acorns. USDA Forest Service Research Paper NE NE-220. Gibson, L.P. (1982) Insects that damage northern red oak acorns. USDA Forest Service Research Paper NE NE-492. Gingrich, S.F. (1967) Measuring and evaluating stocking and stand density in upland hardwood forests in the Central States. Forest Science 13, 38–53.

110

Chapter 2

Godman, R.M. and Mattson, G.A. (1976) Seed crops and regeneration problems of 19 species in northeastern Wisconsin. USDA Forest Service Research Paper NC NC-123. Goodrum, P.D., Reid, V.H. and Boyd, C.E. (1971) Acorn yields, characteristics, and management criteria of oaks for wildlife. Journal of Wildlife Management 35, 520–532. Gottschalk, K.W. (1985) Effects of shading on growth and development of northern red oak, black oak, black cherry, and red maple seedlings. I. Height, diameter, and root/shoot ratio. Proceedings of the 5th Central Hardwood Forest Conference. University of Illinois, pp. 189–195. Gottschalk, K.W. (1990) Gypsy moth effects on mast production. Proceedings of the Workshop on Southern Appalachian Mast Management. USDA Forest Service & University of Tennessee, pp. 42–50. Greulach, V.A. (1973) Plant Function and Structure. Macmillan, New York. Gribko, L.S. and Hix, D.M. (1993) The effect of small rodents on northern red oak acorns in north-central West Virginia. USDA Forest Service General Technical Report NC NC-161, pp. 400–408. Griffin, J.R. (1971) Oak regeneration in the upper Carmel Valley, California. Ecology 552, 862–868. Grimm, E.C. (1984) Fire and other factors controlling the Big Woods vegetation of Minnesota in the mid-nineteenth century. Ecological Monographs 54, 291–311. Grisez, T.J. (1975) Flowering and seed production in seven hardwood species. USDA Forest Service Research Paper NE NE-315. Grubb, P.J. (1977) The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biology Review 52, 107–145. Gullion, G. (1984) Grouse of the North Shore. Willow Creek Press, Oshkosh, Wisconsin. Gysel, L.W. (1957) Acorn production on good, medium, and poor oak sites in southern Michigan. Journal of Forestry 55, 570–574. Haney, G.P. (1962) Growth of thinned white oak sprouts. Journal of Forestry 60, 644–645. Hannon, S.J., Mumme, R.L., Koenig, W.D., Spon, S. and Pitelka, F.A. (1987) Poor acorn crop, dominance, and decline in numbers of acorn woodpeckers. Journal of Animal Ecology 56, 197–207. Hanson, P.J., Dickson, R.E., Isebrands, J.G., Crow, T.R. and Dixon, R.K. (1986) A morphological index of Quercus seedling ontogeny for use in studies of physiology and growth. Tree Physiology 2, 273–281. Harper, J.L. (1977) Population Biology of Plants. Academic Press, London. Harrison, J.S. and Werner, P.A. (1984) Colonization by oak seedlings into a heterogeneous successional habitat. Canadian Journal of Botany 62, 559–563. Head, G.C. (1973) Shedding of roots. In: Kozlowski, T. (ed.) Shedding of Plant Parts. Academic Press, New York, pp. 237–293. Healy, W.M. (1988) Effects of seed-eating birds and mammals on Appalachian hardwood regeneration. Society of American Foresters Publication 88-03, pp. 104–111. Helms, J.A. (ed.) (1998) The Dictionary of Forestry. Society of American Foresters, Bethesda, Maryland. Hildebrand, D.C. and Schroth, M.N. (1967) A new species of Erwinia causing the drippy nut disease of live oaks. Phytopathology 57, 250–253. Holch, A.E. (1931) Development of roots of tree seedlings. Ecology 12, 259–298. Hosner, J.F. (1959) Multi-seeded acorns. Journal of Forestry 57, 127. Janes, D.J., and Nichols, J.M. (1967) Acorn production of nine commercial species of oak in Missouri. School of Forestry, University of Missouri (unpublished manuscript). Jenkins, M.A. and Pallardy, S.G. (1995) The influence of drought on red oak group species growth and mortality in the Missouri Ozarks. Canadian Journal of Forest Research 25, 1119–1127. Johnson, P.S. (1974) Survival and growth of northern red oak seedlings following a prescribed burn. USDA Forest Service Research Note NC NC-177. Johnson, P.S. (1975) Growth and structural development of red oak sprout clumps. Forest Science 21, 413–418. Johnson, P.S. (1977) Predicting oak stump sprouting and sprout development in the Missouri Ozarks. USDA Forest Service Research Paper NC NC-149. Johnson, P.S. (1979a) Shoot elongation of black oak and white oak sprouts. Canadian Journal of Forest Research 9, 489–494. Johnson, P.S. (1979b) Growth potential and field performance of planted oaks. Proceedings of the 1979 J.S. Wright Forestry Conference. Purdue University, pp. 113–119.


111

Johnson, P.S. (1981) Early results of planting English oak in an Ozark clearcut. USDA Forest Service Research Paper NC NC-204. Johnson, P.S. and Rogers, R. (1984) Predicting 25th-year diameters of thinned stump sprouts of northern red oak. Journal of Forestry 82, 616–619. Johnson, P.S., Novinger, S.L. and Mares, W.G. (1984) Root, shoot, and leaf area growth potentials of northern red oak planting stock. Forest Science 30, 1017–1026. Johnson, R.L. (1975) Natural regeneration and development of Nuttall oak and associated species. USDA Forest Service Research Paper SO SO-104. Johnson, W.C. and Adkisson, C.S. (1986) Airlifting the oaks. Natural History 10, 41–43, 48–49. Johnson, W.C. and Webb, T. III. (1989) The role of blue jays (Cyanocitta cristata L.) in the postglacial dispersal of fagaceous trees in eastern North America. Journal of Biogeography 16, 561–571. Johnson, W.C., Adkisson, C.S., Crow, T.R. and Dixon, M.D. (1997) Nut caching by blue jays (Cyanocitta cristata L.): implications for tree demography. American Midland Naturalist 138, 357–370. Johnsson, H. (1952) Ungdomsutvecklingen hos stjälkek, druvek, och rödek. Svenska Slogsvardsföreningens Tidskrift 2, 168–193. Johnston, J.P. (1941) Height-growth periods of oak and pine reproduction in the Missouri Ozarks. Journal of Forestry 39, 67–68. Jones, C.G., Ostfeld, R.S., Richard, M.P., Schauber, E.M. and Wolff, J.O. (1998) Chain reactions linking acorns to gypsy moth outbreaks and Lyme disease risk. Science 279, 1023–1026. Joslin, J.D. and Henderson, G.S. (1987) Organic matter and nutrients associated with fine root turnover in a white oak stand. Forest Science 33, 330–346. Kearby, W.H., Christisen, D.M. and Myers, S.A. (1986) Insects: their biology and impact on acorn crops in Missouri’s upland forests. Missouri Department of Conservation Terrestrial Series 16. Keeley, J.E. (1981) Reproductive cycles and fire regimes. USDA Forest Service General Technical Report WO WO-26, pp. 231–277. Kharitonovich, F.M. (1937) Reforestation by coppice shoots. Sbornik Polezaschitnye Lesnye Polosy (Moscow) (in Russian), pp. 175–239. Kienholz, R. (1941) Seasonal course of height growth in some hardwoods in Connecticut. Ecology 22, 249–258. Kilham, L. (1958) Sealed-in winter stores of red-headed woodpeckers. Wilson Bulletin 70, 107–113. Kilham, L. (1963) Food storing of red-bellied woodpeckers. Wilson Bulletin 75, 227–234. Koenig, W.D. (1980) Acorn storage by acorn woodpeckers in an oak woodland: an energetics analysis. USDA Forest Service General Technical Report. PSW PSW-44. Koenig, W.D. (1990) Oaks, acorns and the acorn woodpecker. Fremontia 18, 77–79. Koenig, W.D., Carmen, W.J., Stanback, M.T. and Mumme, R.L. (1991) Determinants of acorn productivity among five species of oaks in central coastal California. USDA Forest Service General Technical Report PSW PSW-126, pp. 136–142. Koenig, W.D., Mumme, R.L., Carmen, W.J. and Stanback, M.T. (1994) Acorn production by oaks in central coastal California: variation within and among years. Ecology 75, 99–109. Koenig, W.D., Knops, J.M.H., Carmen, W.J., Stanback, M.T. and Mumme, R.L. (1996) Acorn production by oaks in central coastal California: influence of weather at three levels. Canadian Journal of Forest Research 26, 1677–1683. Kolb, T.E. and Steiner, K.C. (1990) Growth and biomass partitioning of northern red oak and yellowpoplar seedlings: effects of shading and grass root competition. Forest Science 36, 34–44. Komarek, E.V. (1970) Insect control–fire for habitat management. Proceedings of the Tall Timbers Fire Ecology Conference 2, pp. 157–171. Korstian, C.F. (1927) Factors controlling germination and early survival in oaks. Yale University School of Forestry Bulletin 19. Kozlowski, T.T. and Pallardy, S.G. (1997) Physiology of Woody Plants, 2nd edn. Academic Press, San Diego. Krajicek, J.E. (1955) Rodents influence red oak regeneration. USDA Forest Service Central States Forest Experiment Station Note, Note 91. Krajicek, J.E. (1960) Some factors affecting oak and black walnut reproduction. Iowa State Journal of Science 34, 631–634. Kramer, P.J. and Decker, J.P. (1944) Relation between light intensity and rate of photosynthesis of loblolly pine and certain hardwoods. Plant Physiology 19, 350–358.

112

Chapter 2

Kramer, P.J. and Kozlowski, T.T. (1979) Physiology of Woody Plants. Academic Press, New York. Kriebel, H.B. (1965) Parental and provenance effects on growth of red oak seedlings. Proceedings, 4th Central States Forest Tree Improvement Conference, pp. 19–25. Kriebel, H.B., Merritt, C. and Stadt, T. (1988) Genetics of growth rate in Quercus rubra: provenance and family effects by the early third decade in the North Central U.S.A. Sylvae Genetica 37, 193–198. Kruger, E.L. and Reich, P.B. (1989) Comparative growth and physiology of stem-pruned and unpruned northern red oak. USDA Forest Service General Technical Report NC NC-132, 302 pp. Kubiske, M.E. and Abrams, M.D. (1992) Photosynthesis, water relations, and leaf morphology of xeric versus mesic Quercus rubra ecotypes in central Pennsylvania in relation to moisture stress. Canadian Journal of Forest Research 22, 1402–1407. Lamson, N.I. (1983) Precommercial thinning increases diameter growth of Appalachian hardwood stump sprouts. Southern Journal of Applied Forestry 7 (May), 93–97. Larson, M.M. (1970) Root regeneration and early growth of red oak seedlings: influence of soil temperature. Forest Science 16, 442–446. Larson, M.M. (1975) Pruning northern red oak nursery seedlings: effects on root regeneration and early growth. Canadian Journal of Forest Research 5, 381–386. Layne, J.N. (1954) The biology of the red squirrel, Tamiasciurus hudsonicus loquax (Bangs), in central New York. Ecological Monographs 24, 227–267. Lee, C.I., Moser, B.C. and Hess, C.E. (1974) Root regeneration of transplanted pin and scarlet oak. New Horizons 1974. Horticultural Research Institute, pp. 10–13. Lewis, V.R. (1989) Host–insect interactions in the California oakworm–coast live oak ecosystem. PhD Dissertation, University of California, Berkeley. Lewis, V.R. (1992) Within-tree distribution of acorns infested by Curculio occidentis (Coleoptera: Curculionidae) and Cydia latiferreana (Lepidoptera: Tortricidae) on the coast live oak. Environmental Entomology 21, 975–982. Liebhold, A., Elkinton, J., Williams, D. and Muzika, R.M. (2000) What causes outbreaks of the gypsy moth in North America? Population Ecology 42, 257–266. Liming, F.G. (1942) Blackjack oak in the Missouri Ozarks. Journal of Forestry 40, 249–252. Liming, F.G. and Johnston, J.P. (1944) Reproduction in oak–hickory forest stands of the Missouri Ozarks. Journal of Forestry 42, 175–180. Loach, K. (1967) Shade tolerance in tree seedlings. 1. Leaf photosynthesis and respiration in plants raised under artificial shade. New Phytologist 66, 607–621. Lockhart, B.R., Hodges, J.D., Toliver, J.R. and Karr, B.L. (1991) Morphology, gas exchange, and carbon-14 allocation patterns in advance cherrybark oak reproduction – preliminary results. USDA Forest Service General Technical Report SE SE-70, Vol. 2, pp. 513–523. Lockhart, B.R., Shelton, M.G. and Guo, Y. (1995). Effects of forest floor amount and composition and acorn placement on emergence and early establishment of white oak. USDA Forest Service General Technical Report SRS SRS-20, pp. 447–453. Longhurst, W.M. (1956) Stump sprouting of oaks in response to seasonal cutting. Journal of Range Management 9, 194–196. Longman, K.A. and Coutts, M.P. (1974) Physiology of the oak tree. In: Morris, M.G. and Perring, F.H. (eds) The British Oak, its History and Natural History. Botanical Society of the British Isles, Classey Ltd, Faringdon, UK, pp. 194–221. Lowell, K.E., Garrett, H.E. and Mitchell, R.J. (1989) Early thinning can improve your stand of coppice-regenerated oak. USDA Forest Service General Technical Report NC NC-132, pp. 53–58. Luckwill, L.C. (1977) The control of growth and fruitfulness of apple trees. In: Luckwill, L.C. and Cutting, C.V. (eds) Physiology of Tree Crops. Academic Press, New York, pp. 237–254. MacRoberts, M.H. and MacRoberts, B.R. (1976) Social organization and behavior of the acorn woodpecker in central coastal California. Ornithological Monographs 21. Marquis, D.A., Eckert, P.L. and Roach, B.A. (1976) Acorn weevils, rodents, and deer all contribute to oak-regeneration difficulties in Pennsylvania. USDA Forest Service Research Paper NE NE-356. Martin, R.E. and Mitchell, R.G. (1981) Possible, potential, probable, and proven fire–insect interactions. Proceedings 1980 Society of American Foresters National Convention, pp. 138–144. Matsuda, K., McBride, J.R. and Kimura, M. (1989) Seedling growth form of oaks. Annals of Botany 64, 439–446. McClaran, M.P. and Bartolome, J.W. (1990) Comparison of actual and predicted blue oak age structures. Journal of Range Management 43, 61–63.


113

McComb, A.L. (1934) The relation between acorn weight and the development of one year chestnut oak seedlings. Journal of Forestry 32, 479–484. McCracken, F.I., Ammon, V., Solomon, J.D. and Nebeker, T.E. (1991) Oak decline in the lower Mississippi River Valley. USDA Forest Service General Technical Report SE SE-70, Vol. 1, pp. 299–306. McDonald, P.M. (1990a) Quercus douglasii Hook. & Arn. Blue oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 631–639. McDonald, P.M. (1990b) Quercus kelloggii Newb. California black oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 661–671. McGee, C.E. (1967) Regeneration in southern Appalachian oak stands. USDA Forest Service Research Note SE SE-72. McGee, C.E. (1968) Northern red oak seedling growth varies by light intensity and seed source. USDA Forest Service Research Note SE, SE-90. McGee, C.E. (1974) Elevation of seed sources and planting sites affects phenology and development of red oak seedlings. Forest Science 20, 160–164. McGee, C.E. (1976) Differences in budbreak between shade-grown and open-grown oak seedlings. Forest Science 22, 484–486. McGee, C.E. (1986) Budbreak for twenty-three upland hardwoods compared under forest canopies and in recent clearcuts. Forest Science 32, 924–935. McGee, C.E. (1988) Spring weather, canopy removal, and early budbreak threaten oak seedlings. Journal of the Elisha Mitchell Science Society 104, 108–115. McGee, C.E. (1997) The effect of oak budbreak patterns on shade tolerance and regeneration. Diversity and Adaptation in Oak Species, Proceedings of Working Party 2.08.05, Oct. 12–17, 1997, Genetics of Quercus, of the International Union of Forest Research Organizations. Pennsylvania State University, University Park, 2, pp. 279–287. McQuilkin, R.A. (1983) Silvical factors affecting the regeneration of oaks and associated species in Pennsylvania. Proceedings Regenerating Hardwood Stands. Pennsylvania State University, pp. 37–65. McQuilkin, R.A. and Musbach, R.A. (1977) Pin oak acorn production on green tree reservoirs in southeastern Missouri. Journal of Wildlife Management 41, 218–225. Merkle, S.A., Feret, P.P., Croxdale, J.G. and Sharik, T.L. (1980) Development of floral primordia in white oak. Forest Science 26, 238–250. Merz, R.W. and Boyce, S.G. (1956) Age of oak ‘seedlings’. Journal of Forestry 54, 774–775. Millington, W.F. and Chaney, W.R. (1973) Shedding of shoots and branches. In: Kozlowski, T.T. (ed.) Shedding of Plant Parts. Academic Press, New York, pp. 149–204. Mogensen, H.L. (1965) A contribution to the anatomical development of the acorn in Quercus L. Iowa State Journal of Science 40, 221–255. Mogensen, H.L. (1975) Ovule abortion in Quercus (Fagaceae). American Journal of Botany 62, 160–165. Mowbray, T.B. and Oosting, H.J. (1968) Vegetation gradients in relation to environment and phenology in a southern Blue Ridge Gorge. Ecological Monographs 38, 309–344. Muick, P.C. and Bartolome, J.W. (1987) Factors associated with oak regeneration in California. USDA Forest Service General Technical Report PSW PSW-100, pp. 86–91. Muller, C.H. (1951) The significance of vegetative reproduction in Quercus. Madroño 11, 129–137. Murtfeldt, M.E. (1894) Acorn insects, primary and secondary. Insect Life (USDA Division of Entomology) 6, 318–324. Musselman, R.C. and Gatherum, G.E. (1969) Effects of light and moisture on red oak seedlings. Iowa State Journal of Science 43, 273–284. Myers, S.A. (1978) Insect impact on acorn production in Missouri upland forests. PhD dissertation, University of Missouri, Columbia. Nakashizuka, T., Takahashi, Y. and Kawaguchi, H. (1997) Production-dependent reproductive allocation of a tall tree species Quercus serrata. Journal of Plant Research 110, 7–13. Neilson, R.P. and Wullstein, L.H. (1980) Catkin freezing and acorn production in gambel oak in Utah, 1978. American Journal of Botany 67, 426–428. Nichols, J.M. (1954) Direct seeding of oak in Missouri. University of Missouri Agriculture Experiment Station Bulletin 609. Nichols, J.O. (1968) Oak mortality in Pennsylvania. Journal of Forestry 66, 681–694.

114

Chapter 2

Olson, D.F., Jr (1974) Quercus L. Oak. USDA Forest Service Agriculture Handbook 450, 88 pp. Ostfeld, R.S., Jones, C.G. and Wolff, J.O. (1996) Of mice and mast. BioScience 46, 323–330. Pallardy, S.G. and Rhoads, J.L. (1993) Morphological adaptations to drought in seedlings of deciduous angiosperms. Canadian Journal of Forest Research 23, 1766–1774. Parker, J. (1949) Effects of variations in the root–leaf ratio on transpiration rate. Plant Physiology 24, 739–743. Parker, J. and Patton, R.L. (1975) Effects of drought and defoliation on some metabolites in roots of black oak seedlings. Canadian Journal of Forest Research 5, 457–463. Parker, W.C., Pallardy, S.G., Hinckley, T.M. and Teskey, R.O. (1982) Seasonal changes in tissue water relations of three woody species of the Quercus–Carya forest type. Ecology 63, 1259–1267. Pavlik, B.M., Muick, P.C., Johnson, S. and Popper, M. (1991) Oaks of California. Cachuma Press, Los Olivos, California. Petrides, G.A. (1988) A Field Guide to Eastern Trees. Houghton Mifflin, Boston, Massachusetts. Pfannmuller, L.A. (1991) Significance of oaks and oak forest communities for nongame wildlife. Proceedings of The Oak Resource in the Upper Midwest Conference. University of Minnesota, pp. 56–64. Phares, R.E. (1971) Growth of red oak (Quercus rubra L.) seedlings in relation to light and nutrients. Ecology 52, 669–672. Pierce, W.D. (1908) A list of parasites known to attack American Rhynchophora. Journal of Economic Entomology 1, 380–396. Post, L.S. (1998) Seed management in Tennessee: development of seed zones for Tennessee and distribution and protection of northern red oak (Quercus rubra L.) acorns. MS thesis, University of Tennessee, Knoxville, Tennessee. Powell, D.S. (1976) Sprouting ability of advance reproduction of undisturbed forest stands in West Virginia. MS thesis, University of West Virginia, Morgantown. Racine, C.H. (1971) Reproduction of three species of oak in relation to vegetational and environmental gradients in the southern Blue Ridge. Bulletin of the Torrey Botanical Club 98, 297–310. Reich, R.B., Teskey, R.O., Johnson, P.S. and Hinckley, T.M. (1980) Periodic root and shoot growth in oak. Forest Science 26, 590–598. Rice, C. and Struve, D.K. (1997) Seedling growth form and water use of selected oak species. Diversity and Adaptation in Oak Species, Proceedings of Working Party 2.08.05, Oct. 12–17, 1997, Genetics of Quercus, of the International Union of Forest Research Organizations. Pennsylvania State University, University Park, 2, pp. 269–278. Roach, B.A. and Gingrich, S.F. (1967) Upland hardwoods can be grown efficiently. Pulp and Paper (April). Romberger, J.A. (1963) Meristems, growth and development in woody plants. USDA Forest Service Technical Bulletin 1293. Sander, I.L. (1971) Height growth of new oak sprouts depends on size of advance reproduction. Journal of Forestry 69, 809–811. Sander, I.L. (1979) Regenerating oaks. Proceedings of the National Silviculture Workshop (USDA Forest Service), pp. 212–221. Scholz, H.F. (1955) Effect of scarification on the initial establishment of northern red oak reproduction. USDA Forest Service Lake States Forest Experiment Station Technical Note 425. Schwartz, C.W. and Schwartz, E.R. (1959) The Wild Mammals of Missouri. University of Missouri Press, Columbia. Scott, V.E., Evans, K.E., Patton, D.R. and Stone, C.P. (1977) Cavity-nesting birds of North American Forests. USDA Forest Service Agriculture Handbook 511. Seidel, K.W. (1972) Drought resistance and internal water balance of oak seedlings. Forest Science 18, 34–40. Shafer, J.D. (1971) Shade tolerance of scarlet oak seedlings. MS thesis, University of Missouri, Columbia. Sharp, W.M. (1958) Evaluating mast yields in the oaks. Pennsylvania State University Agriculture Experiment Station Bulletin 635. Sharp, W.M. and Chisman, H.H. (1961) Flowering and fruiting in the white oaks. I. Staminate flowering through pollen dispersal. Ecology 42, 365–372. Sharp, W.M. and Sprague, V.G. (1967) Flowering and fruiting in the white oaks. Pistillate flowering, acorn development, weather, and yields. Ecology 48, 243–251.


115

Shaw, M.W. (1968) Factors affecting the natural regeneration of sessile oak (Quercus petraea) in North Wales. II. Acorn losses and germination under field conditions. Journal of Ecology 647–660. Short, H.L. (1976) Composition and squirrel use of acorns of black and white oak groups. Journal of Wildlife Management 40, 479–483. Sidney, C. (1948) Acorn weevils of the North Carolina Piedmont: Their biology and method of sampling infestation. MS thesis, Duke University, Durham, North Carolina. Smith, C.C. and Follmer, D. (1972) Food preferences of squirrels. Ecology 53, 82–91. Smith, K.G. (1986) Winter population dynamics of blue jays, red-headed woodpeckers, and northern mockingbirds in the Ozarks. American Midland Naturalist 115, 52–62. Smith, R.L. (1962) Acorn consumption by white-footed mice. West Virginia University Agricultural Experiment Station Bulletin 482T. Sork, V.L. and Bramble, J.E. (1993) Prediction of acorn crops in three species of North American oaks: Quercus alba, Q. rubra and Q. velutina. Annales des Sciences of Forestieres 50 (suppl. 1), 128s–136s. Sork, V.L., Bramble, J. and Sexton, O. (1993) Ecology of mast-fruiting in three species of North American deciduous oaks. Ecology 74, 528–541. Sperry, J.S. and Sullivan, J.E.M. (1992) Xylem embolism in response to freeze–thaw cycles and water stress in ring-porous, diffuse-porous, and conifer species. Plant Physiology 100, 605–613. Stairs, G.R. (1964) Microsporogenesis and embryogenesis in Quercus. Botanical Gazette (Chicago) 125, 115–121. Staley, J.M. (1965) Decline and mortality of red and scarlet oaks. Forest Science 11, 2–17. Stapanian, M.A. and Smith, C.C. (1978) A model for seed scatterhoarding: coevolution of fox squirrels and black walnuts. Ecology 59, 884–896. Stein, W.I. (1990) Quercus garryana Dougl. ex Hook. Oregon white oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 650–660. Steiner, K.C, Abrams, M.D. and Bowersox, T.W. (1993) Advance reproduction and other stand characteristics in Pennsylvania and French stands of northern red oak. USDA Forest Service General Technical Report NC NC-161, pp. 473–483. Tecklin, J. and McCreary, D.D. (1991). Acorn size as a factor in early seedling growth of blue oaks. USDA Forest Service General Technical Report PSW PSW-126, pp. 48–53. Teskey, R.O. (1978) Influence of temperature and moisture on root growth of white oak. MS thesis, University of Missouri, Columbia. Thornburgh, D.A. (1990) Quercus chrysolepis Liebm. Canyon live oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 618–624. Tiedemann, A.R., Clary, W.P. and Barbour, R.J. (1987) Underground systems of Gambel oak (Quercus gambelii) in central Utah. American Journal of Botany 74, 1065–1071. Tripathi, R.S. and Khan, M.L. (1990) Effects of seed weight and microsite characteristics on germination and seedling fitness in two species of Quercus in a subtropical wet hill forest. Oikos 57, 289–296. Tryon, E.H. and Carvell, K.L. (1958) Regeneration under oak stands. West Virginia University Agricultural Experiment Station Bulletin 424T. Tryon, E.H. and Carvell, K.L. (1962) Acorn production and damage. West Virginia University Agricultural Experiment Station Bulletin 466T. Tucker, J.M. (1980) Taxonomy of California oaks. USDA Forest Service General Technical Report PSW PSW-44, pp. 19–29. Turkel, H.S., Rebuck, A.L. and Grove, A.R., Jr (1955) Floral morphology of white oak. Pennsylvania State University Agriculture Experiment Station Bulletin 593. Tyree, M.T. (1989) Cavitation in trees and the hydraulic sufficiency of woody stems. Annales des Sciences Forestieres (suppl.) 46, 330s–337s. Tyree, M.T. and Cochard, H. (1996) Summer and winter embolism in oak: impact on water relations. Annales des Sciences Forestieres 53, 173–180. USDA Forest Service (1985) Insects of eastern forests. Miscellaneous Publication 1426. Van Dersal, W.R. (1940) Utilization of oaks by birds and mammals. Journal of Wildlife Management 4, 404–428. Vivin, P., Aussenac, G. and Levy, G. (1993) Differences in drought resistance among 3 deciduous oak species grown in large boxes. Annales des Sciences Forestieres 50, 221–233. Vogt, A.R. and Cox, G.S. (1970) Evidence for the hormonal control of stump sprouting by oak. Forest Science 16, 165–171.

116

Chapter 2

Wareing, P.F. (1951) Growth studies in woody species IV. The initiation of cambial activity in ringporous species. Physiologia Plantarum. 4, 546–562. Wargo, P.M. and Haack, R.A. (1991) Understanding the physiology of dieback and decline diseases and its management implications for oak. Proceedings of The Oak Resource in the Upper Midwest Conference, University of Minnesota, pp. 147–158. Weigel, D.R. and Johnson, P.S. (1998) Planting northern red oak in the Ozark Highlands: A shelterwood prescription. USDA Forest Service Technical Brief TB-NC TB-NC-6. Weld, L.H. (1922) Note on American gallflies of the family Cynipidae producing galls on acorns with description of new species. Proceedings of the US National Museum 61, pp. 1–32. Welker, J.M. and Menke, J.W. (1990) The influence of simulated browsing on tissue water relations, growth and survival of Quercus douglasii (Hook and Arn.) seedlings under slow and rapid rates of soil drought. Functional Ecology 4, 807–817. Wendel, G.W. (1975) Stump sprout growth and quality of several Appalachian hardwood species after clearcutting. USDA Forest Service Research Paper NE NE-329. Williamson, M.J. (1966) Premature abscissions and white oak acorn crops. Forest Science 12, 19–21. Wilson, B.F. (1968) Red maple stump sprout development the first year. Harvard Forest Paper 18. Wilson, B.F. (1993) Compensatory shoot growth in young black birch and red maple trees. Canadian Journal of Forest Research 23, 302–306. Winston, P.W. (1956) The acorn microsere, with special reference to arthropods. Ecology 37, 120–132. Wolgast, L.J. (1972) Mast production in scrub oak (Quercus ilicifolia) on the coastal plain in New Jersey. PhD dissertation, Rutgers University, Princeton, New Jersey. Wolgast, L.J. and Trout, J.R. (1979) Late spring frost affects yields of bear oak acorns. Journal of Wildlife Management 43, 239–240. Wood, O.M. (1934) A brief record of seed productivity for chestnut oak in southern New Jersey. Journal of Forestry 32, 1014–1016. Wood, O.M. (1938) Seedling reproduction of oak in southern New Jersey. Ecology 19, 276–293. Wright, S.L. (1987) Managing insects affecting oak regeneration by prescribed burning. USDA Forest Service General Technical Report SE SE-46, pp. 186–192. Wright, S.L., Hall, R.W. and Peacock, J.W. (1989) Effects of simulated insect damage on growth and survival of northern red oak (Quercus rubra L.) seedlings. Environmental Entomology 18, 235–239. Wuenscher, J.E. and Kozlowski, T.T. (1971) Relationship of gas-exchange resistance to tree-seedling ecology. Ecology 52, 1016–1023. Yin, X., Perry, J.A. and Dixon, R.K. (1989) Fine-root dynamics and biomass distribution in a Quercus ecosystem following harvest. Forest Ecology and Management 27, 159–177. Zasada, J.C. and Zahner, R. (1969) Vessel element development in the earlywood of red oak (Quercus rubra). Canadian Journal of Botany 47, 1965–1971. Zimmermann, M.H. (1983) Xylem Structure and the Ascent of Sap. Springer-Verlag, New York.

3 Regeneration Ecology II: Population Dynamics

Introduction This chapter is about the establishment and development of populations of juvenile oaks. Variability is a normal characteristic of tree populations, and it can be described in relation to specific tree attributes. For example, an oak forest can be described by the size or age distributions of its member trees, and how those distributions vary in time and space. Populations of one tree species also interact with other species, each with unique ecological requirements and competitive advantages and disadvantages that lead to variation in patterns of establishment, growth and survival. Population variability is further increased by forest disturbances. Predicting the responses of tree populations to forest disturbances, whether natural or of human origin, is fundamental to the practice of silviculture. The fitness or suitability of adult trees to the physical characteristics of an ecosystem (i.e. the site) are often cited as explanations for the observed distribution of species across environmentally heterogeneous landscapes. However, relatively little emphasis has similarly been accorded to species’ regeneration requirements. The concept of the regeneration niche attempts to fill that void. In contrast to the more general notion of species’ niche (Chapter 1), regeneration niche refers specifically to regeneration events and associated ecological conditions. Regeneration niche therefore considers

both the time and place where there is a high probability that a mature tree will be replaced by another tree of the same species (Grubb, 1977). Regeneration niches are ephemeral; their abundance and locations vary over time. The resource requirements of seedlings and seedlings sprouts (e.g. light, nutrients, heat and soil moisture) differ from those of mature trees. Therefore, it has been proposed that niche differences among species coexisting in the same ecosystem may be best expressed during the vulnerable early stages of plant establishment (Grubb, 1977; Latham, 1992; Veblen, 1992). Because all autotrophic plants require essentially the same kinds of resources (but not necessarily the same amounts), niche differentiation among species is likely to be strongly expressed during a life history period when one or more of those resources are limiting. Different species, even among the oaks, adapt to gradients in resource availability and competition in different ways, including how they allocate their growth to stems, roots and leaves (Loach, 1967; Gottschalk, 1985, 1987; Matsuda and McBride, 1986; Matsuda et al., 1989; Kolb and Steiner, 1990; Latham, 1992; Pallardy and Rhoads, 1993; Walters et al., 1993; Callaway and Mahall, 1996; Rice and Struve, 1997). Seedling populations may stratify by species along resource gradients that occur at very small spatial scales. At a spatial scale that corresponds to the area occupied by an adult tree, the distribution 117

118

Chapter 3

of resources may be highly variable (patchy). But at a smaller spatial scale corresponding to the area occupied by a seedling, the same resources may appear very uniform and either favour or inhibit seedling development (Latham, 1992). This variability in the spatial distribution of regeneration niches has been offered as one explanation for the maintenance of diversity in ecosystems (Grubb, 1977; Latham, 1992), and the coexistence of multiple tree species (Veblen, 1992). For the oaks, the critical regeneration events include flowering, fruiting, seed dispersal and germination, seedling establishment, dieback and sprouting, and the growth of oak reproduction. Moreover, numerous biotic and abiotic factors affect the regeneration niche. An oak’s regeneration niche therefore may not pertain to just one ecological factor, but rather to a suite of factors that change over time and result in a corresponding range of regeneration success probabilities. Over time, it is the coupling of regeneration events with ecological conditions that determines probabilities of regeneration success for oaks.

Regeneration Strategy A species’ regeneration niche, i.e. when and where its regeneration requirements are met, is partially expressed by the mechanisms that have evolved to facilitate its regeneration. However, an advantage associated with a particular mechanism in one environment may be a disadvantage in another environment. In a sense, regeneration is one of the problems that a species solves during its evolution. The resulting solution constitutes the evolutionary compromise that defines a species’ regeneration strategy.1 Strategies pertaining to oaks can be considered in relation to reproductive mechanisms, the accumulation and fluctuation of populations of reproduction, and responses to site and forest disturbance. 1This

Reproductive mechanisms: seeding and sprouting Seed production and sprouting are two reproductive mechanisms, or tactics, employed by oaks in their regeneration strategy. Although all oaks rely to some extent on both seeding and sprouting, they differ in their dependence on one mechanism versus the other. Even within a species, regeneration tactics may vary among habitats and disturbance regimes. The large number of oak species, their wide distribution and the disparate habitats they occupy require that regeneration strategies among the oaks vary substantially. At one extreme, oaks of the arid southwestern United States may regenerate almost exclusively by sprouting. Among those species, some reproduce largely from root-like lignotubers and rhizomes (Tiedemann et al., 1987). One such species is Gambel oak, whose seedlings are often only a minor source of reproduction (Harper et al., 1985). Hinckley oak, a shrubby rhizomatous oak restricted to a natural range of a fraction of an acre in western Texas, is known to regenerate only from rhizomes (Muller, 1951). Similarly, the shrub oaks of the fire-prone southern California chaparral depend heavily on their ability to sprout after burning. But even when regeneration is largely dependent on sprouting, some new seedlings eventually must be produced to replace the trees and rootstocks that are inevitably lost to mortality, if a species is to persist. In the Ozark Highlands of southern Missouri, oak regeneration also largely depends on sprouting (Liming and Johnston, 1944; Johnson, 1979; Dey et al., 1996a). These forests occur on relatively droughty sites and are often dominated by some combination of black, white, scarlet, post, southern red and blackjack oaks. Under mature closed-canopy stands, the total density of oak reproduction often varies from 1000 to 3000 seedlings and seedling sprouts per acre. After a good

and related anthropomorphisms are used for conceptual convenience and should not be interpreted as implying that trees ‘plan’ their evolution or exercise choices in the sense that humans do.

Regeneration Ecology II: Population Dynamics

acorn crop, 150–300 new oak seedlings per acre may become established (Sander, 1979a). But the continued dominance of the oaks largely depends on a relatively small number of oak seedling sprouts (200–400 per acre) with large root systems capable of supporting rapid shoot growth after disturbances that release growing space formerly occupied by trees in the overstorey (Johnson, 1979; Sander et al., 1984). Except on the most mesic habitats within the Ozark Highlands, non-oaks seldom successionally displace oaks. Throughout much of the eastern United States, northern red oak occupies the mesic middle ground between wet and dry extremes in soil moisture. Northern red oak expresses its relatively flexible regeneration strategy by producing thousands of new seedlings per acre following bumper acorn crops (Johnson, 1974). In northern red oak forests in France, the number of oak seedlings and seedling sprouts growing beneath the forest canopy may exceed 100,000 per acre (Steiner et al., 1993). Northern red oak also has flexible seedbed requirements (Crow, 1992), potentially rapid shoot growth (Farmer, 1975), the ability to regenerate from seedlings established after final harvest (Johnson et al., 1989), moderate shade tolerance (Sander, 1990), and the capacity to sprout from large stumps of overstorey trees (P.S. Johnson, 1975; Wendel, 1975; Weigel and Johnson, 1998). But unlike the more xerophytic oaks of the Missouri Ozarks, the relatively mesophytic red oak frequently fails to regenerate because of its vulnerability to successional displacement by the more shade tolerant species with which it typically co-occurs (Johnson, 1976; Crow, 1988; Loftis, 1990a; Nowacki et al., 1990; Lorimer, 1993). In bottomland forests of southern United States, seeding appears to be a more important regeneration tactic than in other oak-dominated ecosystems in North America. For example, water oak seedlings established after final harvest can become important members of the succeeding stand (Golden and Loewenstein, 1991; Loewenstein, 1992; Loewenstein and Golden, 1995). But the bottomland oaks

119

also can sprout prolifically, and this facilitates their development in stands disturbed by fire and grazing (Aust et al., 1985). Being neither obligate seeders nor sprouters, these oaks have very flexible regeneration strategies. The relative importance of seeding as a species’ regeneration tactic is sometimes revealed by the number of seedlings that become established after a heavy acorn crop. For example, more than 100,000 Nuttall oaks per acre became established after a bumper acorn crop in one bottomland forest (R.L. Johnson, 1975). If only 0.1% of those seedlings (100 per acre) were competitively successful and well distributed, they would occupy most of the available growing space within two decades following a disturbance that removed the overstorey. Nuttall oak’s seeding strategy is complemented by its rapid height growth (Johnson, 1981), which is necessary for it to successfully compete with the fast-growing and persistent competitors occurring on bottomlands. A species’ seeding strategy also may be reinforced by the dispersal of acorns by rodents and jays to habitats favourable to seed protection, germination, and seedling growth (McQuilkin, 1983; Harrison and Werner, 1984; Sork, 1984; Johnson and Webb, 1989). Nevertheless oaks often fail to regenerate in southern bottomland forests. Competing green ash, sweetgum, and other tree reproduction may outgrow and suppress oak reproduction shortly after standinitiating disturbances occur (Johnson and Krinard, 1983; Aust et al., 1985). Moreover, periods of high oak reproduction density in bottomland forests are often followed by periods of low seedling density because of low seedling survival rates and infrequent acorn crops (R.L. Johnson, 1975). Prolonged periods with little or no oak advance reproduction thus frequently occur. Consequently, domination of bottomland forests by oaks is often limited to one generation. Bottomland oaks nevertheless possess characteristics that, under certain conditions, favour their regeneration over associated species. For example, water oak can

120

Chapter 3

persist in floodplain forests because, unlike its non-oak competitors, its acorns often germinate after flooding occurs. Oak seedlings also avert flood-caused uprooting because of their long taproots and ability to resprout after their tops are damaged (Streng et al., 1989). Because of this persistence, oak reproduction is often relatively well represented in the older and larger size classes of the total reproduction complex. In an East Texas floodplain forest, water oak seedlings ranged from 5% of all 1-year-old seedlings to 32% of all 5-yearold seedlings (Fig. 3.1). Like oak reproduction in uplands, the older and larger reproduction in bottomlands has the highest probability of capturing growing space after a canopy disturbance. These anomalies in regeneration strategy among the oaks emphasize the difficulty of generalizing their regeneration ecology across species and habitats. Different oaks have solved their regeneration problems in different ways. Some species are more flexible than others (Fig. 3.2). In turn, each species’ environment, physiology and genetics shapes its flexibility in regeneration tactic.

100

2,574

1,218

698

Successful regeneration of oaks partly depends on the species that oaks must compete with. The regeneration strategies of competing species may differ greatly from the oaks. As a result, non-oaks may be competitively advantaged or disadvantaged, depending on their adaptations to the particular environment. In addition to basal sprouting and producing seedlings from current seed, non-oaks also may regenerate from root sprouts (e.g. aspens, beech, sassafras and sweetgum) and from seed stored in the forest floor (e.g. black cherry, white and green ashes, and yellow-poplar). The seeds of some species such as pin cherry and briars, unlike the oaks, remain buried in the forest floor and soil where they can accumulate for decades (Marks, 1974; Whitney, 1986). Seeds of other prolific seed producers, such as yellow-poplar, black cherry and sassafras, are shorter-lived but nevertheless may be present in high numbers (Wendel, 1977). When environmental conditions are favourable, such as after clearcutting or burning, stored seeds of non-oaks may germinate in enormous numbers and produce dense populations that often outgrow co-occurring oaks. 267

110 Water oak Sweetgum Ironwood Red maple American elm Other

Seedlings (%)

80

60

40

20

0 1

2

3 4 Seedling age (years)

5

Fig. 3.1. Species composition of reproduction under an east Texas floodplain forest in relation to seedling age. Numbers of trees in each age class are shown above bars. Other species include blackgum, deciduous holly, American holly and Sebastian bush. (Adapted from Streng et al., 1989, with permission.)


121

100

Relative dependence on seeding

Nuttall oak 80 Northern red oak 60 White oak 40

Black oak

20 Post oak Gambel oak 0 0

20

40

60

80

100

Relative dependence on sprouting Fig. 3.2. Conceptualized dependence of six oak species on seeding versus sprouting as a regeneration tactic. Diameters of circles are proportionate to each species’ postulated flexibility in selecting the alternative tactic.

After heavy thinning or complete overstorey removal, competition from non-oak root sprouts and stump sprouts may suppress and kill established oak reproduction (Beck, 1970; Beck and Hooper, 1986). Grasses, sedges, ferns, vines and shrubs also can seriously interfere with the regeneration of oaks (Jarvis, 1964; Beck, 1970; Hanson and Dixon, 1985; Bowersox and McCormick, 1987; Horsley, 1991; Smith and Vankat, 1991; Johnson 1992). In the oak savannas and woodlands of Oregon and California, competition from exotic annual grasses that have replaced native perennial bunchgrasses is a probable cause of oak regeneration failures (Gordon et al., 1989; Welker and Menke, 1990; Barnhart et al., 1991; Danielsen and Halvorson, 1991; Pavlik et al., 1991; Riegel et al., 1992). Whether or not oaks regenerate successfully thus depends in no small part on the competition environment.

Accumulation of oak reproduction In many oak-dominated ecosystems, the oak reproduction beneath the parent

stand includes seedlings and seedling sprouts that originate from several acorn crops. This accumulation of reproduction results from the combined effects of periodic seed production, the relatively large food reserves in acorns that sustain seedlings through the first year, the high sprouting capacity of seedlings, drought tolerance and the ability of seedlings to persist under at least moderate shade. A small proportion of the seedlings originating from a single acorn crop (cohort) may survive for several decades. Silviculturists sometimes refer to the resulting accumulation of seedlings and seedling sprouts as advance reproduction because, in the management of even-aged forests, it is present in advance of final harvest. Its presence and development largely determines the importance of oak after the occurrence of natural or human events that greatly reduce parent stand density. The capacity of oak reproduction to accumulate over several acorn crops may compensate, in part, for the inability of acorns to remain viable for more than a few months during a single dormant season.

122

Chapter 3

The term accumulation, as used here, is not meant to imply a continuous, unending increase in oak reproduction density with time. Rather, it refers to the episodic addition of new seedlings to one or more established cohorts. The total population of seedlings and seedling sprouts beneath an oak stand responds to the prevailing overstorey condition, which is continually changing and affecting established oak reproduction and other understorey vegetation. However, the changes in oak reproduction occur much faster than in the parent stand. Consequently, relatively short cycles of the birth, growth and death of individual cohorts of oak reproduction are embedded in longer-term changes in overstorey species composition and size structure. An accumulated population of oak reproduction can sometimes capture all or some of the growing space in canopy gaps or larger openings. Its success in doing so usually depends on the presence of seedling sprouts with relatively large roots. True seedlings and small seedling sprouts usually have root systems that are too small to support rapid shoot growth (Sander, 1971). And stump sprouts from overstorey trees, by themselves, may not be numerous enough to capture all of the available growing space (Sander et al., 1984). Nevertheless, it is the number, size and spatial distribution of all three classes of reproduction that express the total oak regeneration potential of a stand (Sander et al., 1984). Because this potential is implicit in the advance reproduction and the parent stand itself, the new stand is essentially a ‘legacy’ (sensu Franklin et al., 1989) of the parent stand. Sustaining oak-dominated forests thus largely depends on perpetuating propagules from one generation to the next. Whether or not the overstorey is eliminated in small gaps or over large areas, the resulting spatial units of reproduction are even-aged.2 Although oaks have not always been regarded as species well adapted to 2However,

capturing small canopy gaps (Ehrenfeld, 1980; Crow, 1988), they have the potential to do so if advance reproduction of sufficient size is present. For example, 10-ft tall advance reproduction of northern red oak captured canopy gaps as small as 1/25 acre in a mixed sugar maple–oak stand in southwestern Wisconsin (Lorimer, 1983). This is the approximate area occupied by the crowns of five dominant or co-dominant northern red oaks 15 inches dbh in a fully stocked stand. The accumulation of oak advance reproduction is most obvious in xeric uplands where recurrent dieback and resprouting of reproduction produce multiple stems of varying size, some of which are attached to large root systems. Large roots may be several decades older than their living stems. The true age of a seedling sprout thus is recorded in the rings of its taproot, and not its stem, which is set back to age zero with each complete dieback.3 In a southeastern Ohio oak–hickory stand, the roots of oak reproduction were up to 32 years older than the stems (Merz and Boyce, 1956). In West Virginia, root ages of northern red and white oaks were up to 8 and 10 years older, respectively, than their stems (Fig. 3.3). Accordingly, the range of ages of the roots of advance reproduction and the number per acre of older living roots is an indicator of an oak stand’s capacity to accumulate oak reproduction.

Variation in the accumulation process In the oak forests of the eastern United States, the accumulation of oak reproduction generally increases with decreasing site quality and overstorey density. For oak forests in West Virginia, the relation is represented by the decreasing average age of root systems of oak reproduction (age measured at the root collar) with increasing overstorey density and site index (Fig. 3.4A). Accumulation of oaks tends to be

if their age structure were based on age from germination (‘age from birth’), they could be uneven-aged, depending on the length of the reproduction accumulation period. 3Roots of oaks are difficult to age because their annual rings are not easily distinguishable even through a light microscope. Accurate age determination therefore usually requires special techniques such as Xradiography (Renton et al., 1974; Powell, 1976).


A

20

0.8 0.6

16

0.4

12 8

Mean root age (years)

24

Root age (years)

40

Root diameter (inches)

30

20

Site index

A

35 50 65

80 10

0.2 4

True seedlings (1:1)

0 60

0 0 24

4

8

12

16

70 80 Crown cover (%)

90

B 0.8 0.6

16 12 8 0.4 4 0.2

True seedlings (1:1)

0.8 Root diameter (inches)

20 Root age (years)

123

Site index

B

35 0.6 50 0.4 65

0 0

4

8 12 Stem age (years)

80

16

Fig. 3.3. Relation of root age to stem age and root diameter for (A) white oak, and (B) northern red oak reproduction in a random sample of 56 stems of each species beneath West Virginia oak stands. Based on the linear regression models: (A) Root age = 1.97 + 0.671*SA + 10.0*RD; R 2 = 0.86. (B) Root age = 0.199 + 0.735*SA + 8.98*RD; R 2 = 0.90. In both equations, SA = stem age and RD = root diameter. Root diameters were measured just below the root collar. Estimates are shown for the approximate range of observed values of stem age and root diameter. (Adapted from Powell, 1976, used with permission.)

greatest on the drier sites and where overstorey density is low. Root diameter is positively correlated with root age (when both are measured at the root collar) so the accumulation process is similarly evident in the relation of reproduction basal diameter to site quality and overstorey density (Fig. 3.4B). However, oaks in arid and semi-arid regions behave differently. Because of

0.2

60

70 80 Crown cover (%)

90

Fig. 3.4. Mean age (A) and diameter (B) of roots of oak reproduction in relation to overstorey crown cover and site index in West Virginia forests. Ages and diameters were measured just below the root collar on a random sample of 46 plots. Estimates are based on linear regression models: (A) Root age = 63.8 – 0.391*CC – 0.205*SI – 0.241*SLP; R 2 = 0.74. (B) Root diameter = 1.99 – 0.0172*CC–0.0185*SI + 0.000185*CC*SI – 0.00285*SLP; R 2 = 0.80. For both models, CC = crown cover (%), SI = oak site index (ft at an index age of 50 years), SLP = slope percentage. For this graph slope percentage is held constant at 25, and each data point represents seven to ten trees. Estimates are shown for the approximate range of observed values of crown cover and site index. The estimates represent the average of the sample of scarlet, chestnut, black, white and northern red oak reproduction. (Adapted from Matney, 1974, used with permission.)

124

Chapter 3

extreme heat, solar radiation and moisture deficiency, establishment and survival of oak reproduction is often favoured on the less severe sites such as north-facing slopes and under partial shade rather than in canopy gaps (Callaway and D’Antonio, 1991; Callaway et al., 1991; Williams et al., 1991; Callaway, 1992; Keeley, 1992). Wherever the accumulation of oak reproduction occurs, it results from survival rates sufficient for reproduction to build up over several successive acorn crops. Because the size and age of oak advance reproduction are positively correlated, properties of the accumulation process are sometimes implicit in the size distribution of the reproduction. The regeneration dynamics of black oak–white oak/Vaccinium forests in northern Lower Michigan (Host and Pregitzer, 1991) (Fig. 1.8) provide an example. These forests occur on droughty outwash sands where the site index for black oak ranges from about 50 to 60 ft at an index age of 50 years (Cleland et al., 1993). The density of oak reproduction averages about 11,000 seedlings or seedling sprouts per acre (Johnson, 1992). However, the frequency of occurrence of black oak and white oak reproduction in successively larger basal diameter classes decreases at an exponential rate (Fig. 3.5). This relation suggests that only a fraction of the reproduction in each size class survives to become a member of the next larger size class. The recruitment or ingrowth of oak reproduction into successively larger size classes involves three population processes that jointly determine the accumulation rate: (i) the periodic establishment of new seedlings (seedling input); (ii) growth rate; and (iii) survival rate. If, in a given ecosystem, all three processes are in equilibrium so that seedling input balances seedling mortality and seedling growth remains relatively constant, then the associated size and age distribution of survivors should remain constant. However, constancy is an unlikely attribute of an oak forest. Variable weather, fluctuating acorn production, fluctuating populations of acorn consumers, changes

in stand composition and structure that affect acorn production, seedling growth and seedling survival are some of the factors that rule against constancy. The diameter distribution of oak reproduction at one point in time nevertheless reflects, in some way, the recent history of seedling input, drain and growth in a given stand. Population curves like those in Fig. 3.5 therefore might be interpreted as representing a transient recruitment rate of oak reproduction into successively larger size classes under recently prevailing conditions. Such rates accordingly would reflect the nature of the accumulation process. A transient recruitment rate can be estimated by fitting certain mathematical functions to observed field data. The negative exponential function has been widely used to express the constant exponential decrease in numbers of individuals within a cohort per unit time. For the black oak curve in Fig. 3.5, the slope coefficient of the function (3.1) defines the rate of change in oak seedling and seedling sprout density in relation to increasing basal diameter. However, a more useful expression of the curve’s slope is given by the constant negative exponential rate, k, where k = eb

[3.1]

e is the base of the natural logarithm, and b is the curve’s slope coefficient. Equation 3.1 thus restates b as a negative exponential rate of depletion for increasing basal diameter classes. Accordingly, k corresponds to the survival rate of seedlings or seedling sprouts per unit of increase in diameter (as opposed to change over time). Therefore k represents the probability of a seedling or seedling sprout surviving long enough to attain a basal diameter of 1 inch. For the black oak data, k is approximately 0.000001. However, because 1 inch is beyond the observed data range, it may be more meaningful to derive a survival rate per 0.1 inch of basal diameter increase. If we call this rate k0.1, then k0.1 = eb/10

[3.2]

which yields a rate of 0.27 per 0.1 inch.


125

Overstorey trees (no./acre)

Seedlings and seedling sprouts per acre

2800 2400 2000 1600 1200

20 16 12 8 4 0 2

800

6 10 14 18 22 Dbh (inches)

400 0 0.0

0.1

0.2

0.3

0.4

0.5

Basal diameter (inches)

White oak

Black oak

Fig. 3.5. The density of black oak and white oak seedlings and seedling sprouts in relation to their basal diameter under a relatively undisturbed stand on a droughty outwash sand in northern Lower Michigan (black oak–white oak/Vaccinium type as defined by Host and Pregitzer, 1991). The black oak curve is given by the negative exponential model: N = 6581.96(e–13.107d ) where N is number of seedlings and seedling sprouts per acre, e is the base of the natural logarithm, and d is the diameter of seedlings or seedling sprouts measured in inches at the ground line. The white oak curve is given by the power function model: N = 3.81d –1.886 The inset graph shows the diameter distribution of the overstorey, which is at 96% stocking based on Gingrich’s (1967) stocking equation. (Authors’ data.)

If current stand conditions were sustained, we would expect about 27% of the population of black oak reproduction to survive to be recruited into the next larger 0.1 inch basal diameter class, 7% (0.272) into the second larger 0.1 inch class, and 0.5% (0.273) in the third larger class. For larger diameter classes, recruitment drops to 0.1% or less. In contrast to black oak’s constant recruitment rate, white oak’s rate continually changes with increasing basal diameter, and this relationship is better described mathematically by the power function (Fig. 3.5). The probability that a white oak seedling or seedling sprout of a given initial size survives to grow into the next larger size class increases as size increases. Assuming that rates are essen-

tially constant over short intervals of basal diameter, the recruitment rate from the 0.1 to the 0.2 inch diameter class is about 18% whereas the rate from the 0.2 to 0.3 inch class is 30% (Table 3.1). The magnitude of increase in recruitment rates with increasing size of reproduction (Table 3.1) suggests that, for a given number of initially established seedlings, white oak is a more aggressive accumulator of seedlings than is black oak in this ecosystem. This is consistent with white oak’s greater shade tolerance, which is also evident in the differences between the overstorey diameter distributions of the two species (Fig. 3.5 inset). The relatively large numbers of white oaks in the small overstorey diameter classes largely

126

Chapter 3

Table 3.1. The estimated probability (P) that a seedling or seedling sprout of white oak advance reproduction of an initial basal diameter survives to a specified future basal diameter class in black oak–white oak/Vaccinium forests in northern Lower Michigan.a Future basal diameter class (inches) Initial basal diameter (inches)

0.2

0.3

0.1 0.2 0.3 0.4

0.18 – – –

0.16 0.30 – –

0.4

0.5

0.15 0.27 0.41 –

0.14 0.24 0.37 0.52

P

aBased on fitting the observed frequency distribution of reproduction by basal diameter classes at one point in time to the power function equation: N = 3.81D1.886, where N is the estimated number of seedlings and seedling sprouts per acre, and D is basal diameter in inches at the ground line for a seedling or for the largest stem for rootstocks with multiple stems. P is derived from Equation 3.2. The values of P shown assume that rates of change within 0.1 inch intervals of basal diameter are constant. (Authors’ data.)

represent suppressed trees. Where a canopy gap occurs, larger seedling sprouts may be recruited into the overstorey. The lower end of the overstorey diameter distribution therefore may include both recently recruited trees, and trees declining in growth and vigour that are succumbing to inter-tree competition. The paucity of black oak in the smaller diameter classes reflects its shade intolerance and inability to endure in the sub-canopy, and its low probability of recruitment into the overstorey. Where overstorey density remains uniformly high, the reproduction of both species remains in a largely suppressed state. To remain in that state, the oak reproduction must be continually renewed in endless short cycles of seedling input and mortality. In conceptualizing the reproduction accumulation process, it is convenient to label ecosystems where oak reproduction has a natural propensity to accumulate over long periods as intrinsic accumulators and those with little such propensity as recalcitrant accumulators. Oak reproduction probably accumulates, at least at times, over several acorn crops wherever oaks naturally occur. Accumulation is facilitated in part by the relatively large food reserves of acorns that sustain

seedlings, even when they grow under low light, through the critical first year of establishment. During germination, roots emerge before shoots (Fig. 2.15). Rapid taproot growth, deep penetration, recurrent shoot dieback and resprouting favour root development over shoot development and thus reinforce the accumulation process. The accumulation of oak reproduction even in mesic and wet forests therefore may be more pronounced than that of associated non-oaks (e.g. in the bottomland oak forests of Texas) (Fig. 3.1). A more rigorously defined classification of accumulator types incorporates the known or assumed capacity of oak advance reproduction to replace the parent stand and its relative constancy in doing so. In the eastern United States, a high capacity for replacement of overstorey oaks by oak reproduction occurs most frequently in the drier ecosystems. There, the accumulation of oak reproduction is often intrinsic and relatively independent of forces external to the ecosystem itself. Oaks in such ecosystems tend to be self-replacing, and therefore are not readily displaced successionally by non-oaks. These ecosystems therefore are relatively resilient, i.e. they tend to return quickly to their predisturbance state after disturbance.


One of the largest North American ecosystems of this type is the Ozark Highlands of Missouri, where oak reproduction characteristically accumulates in the understorey for decades, even in the absence of large-scale disturbances. The accumulation of oak reproduction is nevertheless strongly correlated with topographic features such as slope position and aspect (Sander et al., 1984; Dey, 1991; Dey et al., 1996b), which collectively influence light, heat and soil moisture. Oak forests that similarly accumulate oak reproduction occur on the drier sites in the Ohio Valley (Minckler and Woerheide, 1965), the Appalachians (Trimble, 1973; Ross et al., 1986) and the Lake States (Johnson, 1966, 1992; Arend and Scholz, 1969). Ecosystems that are intrinsic accumulators do not necessarily accumulate large numbers of oak reproduction. For example, about 1200 seedlings and seedling sprouts per acre characterize some Ozark oak stands (Sander, 1979b). Only 150–300 new oak seedlings per acre may become established, even after a good acorn crop (Sander, 1979b; McQuilkin, 1983). In contrast, more than 100,000 Nuttall oak seedlings per acre may become established after a single bumper acorn crop in bottomland forests that are recalcitrant accumulators (R.L. Johnson, 1975). Thus, large numbers of oak seedlings by themselves may not identify ecosystems that intrinsically accumulate oak reproduction. These anomalies emphasize the degree to which regeneration processes can vary among different kinds of oak forests. Recalcitrant accumulation of oak reproduction is characteristic of mesic and hydric ecosystems. There, oak reproduction may accumulate only after prolonged or recurrent disturbances spanning successive acorn crops. In the absence of disturbance, these oak forests usually are successionally displaced by non-oaks (Carvell and Tryon, 1961; Trimble, 1973; R.L. Johnson, 1975; Loftis, 1990a; Nowacki et al., 1990; Will-Wolf, 1991; Abrams and Nowacki, 1992). In the absence of disturbance, a dense overstorey with a subcanopy of small trees often develops

127

(Braun, 1972; Loftis, 1990b). Vertical stratification of tree crowns together with high total stand density creates extremely low light intensities on the forest floor. Nevertheless, reported oak reproduction densities in North American forests exceed 50,000 seedlings and seedling sprouts per acre in mesic habitats (Tryon and Carvell, 1958) and 100,000 per acre in bottomlands (R.L. Johnson, 1975). But such ecosystems also may be frequently depauperate of oak reproduction (Sander, 1983). When oak reproduction does occur, few age cohorts are represented because of low survival rates. Most of the seedlings from a single cohort often die before the next good acorn crop occurs because respiration rates of seedlings growing in deep shade often exceed their photosynthetic rates (Hanson et al., 1987; Dey and Parker, 1996). Microsites where oak seedlings become established in large numbers may not be where they ultimately survive (Johnson, 1966; Harrison and Werner, 1984). Whereas the cool, moist microclimate of northeastfacing slopes may favour initial establishment, the more favourable places for long-term survival, root development and thus accumulation occur on the more southerly or neutral aspects (Carvell and Tryon, 1961; Sander et al., 1984; Walters, 1990). In eastern forests, there is a general inverse relation between site quality and regeneration success: the better the site the greater the competition and the more difficult it is to regenerate oaks (Arend and Scholz, 1969; Trimble, 1973; Lorimer, 1989; Loftis, 1990b). The accumulation of oak reproduction in relation to combined site and overstorey factors is indirectly illustrated by the demographics of reproduction density in upland forests of southeastern Ohio (Walters, 1990). There, forests are comprised of black, white, scarlet, northern red and chestnut oaks mixed with maples, American beech, yellow-poplar and other hardwoods. The region is characterized by heavily dissected low hills with relatively short slopes (200–600 ft long) and slope gradients ranging from less than 10 to over 50%. Site quality is highly variable and

Chapter 3

In the same region, the combined density of maple, American beech, yellow-poplar and other non-oak reproduction differs markedly from the oaks. The non-oaks attain highest densities on north-facing slopes and lowest densities on south-facing slopes. Maximum and minimum densities of the non-oaks tend to occur at 1° and 181° azimuth, respectively (Fig. 3.6B). But like the oaks, non-oak densities change with changing slope gradient and overstorey density. For a given aspect, highest densities occur on steep slopes and under low overstorey densities. Collectively, these landscape-level models illustrate important

A 1.5

10% slopes

1.0 30% slopes 360 270 180 90

re

es

)

0.5

h

ut im Az

80 70 Over 60 0 store 50 40 y sto cking (%)

(d

eg

0.0

B

30% slopes

4 3

10% slopes

80 70 Over 60 0 store 50 40 y sto cking (%)

re

0

360 270 180 90

eg

1

es )

2

(d

Seedlings and seedling sprouts (thousands/acre)


depends on soil and topographic factors. Oak reproduction densities tend to be greatest on hot southwesterly facing slopes and least on cool northeasterly facing slopes (Fig. 3.6). Maximum and minimum densities of oak reproduction greater than 1 ft tall occur at azimuths of 203° and 23°, respectively. However, azimuth effects change with slope gradient and overstorey density. On a given aspect, oak reproduction density is greatest on gentle slopes and decreases with increasing slope gradient. Moreover, for any given combination of slope and azimuth, reproduction density decreases with increasing overstorey density.

Az im ut h

128

Fig. 3.6. Estimated densities of oak (A) and non-oak (B) reproduction beneath stands in southeastern Ohio in relation to overstorey stocking, slope aspect (degrees azimuth) and slope gradient (%). Includes reproduction at least 1 ft tall and up to 1 inch dbh. Oaks include black, white, scarlet, northern red and chestnut oaks. Non-oaks include red maple, yellow-poplar, hickories, sugar maple, American beech, white ash, American basswood, yellow buckeye, sycamore, black gum, black walnut, elms and aspen. The models explain 25% and 51% of the variation in oak and non-oak reproduction density, respectively. Stocking was defined by stocking (relative density) equations. (Adapted from Walters, 1990, used with permission.)


differences between the demographics of oak and non-oak reproduction. The environmental conditions that favour successful oak regeneration, and thus the accumulation of oak reproduction, can be conceptualized as a regeneration window (Fig. 3.7). The window is relatively narrow in mesic sites where the slow initial growth of oak reproduction must be matched by a light intensity sufficient for seedling growth but low enough to discourage the development of competitors. At the expense of the oaks, which are intermediate in shade tolerance, a lower light intensity favours tolerant species and a higher light intensity favours intolerant species. As site becomes drier, the window expands in response to the drought tolerance of the oaks that occur there and the concomitant exclusion of long-lived competitors that are less tolerant of drought. Reduced stand density also tends to create warmer and drier microclimates than those prevailing in the undisturbed forest

129

(Dey and Parker, 1996). ‘Xerifying’ disturbances such as fire, grazing and insect defoliation may consequently widen the oak’s regeneration window. The regeneration window emphasizes the gradational nature of the oak regeneration process with respect to light and moisture. The propensity of ecosystems to accumulate oak reproduction is similarly gradational. Accordingly, it is useful to consider an intermediate category: ambivalent accumulators of oak reproduction. These can be defined as ecosystems with uncertain or pliant tendencies towards accumulating oak reproduction. Such ecosystems would be expected to lie near the centre of the regeneration window (xero-mesic ecosystems). There, even minor disturbances could produce pronounced shifts in the accumulation of oak reproduction. Identifying these forests is important because they are likely to be especially amenable to silvicultural manipulation (Bakken and Cook, 1998).

Open

Light gradient

The oak regeneration window

Closed canopy Moist

Dry Moisture gradient

Fig. 3.7. The ‘regeneration window’ for oaks in eastern United States forests in relation to light and soil moisture. The window (open area) defines the region most favourable for successful oak regeneration. This region is relatively narrow on moist sites but widens with increasing dryness. On moist sites, intermediate light intensities offer the best opportunities for oak seedling survival and growth. Lower light intensities are insufficient to meet the oak’s minimum light requirements for photosynthesis but are sufficiently high to inhibit the development of many competitors. As site conditions become drier, the window widens because of the oak’s drought tolerance and the exclusion of competitors that are less drought resistant. (Adapted from Hodges and Gardiner, 1993.)

130

Chapter 3

Categorizing ecosystems by their propensity to accumulate oak reproduction can be incorporated into ecological classification systems to characterize oak’s replacement potential within each ecological unit. Population characteristics useful for deriving such ratings include the expected number and size distribution of oak reproduction (basal diameters and/or heights), and the relative constancy of those characteristics over time within an ecological classification unit. Most classification systems generally describe reproduction characteristics and the successional status of oaks within classification units (e.g. Smalley, 1978, 1984; Hix, 1988; Kotar et al., 1988; Cleland et al., 1993). To date, none specifically considers the replacement potential of oaks based on observed reproduction characteristics and associated successional relations, but oak accumulation types can sometimes be inferred from existing descriptions of the ecological units (Table 3.2).

Disturbance effects The presence of established oaks in the overstorey of a stand is often attributed to past stand disturbances. This is especially true in stands that are recalcitrant accumulators of oak reproduction even though there is often little or no evidence of previous disturbances that might explain the origin of the oaks. In the absence of future disturbances of the appropriate frequency and intensity, such oak stands are destined to successional replacement by other species. Past forest disturbances are often recorded in the spatial variability of overstorey density and size structure. In turn, variability in the overstorey usually is associated with variability in oak reproduction density. This is the case in xeric oak forests on outwash sands in northern Lower Michigan. There, more than half of the spatial variation in the density of white oak and black oak reproduction is explained by variation in overstorey density and the basal area of large oaks (Johnson, 1992) (Fig. 3.8). In this example, large trees are defined as those at least 12 inches dbh

(white oak) or 14 inches dbh (black oak) and are of a size generally considered to be better acorn producers than small trees (Downs, 1944). The reproduction density of both white and black oak increases as the basal area of large trees increases. Reproduction of these two species also differs in reaction to overstorey density (Fig. 3.8). Whereas white oak reproduction density decreases with increasing overstorey density, black oak reproduction density gradually increases with increasing overstorey basal area before declining at about 85 ft2 acre−1. Moreover, the two species differ in their apparent sensitivity to overstorey density. The rapid decrease in white oak reproduction density per unit increase in overstorey density contrasts with relatively small changes in black oak reproduction density with change in overstorey density. The two species similarly differ in their sensitivity to the basal area of large diameter trees. White oak reproduction density increases more rapidly per unit basal area of large trees than does black oak. This relation may be associated with white oak’s greater acorn production per unit crown area than black oak’s (Myers, 1978). Consequently, the total maximum density of oak reproduction occurs under conditions that favour white oak reproduction. Despite its regeneration disadvantages, black oak typically maintains a position of dominance in these forests. This persistence may be related to its height and diameter growth, which are greater than that of white oak during the first 100 years of stand development (Trimble, 1960; Carmean, 1971). However, other factors, including stand history and site quality also influence oak reproduction density (Carvell and Tryon, 1961; Arend and Scholz, 1969; Ross et al., 1986; Nowacki et al., 1990; Walters, 1990). Although relations between overstorey density and oak advance reproduction density imply disturbance effects (e.g. Figs 3.7 and 3.8), disturbance effects are directly considered in a predictive model for West Virginia forests (Fig. 3.9). The model expresses disturbance as a subjectively derived index ranging from 0 (none) to 16


131

Table 3.2. Inferred oak reproduction accumulation types for upland plant associations on the HuronManistee National Forests in Michigan.a See Fig. 1.8 for a schematic illustration of the landforms listed in this table. Plant association Northern pin oak–white oak/ Deschampsia Black oak–white oak/ Vaccinium Mixed oak–red maple/ Trientalis Northern red oak–red maple/ Viburnum Northern red oak–red maple/ Desmodium Sugar maple– beech/ Maianthemum Sugar maple– white ash/ Osmorhiza Sugar maple– white ash/ Caulophyllum

Soil and landform characteristics

Characteristic overstorey/understorey compositionb

Site index (ft)c

Inferred accumulation typed

Excessively well-drained sands of outwash plains

Northern pin, black, and white oaks; jack pine/oaks, black cherry

N. Pin oak: 48

Intrinsic

Excessively well-drained sands of outwash plains

Black, northern pin, and white oaks/red maple, oaks

Black oak: 50–56

Intrinsic

Well to excessively well-drained sands of overwashed moraines and kame terraces Moisture-enriched, well-drained sands of moraines and ice-contact topography

White, black, and northern red oaks; red maple/red maple, witch-hazel Northern red and white oaks; red maple/red maple, black cherry, flowering dogwood, and witch-hazel Northern red and white oaks; red maple/red maple, black cherry, flowering dogwood, and witch-hazel Sugar maple, American beech, northern red oak, red maple/ sugar maple, red maple, American beech Sugar maple, white ash, northern red oak, American basswood, red maple/sugar, American beech Sugar maple, white ash, northern red oak, American basswood, red maple/sugar maple, American beech

N. Red oak: Ambivalent 61–65

Moderately well-drained sandy loams over loamy substrata of ground moraines and glacial lake beds Well-drained morainal medium to fine sands

Moisture- and nutrient-enriched morainal sands Well-drained to moderately well-drained sands over fine loamy material

N. Red oak: Recalcitrant 77

N. Red oak: Recalcitrant 85

N. Red oak: Recalcitrant 76–88

N. Red oak: Recalcitrant 86 N. Red oak: Recalcitrant 76–88

aAdapted

from Cleland and others (1993). composition of relatively undisturbed natural stands. Understorey includes trees and shrubs 1–3 inches dbh. an index age of 50 years. dBased on Cleland and others (1993), Host and Pregitzer (1991), Johnson (1992), and other sources. bCharacteristic cAt

(heavy). It considers intensity of disturbance and time since disturbance events that include burning, grazing and logging (Carvell and Tryon, 1961). Although the proportion of variation in reproduction density is unspecified, the factors included in the model (light, aspect and disturbance) are statistically significant. For a given disturbance intensity, the density of oak reproduction increases with increasing heat associated with slope aspect progressing from northeast (coolest) to southwest (hottest), and increasing light. Historically, fire was the pre-eminent disturbance factor that shaped and maintained environments favouring oaks. The extensive oak savannas the early European settlers encountered in the midwestern United States were products of a centurieslong fire history on a spatial scale unlikely

to be repeated. Although the ecological disturbances that created and sustained those savannas have long disappeared, their effects are still evident in some regions. Within the Forest–Prairie Transition Region of the Midwest, many of today’s closedcanopy oak forests probably originated from the savannas present before settlement by Europeans (Beilmann and Brenner, 1951; Grimm, 1984; Guyette and Cutter, 1991; Abrams, 1992). The historical record indicates that throughout much of North America during the presettlement and early settlement eras, fire created and sustained the conditions necessary for perpetuating oaks where they would not otherwise have occurred (Day, 1953; McClaran and Bartolome, 1989; Johnson, 1993; Lorimer, 1993; Guyette and Dey, 1995).

132

Chapter 3

16

SppBA White oak Black oak


16 12

13 10

8 7 4 28 16 4

4

0 60

70

80

90

100

Stand basal area (ft2/acre) Fig. 3.8. Estimated densities of black oak and white oak reproduction in relation to total stand basal area and the basal area of large overstorey trees of the same species (SppBA) in xeric oak forests in northern Lower Michigan. The trees comprising SppBA are the presumed primary acorn producers. For white oak, SppBA includes trees ≥12 inches dbh, and for black oak includes trees ≥14 inches dbh. Based on linear regression models that account for 55% and 63% of the observed variation in reproduction density of white oak and black oak, respectively. (Modified from Johnson, 1992, used with permission from Elsevier Science.) 40

50

Moderate disturbance

30

25

20

50

Light (%)

Oak reproduction density (thousands/acre)

Light disturbance

10 25 5 5 0 0 NE

45 N&E

90 SE & NW

135 S

180 SW

Aspect scale and aspect Fig. 3.9. The estimated density of oak seedlings and seedling sprouts beneath mixed oak stands in West Virginia in relation to slope aspect, light and disturbance. Includes white, black, northern red, scarlet and chestnut oak reproduction from 1 ft tall to 0.6 inch dbh. Aspect scale is displayed in degrees departure from northeast (45° azimuth). Light is expressed as the amount measured by a photoelectric cell beneath the overstorey as a percentage of that measured in the open. Disturbance is expressed as a qualitative index based on an arbitrary scale from 0 (none) to 16 (heavy) that considers intensity of and time since grazing, fire and logging; estimates for values of 2 (light disturbance) and 6 (moderate disturbance) are shown. (Adapted from Carvell and Tryon, 1961 by permission of the Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)


When fires were frequent and intense, open-grown oak savannas were created. Those events produced communities comprised of a few large thick-barked, fireresistant overstorey trees per acre, a unique and diverse grass and forb flora, and abundant oak reproduction largely comprised of seedling sprouts (Curtis, 1959; Haney and Apfelbaum, 1990). Oak savannas, which once dominated the landscape throughout much of the central and eastern states, are today rare plant communities (Haney and Apfelbaum, 1990). Following the cessation of burning and grazing, some savannas developed into closed-canopy forests; many others were converted to pasture and cultivated fields (Beilmann and Brenner, 1951; Curtis, 1959; Thilenius, 1968; Whitney, 1994). Recurrent fire promotes the accumulation of oak reproduction in various ways. Fire can destroy seed stored in the forest floor and soil and thereby eliminate or reduce post-fire competition from some shrub, herb and other tree species (Fig. 3.10). However, in some ecosystems, that effect may be offset by the presence of species whose seeds are resistant to fire and whose germination is stimulated by fire and the post-fire environment. The lat-

133

ter include several chaparral species associated with oaks of southwestern United States (Keeley, 1991) and pin cherry in the eastern forests (Wendel, 1990). In other species, seeds may be stored in the humus and soil where they are relatively unaffected by fire. Burning also can reduce overstorey and understorey density (DeSelm et al., 1991), which thereby increases light on the forest floor and reduces competition for other site resources. Within a species, small-diameter trees are the most vulnerable to fire because their thinner bark offers less insulation against potentially lethal heat (Harmon, 1984; Hengst and Dawson, 1993). Although species vary widely in bark thickness for a given diameter (Fig. 3.11), other physical properties of bark also vary among species. The effectiveness of bark in insulating a tree from heat depends on three physical properties: (i) thermal conductivity (ability to transfer heat); (ii) specific heat (ability to absorb heat); and (iii) thermal diffusivity (ratio of thermal conductivity to the product of specific heat and bark density) (Martin, 1962). Consequently, the survival of trees of a given diameter subjected to a given duration and intensity of heat varies among

Fig. 3.10. Fire eliminates many of the oak’s competitors by destroying or reducing seed stored in the forest floor and also by killing established fire-sensitive competitors. (USDA Forest Service, North Central Research Station photograph.)

134

Chapter 3

1.0

Blackgum

Bark thickness (inches)

A 0.8 0.6

Chestnut oak

0.4

Sourwood Virginia pine

0.2

Red maple 0.0 0

4

8

12

16

Bark thickness (inches)

2.0

20 Bur oak

B 1.6

White oak 1.2

Northern red oak

0.8

Sugar maple

0.4 Yellow poplar 0.0 0

10

20

30

40

Dbh (inches) Fig. 3.11. Relation between estimated bark thickness and dbh of trees in eastern United States. (A) Five species in dry pine–oak forests in the southern Appalachians. Shaded circles identify the bark thickness and the correlated dbh needed to assure that 50% of trees are not top-killed by a ‘low-intensity’ fire. The proportion of the variation in bark thickness accounted for by dbh ranged from 69% (Virginia pine) to 96% (blackgum). (Adapted from Harmon, 1984.) (B) Five species in Illinois. The proportion of the variation in bark thickness accounted for by dbh ranged from 56% (northern red oak) to 93% (bur oak). (Adapted from Hengst and Dawson, 1993.)

species. Such differences were observed in pine–oak stands on dry sites in the southern Appalachians where low-intensity fires primarily consumed leaf litter. There, the bark thickness required for 50% survival of a population of trees ranged from less than 0.1 inch to over 0.2 inch at dbh, depending on species (Fig. 3.11). Although chestnut oak is generally considered resistant to fire damage, the amount of bark required for the survival of chestnut oak is greater than that required for some non-oaks such as red maple (generally regarded as fire-sensitive). But because the chestnut oak can grow faster in diameter than red maple (at least on dry sites), the former can more

quickly attain an effective fire-insulating bark thickness (Harmon, 1984). In the Ozark Highlands of southeastern Missouri, oak species differed in their response to fire frequency over a 33- to 35year study period (Huddle and Pallardy, 1995). During that time, experimental plots were burned annually, periodically (every 4 years), or not burned. For plots burned annually or periodically, average survival of post oak (a white oak) was greater than that of red oaks (black, scarlet and southern red oaks combined) after statistically accounting for differences in initial (preburn) diameter between the two species groups. Among all treatments, tree survival


increased with increasing initial dbh up to 7 or 8 inches regardless of species and treatment. For post oak, fire frequency had no significant effect on the relation between survival and initial dbh. In contrast, the relation between survival of red oaks and initial dbh differed significantly among the three treatments. Differences in survival generally increased with increasing dbh. Among trees less than 7 inches dbh, lowest survival occurred within periodically burned plots and highest survival occurred within unburned plots. Hickories also responded negatively to burning (Fig. 3.12). Differences among species’ growth characteristics thus may interact with tree size, bark, and other species’ characteristics to determine survival of burned trees. Moreover, fire frequency and pre-burn conditions may further affect these differences. The long-term survival of trees subjected to low-intensity surface fires in the Great Smoky Mountains National Park in eastern Tennessee is largely dependent on tree growth rate and bark thickness. Fire intervals of less than 14 years favour the relatively fast-growing species (pines,

blackgum, sourwood and chestnut oak) over slow-growing species (red maple and hickories) and some fast-growing but relatively thin-barked species such as scarlet oak (Harmon, 1984). Longer intervals favour red maple, hickories and the thinbarked scarlet oak by allowing them to attain a sufficient size and bark thickness to resist fire damage. Moreover, once the more fire-sensitive species reach a safe size, similar future fires are less likely to effectively alter overstorey composition. In mixed white oak–yellow-poplar stands in the North Carolina Piedmont (red oak site index 75–80), responses to highintensity fires illustrate the superior resistance of oaks to top-kill (Maslen, 1989). The burns, described as strip head fires, produced flames 3–10 ft high that consumed most of the litter and caused wounding of the residual stand. Seven years after a single burn, top-kill from burning was observed as well as natural dieback (unrelated to burning) on paired burned and unburned areas. Although all reproduction of all species was top-killed by burning, oak saplings and small poles (trees > 12 ft

1.0

Post oak Red oaks (control)

0.8 Survival probability

135

Red oaks (annual burn)

0.6

Red oaks (periodic burn)

0.4

Hickories (control)

0.2 Hickories (all burns)

0.0 0

2

4 6 Initial dbh (inches)

8

Fig. 3.12. Estimated survival probabilities for oaks and hickories in relation to initial (pre-treatment) dbh and frequency of burning during a 33–35 year period in southeastern Missouri. Periodic burns occurred every 4 years. Species in the red oak group include black, scarlet and southern red oaks; hickories include shagbark, mockernut and black hickories. Survival of post oaks in burned plots did not differ significantly (P < 0.05) from trees in control plots. Probability estimates were derived by logistic regression. (From Huddle and Pallardy, 1996.)

136

Chapter 3

tall but ≤ 6 inches dbh) were more resistant to top-kill by fire than non-oaks of the same size (Fig. 3.13A). In unburned areas, the probability of top-kill among non-oaks was consistently lower than that of oaks across all size classes observed except the largest (Fig. 3.13B). Because virtually all the hardwood species associated with oaks can sprout, top-kill from one or a few single low-intensity burns may not effectively reduce or

eliminate these competitors (Johnson, 1974; Nyland et al., 1983; Walters, 1990). However, through repeated low-intensity burns (Langdon, 1981) and sometimes through single high-intensity burns (Maslen, 1989), differences among species in fire-caused mortality of tops, root systems, dormant buds near the root collar and decay resistance confer a competitive advantage to the oaks (Fig. 3.14). Oaks therefore are ‘fire persistent’ compared to

1.0

Oaks (n = 176) Non-oaks (n = 215)

Probability of no top-kill

A Burned 0.8

0.6

0.4

0.2

0.0

0 1–2 ft

0

0

3–5 ft 6–12 ft >12 ft 2–6 in 7–12 in –12 ft 2–6 in 7–12 in –12 ft 2–6 in 7–12 in – < 2 in Tree size class

Fig. 3.14. Survival probabilities for trees in immature mixed oak stands on mesic sites (red oak site index 75–80) in the North Carolina Piedmont. (A) After a ‘high-intensity’ prescribed burn, and (B) on adjacent unburned areas. Size classes are expressed as heights (ft) for reproduction (trees an acre). It usually involves a disturbance originating from forces outside (exogenous to) the stand such as fire, tornadoes, hurricanes and timber harvest. These events abruptly increase light, moisture and nutrients, which are also accompanied by significant changes in microclimate. This mode of regeneration often facilitates the temporary coexistence of trees of all shade tolerances (Roach and Gingrich, 1968; Dunn et al., 1983; Smith, 1986). Coexistence is facilitated by shade tolerant species originating from advance reproduction together with less tolerant species originating from new or buried seed and sprouts. In some mesic forests, this may produce stands of high tree diversity (Loftis, 1983; McGee, 1987; Smith and Miller, 1987). Even-aged meth-

Fig. 3.23. A northern red oak sapling growing in a canopy gap in a xero-mesic oak-mixed hardwood stand in northern Wisconsin. This sapling’s potential to eventually dominate the gap depends on its height growth and the rate of crown closure of surrounding trees. Adjacent trees in the main canopy include northern red oak, bigtooth aspen, paper birch, red maple and white ash. (USDA Forest Service, North Central Research Station photograph.)


ods of silviculture such as clearcutting, shelterwood, seed tree and their variants simulate the catastrophic mode of regeneration through the removal of all or most of the overstorey in one or a few steps (Marquis and Johnson, 1989). Oaks are well adapted to the catastrophic mode of regeneration, especially in xeric ecosystems where oak reproduction intrinsically accumulates. However, in mesic and hydric forests, successful regeneration of oaks may depend on disturbances from fire, grazing or flooding that precede a later catastrophic event that significantly reduces overstorey density. The intensity, frequency and recency of minor disturbances combined with later catastrophic disturbances collectively direct ecological succession (GlennLewin and van der Maarel, 1992). The continuous regeneration mode pertains to species that can attain maturity in the absence of a canopy opening. In the oak forests of the Eastern Deciduous Region of the United States, these include shade tolerant species such as flowering dogwood, redbud, American hornbeam, serviceberry and sourwood. These species usually are relegated to the sub-canopy because their maximum attainable heights seldom exceed 35 ft (Burns and Honkala, 1990). However, they do not require canopy gaps for seed production, germination or early growth. Terborgh (1985) theorized that the tops of the relatively flat, spreading crowns of flowering dogwood and redbud occur at a predictable distance below the main canopy. That distance occurs where the beams of solar radiation penetrating different canopy gaps most frequently intersect along the daily solar path during the growing season. Terborgh reasoned that the distinct layers of vegetation in a forest may be adaptations to relatively uniform light fields that occur at fixed distances below the main tree canopy, which in turn are related to the shape, size and distribution of tree crowns and gaps, together with angles of direct-beam solar radiation. Accordingly, vegetation would occur as distinct vertical layers, as commonly observed, rather than as a vertical continuum of tree crowns (Fig. 3.24).

151

Although Veblen (1992, p. 166) offers that ‘… the continuous mode of regeneration appears to be rare’, some shade tolerant species such as sugar maple and American beech are capable of growing directly into the overstorey if canopy density is not extremely high (Spurr and Barnes, 1980; Canham, 1989; Godman et al., 1990). In oak forests, high overstorey densities reduce light reaching the forest floor to levels that are usually insufficient for sustaining the growth of oak reproduction into the overstorey. The regeneration of oaks thus depends on forest disturbance, the number, size and spatial distribution of oak advance reproduction present at the time of disturbance, and other factors associated with site quality and competition (Sander, 1971; Sander and Clark, 1971; Trimble, 1973; Sander et al., 1984). When overstorey density is reduced, the oak advance reproduction may capture the vacated growing space. The resulting recruitment of oaks into the overstorey is slow in small canopy gaps that admit little light to the forest floor and more rapid in larger openings – provided that adequate reproduction is present (Fig. 3.25). Other factors being equal, the growth rate of oak advance reproduction after overstorey removal depends on its pre-disturbance basal diameter (Sander, 1971; Johnson, 1979), which is correlated with total root mass (Canadell and Rodà, 1991) (Fig. 2.27). In oak stands that accumulate oak advance reproduction, stand regeneration potential and succession largely depend on the composition and structure of that reproduction. This type of succession follows the ‘initial floristics’ model (Egler, 1954). Succession is driven by an initial floristics when the condition of a plant community at the time of disturbance largely controls its future development. Similar ideas are implicit in the ‘legacy’ concept of Franklin et al. (1989) wherein the future state of an ecosystem is perceived as partly, if not largely, inherited from propagules carried over from the previous state.

152

Chapter 3

Fig. 3.24. Flowering dogwood (seen here in full spring bloom) forms a distinct and permanent subcanopy in this oak–hickory forest. (USDA Forest Service, North Central Research Station photograph.)

A

B

C

Fig. 3.25. Recruitment of oak reproduction into the overstorey. (A) No recruitment; advance reproduction (trees below the thin horizontal line) is suppressed by high overstorey density. (B) Some recruitment occurs in the centre of small canopy gaps. (C) Recruitment accelerates with increasing gap size. Maximum height growth and recruitment occur in centres of gaps where light is maximal and competition from border trees is minimal.

The dependence of the future condition of a stand on its initial state is an important concept in the development and application of predictive regeneration models (e.g. Sander et al., 1984; Loftis, 1990a; Marquis et al., 1992; Johnson and Deen, 1993) (Fig. 3.26). When overstorey density is reduced by fire or timber harvesting, the overstorey itself is often an important component of the initial

state of an oak stand because a significant proportion of the reproduction may originate from dormant buds at the bases of top-killed trees or stump sprouts (P.S. Johnson, 1975, 1977; Wendel, 1975; Lamson, 1988; Weigel and Johnson, 1998). The importance of this source of reproduction to the future stand depends on the proportion of trees that produce basal or stump sprouts. In general, older


153

ASPECT

NW (neutral)

0.37 0.42

NE (cool)

0.28 0.10

0.13

0.27 0.15

0.33 0.12

0.10

0.20 0.08

Lower Middle Upper Slope position

0.32 0.37 0.23

0.37

0.12

0.13

0.42 0.15

0.13

0.09

0.27 0.10

SW (hot)

SE (neutral)

Initial size of advance reproduction (height, basal diameter) 4 ft, 0.5 in

P

2 ft, 0.25 in

P

P = probability of attaining intermediate or larger crown class 21 years after overstorey removal.

Fig. 3.26. Estimated probabilities that oak advance reproduction will attain an intermediate or larger crown class 21 years after complete overstorey removal in the Ozark Highlands of Missouri. Probabilities are shown in relation to aspect for two preharvest size classes (4 ft tall, 0.5 inch in basal diameter, and 2 ft tall, 0.25 inch in basal diameter). The probabilities illustrate the predictive value of the size of oak advance reproduction (i.e. the ‘initial state’) as predictors of a forest’s future state. Probabilities apply to black, white, scarlet, northern red and post oaks, and are based on the predictive regeneration model ACORn. (Adapted from Dey et al., 1996b.)

and larger trees are less likely to produce stump sprouts than smaller and younger trees (Roth and Sleeth, 1939; P.S. Johnson, 1975, 1977; Ross et al., 1986) (Fig. 2.24). But even in older stands, stump sprouts may contribute significantly to the future occupancy of growing space (stocking) because, despite their small numbers, growth and survival rates of stump sprouts are usually high (Johnson and Rogers, 1984; Lamson, 1988).

The ecological significance of the predisturbance state of an oak forest is not realized until the inhibiting effect of the overstorey is released. The effect accordingly has been referred to as the ‘inhibition’ model of succession (Connell and Slatyer, 1977). Accordingly ‘… replacement occurs only when resources are released by the damage or death of the previous occupant [of the site]’ (Connell and Slatyer, 1977,

154

Chapter 3

p. 1138). An oak forest’s regeneration potential therefore is often encoded in the composition and structure of its advance reproduction and the overstorey. But this potential is not expressed until it is freed from the inhibitory effect of the overstorey. In a mature oak stand, site resources are partitioned among the various vegetative layers. Large trees utilize resources (e.g. soil moisture and nutrients) from the relatively large and heterogeneous spaces they occupy, whereas smaller trees and other low vegetation are limited to relatively small and homogeneous microsites. The early postdisturbance years after overstorey removal are characterized by the direct competition of oak reproduction with a chaotic mixture of

herbaceous, shrub and other tree reproduction, which collectively occupy relatively narrow above- and below-ground strata. Under these conditions, herbaceous and shrub species interfere directly with the growth and survival of tree reproduction. This period is characterized by rapid competitive sorting among and within species. During this period, the composition and structure of the future forest are often neither obvious nor apparently predictable. The oaks may win or lose in the competitive sorting process, depending on antecedent states and events (Fig. 3.27). Nevertheless, a more orderly partitioning of resources gradually reemerges but only becomes visually conspicuous after tree crowns again vertically stratify.

Forest type Non-oak Accumulation of oak reproduction: RECALCITRANT

Mixed oak Oak

Accumulation mediated by disturbance Oaks win Mixed outcome Oaks lose or no oaks present

Not mediated Accumulation of oak reproduction: INTRINSIC

Competitive sorting

Advance tree reproduction and seed bank inhibited by overstorey

Disturbance-mediated release from overstorey inhibition Fig. 3.27. Disturbance-mediated successional pathways in relation to intrinsic and recalcitrant patterns of accumulation of oak reproduction. In the eastern United States, accumulation is usually an intrinsic characteristic of xeric forests. In contrast, accumulation is generally recalcitrant in mesic and hydric forests in the absence of disturbances such as recurrent fire. In the western United States (in Mediterranean and semi-desert climates), accumulation may be favoured in the more moist ecosystems and microhabitats. The outcome of competitive sorting among species during the first two decades after disturbance is largely determined by advance reproduction characteristics and site factors. In eastern United States, oaks typically win the post-disturbance race to capture growing space in xeric ecosystems. In mesic and hydric ecosystems, oak’s ascendance to dominance is more variable, and highly dependent on the type, frequency and intensity of disturbance, which in turn influences the accumulation process. Oak reproduction also can occur and accumulate beneath non-oak types via acorn dispersal by birds and mammals.


Putting the necessary information on oak regeneration potential into a practical and useful framework for predicting the future composition and structure of oak stands requires quantitative regeneration models.

Modelling theory and objectives In their broadest sense, models represent ‘… in some way the form and/or the function of real-world entities and processes’ (Kimmins, 1987, p. 460). Models therefore may range from unrevealed thoughts to more externally represented ideas expressed by words, pictures, graphs, charts, mathematical equations and computer simulations (Kimmins, 1987). The latter has largely made modelling the valuable tool that it has become for predicting and understanding the behaviour of ecosystems generally and forests specifically. Many predictive models have been developed including the successional models of ecologists (e.g. Shugart, 1984; Urban and Shugart, 1992; Botkin, 1993), and the growth and yield models of silviculturists (e.g. Belcher et al., 1982; Hilt, 1985; also see Chapter 10). It has been suggested that complete knowledge of the state of an ecosystem (i.e. its multidimensional structure or ‘spacestate’) at any given instant cannot be fully specified because of the large number of factors, living and non-living, that define ecosystems (Margalef, 1963; Jørgensen, 1990). Even if such a model could be constructed, its complexity would probably render it useless. This produces the conundrum that an ecosystem model can only approach reality through increasing complexity, which in turn increasingly reduces its utility. Nevertheless, relatively simple ecosystem models can have surprisingly predictive power (Urban and Shugart, 1992), and such models have become widely used to predict forest behaviour. It is none the less important to recognize the limitations of models. The suitability of a model should be evaluated by how well it

155

meets the objectives of its intended application (Buchman and Shifley, 1983; Blake et al., 1990; Bruce, 1990). Forests represent hierarchies of biotic organization ranging from the cellular level upwards to the individual organism level and extending further upwards to species’ populations, associations of species and beyond. Depending on its purpose, a model should be designed to consider specific levels of detail within hierarchies (Allen and Starr, 1982). Models developed for silviculturists usually focus on predicting population phenomena such as time-dependent changes in the number and size of trees in a stand. Providing a silvicultural explanation of phenomena at the population level typically requires consideration of detail at least one level lower, i.e. for the individual trees in the population (Botkin, 1993). If the ‘explanatory’ level of a model is set too far below the ‘predictive’ level, the detail required by the model may exceed the limits of available knowledge or the ability to meaningfully build such details into a model. On the other hand, a more hierarchically detailed model may have more explanatory power. The ‘bottom-up’ approach (aggregating model components from fine to coarser scales) is consistent with the scientific method and the building of scientific theories (Lewis, 1990). It is also the most common approach used in modelling forest dynamics. However, a ‘top-down’ approach to modelling also is possible and begins with a relatively coarse scale, but adds details in subordinate levels (Landsberg, 1986). In addition to hierarchical structure, models also possess other properties that determine their usefulness for meeting defined objectives. Generality, realism and accuracy are three such properties (Levins, 1966; Sharpe, 1990; Botkin, 1993). Generality refers to the range of situations that a model can be applied to. Realism refers to the qualitative similarity between model projections and the real world (e.g. as demonstrated by the similarity in the shapes of projected and actual response curves). Accuracy4 refers to the quantitative

4We use the term accuracy, as Botkin (1993) did, instead of precision as originally used by Levins (1966) and later by Sharpe (1990) to discuss these model properties. In the context of models, accuracy is more consistent with its parallel meaning in statistics and the sciences in general. Also see Sokal and Rohlf (1969).

156

Chapter 3

closeness of model projections to the real world. The goal or objective of the model determines the relative importance of each of these properties. However, there are limitations to each of the various approaches to modelling because each tends to rely on one of three essentially mutually exclusive approaches that are either: (i) statistical, (ii) mathematical, or (iii) science-based. As a consequence, one of the three properties is likely to be sacrificed in any one model (Levins, 1966; Sharpe, 1990). In their construction, statistically derived models require relatively large amounts of data. Fitting curves to data using regression analysis often is used to derive relations among variables. The resulting models tend to lack generality because their efficacy in application is limited to the conditions represented by the data on which they are based. Nevertheless, within that limitation, such models are potentially (but not necessarily) accurate and realistic. Most of the models developed for silvicultural applications, including regeneration and growth and yield models, fall into this category. In contrast, mathematical models are derived from abstract relations assumed to describe systems within narrowly defined conditions. They may require no data and are often designed to investigate generalized theories of ecosystem behaviour such as stability and response characteristics (e.g. Gatto and Rinaldi, 1987). Such models may possess accuracy and generality, but often sacrifice realism. Realism is lost because of the recognized unrealistic assumptions about the system that must be made in order to conform the system to the model’s known limitations. Mathematical models are usually used to explore theory and are not generally applied to predict outcomes for real ecosystems (Prentice and Helmisaari, 1991). Sharpe (1990) defined a third class of models that he termed science-based models. They are more common to ecology than silviculture and are derived from available information on species’ characteristics such as growth rates, longevity, reaction to stress factors, and other natural history

characteristics. In their construction, science-based models require less data than statistical models but more data than mathematical models. The approach has been widely used to develop forest succession models (e.g. Shugart, 1984; Prentice and Helmisaari, 1991; Botkin, 1993) and to a lesser extent regeneration models (e.g. Waldrop et al., 1986). Science-based models are potentially realistic and general but sacrifice accuracy. Accuracy is lost because of the generalized database from which they are derived. Each of these three classes of models is also associated with specific methods and problems of validation (Sharpe, 1990). The users of models in forestry are largely forest managers who are often concerned with the consequences of silvicultural actions. Accurately predicting the outcome of silvicultural actions is important to managers. However, accurate prediction does not infer that ecological or physiological processes are well understood or even explicitly considered in model development. On the contrary, models with high accuracy often are not based on the mechanisms that directly explain the predicted phenomenon. High accuracy and low explanatory power characterize most of the current forest growth and yield models (Sharpe, 1990; Prentice and Helmisaari, 1991). Such models are largely empirical in that the variables used to make predictions are only indirectly related to the actual mechanisms that govern stand development. For example, a model predicting growth of individual trees may employ predictors such as the initial dbh, the crown class, the ratio of the living crown to its total live height (live crown ratio) and some measure of site quality (e.g. site index). Of course, the underlying causes of tree growth (and death) are rates of photosynthesis, respiration, uptake of water and nutrients, and other physiological mechanisms and the stress factors that limit those processes. Models based on causative physiological mechanisms are sometimes referred to as process models (Blake et al., 1990; Bruce, 1990). There is strong interest in the


development of ecosystem models that directly consider physiological processes such as photosynthesis and carbon allocation in trees (Dixon et al., 1990). Their potential advantages lie in their greater generality of application and explanatory power over empirical models (Blake et al., 1990; Isebrands et al., 1990; Sharpe, 1990). However, the advantages of generality are likely to be counterbalanced by reduced accuracy and increased complexity of design and efficiency of application (Levins, 1966; Sharpe, 1990). It also can be argued that all models are empirical because causation at some scale (e.g. molecular) is not addressed by the model.

Stand-level regeneration models: purpose, problems and limitations From an ecological perspective, regeneration models could be defined as predictors of the outcome of short-term secondary succession associated with planned disturbances. They are designed to function as silvicultural tools for assessing the adequacy of the regeneration potential of stands for meeting pre-defined regeneration goals. Successional models usually do not fulfil that function adequately because they usually do not provide the stand-level accuracy needed by silviculturists (Waldrop et al., 1986). Growth and yield models for established stands generally qualify as possessing accuracy but they usually do not explicitly consider regeneration. For example, changes in tree populations over the first two decades after final harvest of even-aged stands usually are not included in (or are only coarsely simulated by) growth and yield models. This leaves an informational void on stand development for one-fifth to one-fourth of the 80to 100-year rotations typically used in managing oak forests (Chapter 7). Like all models, regeneration models are imperfect in mimicking real ecosystems because the models are simplifications of reality. Although regeneration models are largely empirical, they may be more precise and realistic than some

157

other kinds of forest dynamics models. However, they typically possess little generality and explanatory power because they are usually statistically derived and based on correlations rather than on causative mechanisms. Several predictive regeneration models nevertheless have been developed for oak and associated mixed hardwood forests (Sander et al., 1984; Waldrop et al., 1986; Loftis, 1990a; Marquis et al., 1992; Johnson and Deen, 1993; Dey et al., 1996a,b). While these models differ in their details and complexity, they are all based partly or wholly on the successional concepts of initial floristics (Egler, 1954) and inhibition (Connell and Slatyer, 1977). In other words, they assume that the future state of a forest is encoded in its current or initial state. Moreover, overstorey removal or its reduction in density is required to release the future stand from the inhibiting effects of the present overstorey. Because such models are predictors of short-term secondary succession, they are potentially useful as tools for guiding silvicultural decisions. In application, regeneration models are usually designed for specific geographic regions or ecosystems, and they require quantitative information on the initial state of a stand including current vegetation and site characteristics. For models to be silviculturally practical, the initial state, or predictor variables, must be easily measurable and have predictive power. Those requirements create practical and theoretical problems in the development of regeneration models including: (i) the selection of a relatively small but useful set of easily measured predictors; and (ii) the relatively large variation in tree growth and survival during the post-disturbance regeneration period. These problems are further confounded by the relatively long period of observation (up to two decades) required to obtain data from which to build some types of models. Predictors of the post-disturbance state usually include measurements and/or counts of advance reproduction obtained from sample plots within stands (e.g.

158

Chapter 3

Sander et al., 1984; Marquis et al., 1992; Dey et al., 1996b). Information on advance reproduction heights and/or diameters is a requirement for applying many regeneration models. Although those variables are easily measured attributes of reproduction, they may not be the most important determinants of a stand’s future state. For example, the growth potential of an oak seedling may be more directly and accurately related to its root mass and leaf area (Johnson, 1979). But because those measurements are impractical to obtain in silvicultural practice, height and diameter are the measurements of necessity. Other sources of variation considered in some models include the density and size of competing shrubs and herbs. The ‘initial floristics’ approach used in prediction also may limit the predictive power of models because it downplays potential contributions of trees originating from seed and stump sprouts.

Fluctuations in weather, insect and mammal populations that affect regeneration processes may be impossible or difficult to predict. Models therefore may lack predictive power because of their inability to consider important factors, both known and unknown. Building regeneration models is also complicated by the inherent variation associated with the disturbed states that characterize regenerating forests. The sudden post-disturbance increase in light triggers the germination of seeds of many tree, shrub and herbaceous species. Some seedlings arise from seed buried in the forest floor while some are carried in from outside the disturbed area. Advance tree reproduction also is released from suppression and new sprouts may develop from dormant basal buds of cut trees. This produces a chaotic mix of established plants and new propagules that are suddenly competing for space

INPUT Stand and site characteristics: Overstorey Advance reproduction

REGENERATION SIMULATOR

Overstorey harvested

OUTPUT Regenerated stand characteristics: Species Tree size distributions Stocking

Fig. 3.28. Data input and output elements for a representative regeneration simulation model. Input requirements and output vary among available models (see Chapter 7).


in the new environment. The result is a relatively unstable ecological state that produces changes in stand composition and structure that are difficult to predict. During this period, stand development often is so uncertain that the outcome may best be expressed probabilistically (e.g. Fig. 3.26). Despite these problems, regeneration models attempt to bring order to the chaos of this least predictable stage of stand development. Regeneration models generally fall into one of two categories: (i) decision guides for evaluating the adequacy of a stand’s regeneration potential for meeting defined silvicultural objectives; and (ii) simulation models that predict future stand composi-

159

tion and structure. The former are largely limited to providing a ‘yes’ or ‘no’ answer to the question of the adequacy of regeneration potential. Such models also may be partially or wholly based on experience or ‘expert opinion’. This class of models may range from the relatively simple to the complex, and their application often does not require a computer. In contrast, simulation models predict future stand composition and structure. Because of their complexity, application requires computer software (Fig. 3.28). Both types of regeneration models have been developed for oak and mixed forests in several regions (Chapter 7).

References Abrams, M.D. (1992) Fire and the development of oak forests. BioScience 42, 346–353. Abrams, M.D. and Nowacki, G.J. (1992) Historical variation in fire, oak recruitment, and post-logging accelerated succession in central Pennsylvania. Bulletin of the Torrey Botanical Club 119, 19–28. Abrams, M.D. and Scott, M.L. (1989) Disturbance-mediated accelerated succession in two Michigan forest types. Forest Science 35, 42–49. Allen, T.F.H. and Starr, T.B. (1982) Hierarchy: Perspectives for Ecological Complexity. University of Chicago Press. Arend, J.L. and Scholz, H.F. (1969) Oak forests of the Lake States and their management. USDA Forest Service Research Paper NC NC-31. Auchmoody, L.R. and Smith, H.C. (1993) Survival of northern red oak acorns after fall burning. USDA Forest Service Research Paper NE NE-678. Aust, W.M., Hodges, J.D. and Johnson, R.L. (1985) The origin, growth and development of natural, pure, even-aged stands of bottomland oak. USDA Forest Service General Technical Report SO SO-54, pp. 163–170. Bakken, P.N. and Cook, J.E. (1998) Regeneration potential of six habitat types common to northcentral Wisconsin. Northern Journal of Applied Forestry 15, 116–123. Barnhart, S.J., McBride, J.R. and Warner, P. (1991) Oak seedling establishment in relation to environmental factors at Annadel State Park. USDA Forest Service General Technical Report PSW PSW126, pp. 25–30. Beck, D.E. (1970) Effect of competition on survival and height growth of red oak seedlings. USDA Forest Service Research Paper SE SE-56. Beck, D.E. and Hooper, R.M. (1986) Development of a southern Appalachian hardwood stand after clearcutting. Southern Journal of Applied Forestry 10, 168–172. Beilmann, A.P. and Brenner, L.G. (1951) The recent intrusion of forests in the Ozarks. Annals of the Missouri Botanical Garden 38, 261–282. Belcher, D.M., Holdaway, M.R. and Brand, G.J. (1982) A description of STEMS, the stand and tree evaluation and modeling system. USDA Forest Service General Technical Report NC NC-79. Blake, J.I., Somers, G.L. and Ruark, G.A. (1990) Perspectives on process modeling of forest growth responses to environmental stress. In: Dixon, R.K., Meldahl, R.S., Ruark, G.A. and Warren, W.G. (eds) Process Modeling of Forest Growth Responses to Environmental Stress. Timber Press, Portland, Oregon, pp. 9–17. Borchert, M.I., Davis, F.W., Michaelsen, J. and Oyler, L.D. (1989) Interactions of factors affecting seedling recruitment of blue oak (Quercus douglasii) in California. Ecology 70, 389–404. Bormann, F.H. and Likens, G.E. (1979) Pattern and Process in a Forested Ecosystem. Springer-Verlag, New York.

160

Chapter 3

Botkin, D.B. (1993) Forest Dynamics, an Ecological Model. Oxford University Press, New York. Bowersox, T.W. and McCormick, L.H. (1987) Herbaceous communities reduce the juvenile growth of northern red oak, white ash, yellow poplar, but not white pine. Proceedings of the Central Hardwood Forest Conference VI. University of Tennessee, Knoxville, pp. 39–42. Braun, E.L. (1972) Deciduous Forests of Eastern North America. Hafner, New York. Bruce, D. (1990) Development of empirical forest growth models. In: Dixon, R.K., Meldahl, R.S., Ruark, G.E. and Warren, W.G. (eds) Process Modeling of Forest Growth Responses to Environmental Stress. Timber Press, Portland, Oregon, pp. 191–199. Buchman, R.G. and Shifley, S.R. (1983) Guide to evaluating forest growth projection systems. Journal of Forestry 81, 232–234, 254. Burns, R.M. and Honkala, B.H. (tech. coords.) (1990) Silvics of North America. USDA Forest Service Agriculture Handbook 654, Vol. 2. Callaway, R.M. (1992) Effect of shrubs on recruitment of Quercus douglasii and Quercus lobata in California. Ecology 72, 2118–2128. Callaway, R.M. and D’Antonio, C.M. (1991) Shrub facilitation of coast live oak establishment in central California. Madroño 38, 158–169. Callaway, R.M. and Mahall, B.E. (1996) Variation in leaf structure and function in Quercus douglasii trees differing in root architecture and drought history. International Journal of Plant Sciences 157, 129–135. Callaway, R.M., Nadkarni, N.M. and Mahall, B.E. (1991) Facilitation and interference of Quercus douglasii on understory productivity in central California. Ecology 72, 1484–1499. Canadell, J. and Rodà, F. (1991) Root biomass of Quercus ilex in a montane Mediterranean forest. Canadian Journal of Forest Research 21, 1771–1778. Canham, C.D. (1989) Different responses to gaps among shade-tolerant tree species. Ecology 69, 548–550. Carmean, W.H. (1971) Site index curves for black, white, scarlet and chestnut oaks in the Central States. USDA Forest Service Research Paper NC NC-62. Carvell, K.L. and Tryon, E.H. (1961) The effect of environmental factors on the abundance of oak regeneration beneath mature oak stands. Forest Science 7, 98–105. Cecich, R.A. (1991) Seed production in oak. Proceedings of The Oak Resource in the Upper Midwest Conference. University of Minnesota, St Paul, pp. 125–131. Christisen, D.M. and Kearby, W.H. (1984) Mast measurement and production in Missouri (with special reference to acorns). Missouri Department of Conservation Terrestrial Series 13. Cleland, D.T., Hart, J.B., Host, G.E., Pregitzer, K.S. and Ramm, C.W. (1993) Ecological Classification and Inventory System of the Huron–Manistee National Forests. USDA Forest Service, Washington, DC. Connell, J.H. and Slatyer, R.O. (1977) Mechanisms of succession in natural communities and their role in community stability and organization. American Naturalist 111, 1119–1144. Crow, T.R. (1988) Reproductive mode and mechanisms for self-replacement of northern red oak (Quercus rubra) – a review. Forest Science 34, 19–40. Crow, T.R. (1992) Population dynamics and growth patterns for a cohort of northern red oak (Quercus rubra) seedlings. Oecologia 91, 192–200. Curtis, J.T. (1959) The Vegetation of Wisconsin. University of Wisconsin Press, Madison. Danielsen, K.C. and Halvorson, W.L. (1991) Valley oak seedling growth associated with selected grass species. USDA Forest Service General Technical Report PSW PSW-126, pp. 9–13. Davis, F.W., Borchert, M., Harvey, L.E. and Michaelsen, J.C. (1991) Factors affecting seedling survivorship of blue oak (Quercus douglasii H. & A.) in central California. USDA Forest Service General Technical Report PSW PSW-126, pp. 81–86. Day, G.M. (1953) The Indian as an ecological factor in the northeastern forest. Ecology 34, 329–346. DeSelm, H.R., Clebsch, E.E.C. and Bennie, J.C. (1991) Effects of 27 years of prescribed fire on an oak forest and its soils in middle Tennessee. USDA Forest Service General Technical Report SE SE-70, Vol. 1, pp. 409–417. Dey, D.C. (1991) A comprehensive Ozark regenerator. PhD dissertation, University of Missouri, Columbia. Dey, D.C. and Parker, W.C. (1996) Regeneration of red oak (Quercus rubra L.) using shelterwood systems: ecophysiology, silviculture and management recommendations. Ontario Forest Research Institute Forest Research Information Paper 126. Dey, D.C., Johnson, P.S. and Garrett, H.E. (1996a) Modeling the regeneration of oak stands in the Missouri Ozark Highlands. Canadian Journal of Forest Research 26, 573–583.


161

Dey, D.C., Ter-Mikaelian, M., Johnson, P.S. and Shifley, S.R. (1996b) Users’ guide to ACORn: a comprehensive Ozark regeneration simulator. USDA Forest Service General Technical Report NC NC-180. Dixon, R.K., Meldahl, R.S., Ruark, G.A. and Warren, W.G. (eds) (1990) Process Modeling of Forest Growth Responses to Environmental Stress. Timber Press, Portland, Oregon. Downs, A.A. (1944) Estimating acorn crops for wildlife in the southern Appalachians. Journal of Wildlife Management 8, 339–340. Dunn, C.P., Guntenspergen, G.R. and Dorney, J.R. (1983) Catastrophic wind disturbance in an oldgrowth hemlock–hardwood forest, Wisconsin. Canadian Journal of Botany 61, 211–217. Egler, F.E. (1954) Vegetation science concepts I. Initial floristic composition. A factor in old-field vegetation management. Vegetatio 4, 412–417. Ehrenfeld, J.G. (1980) Understory response to canopy gaps of varying size in a mature oak forest. Bulletin of the Torrey Botanical Club 107, 29–41. Farmer, R.E., Jr (1975) Growth and assimilation rate of juvenile northern red oak: effects of light and temperature. Forest Science 21, 373–381. Fischer, B.C. (1979) Managing light in the selection method. Proceedings 1979 J.S. Wright Forestry Conference. Purdue University, West Lafayette, Indiana, pp. 43–53. Franklin, J.F., Perry, D.A., Schowalter, T.D., Harmon, M.E., McKee, A. and Spies, T.A. (1989) Importance of ecological diversity in maintaining long-term site productivity. In: Perry, D.A., Meurisse, R., Thomas, B. et al. (eds) Maintaining the Long-term Productivity of Pacific Northwest Forest Ecosystems. Timber Press, Portland, Oregon, pp. 82–97. Galford, J.R., Peacock, J.W. and Wright, S.L. (1988) Insects and other pests affecting oak regeneration. Society of American Foresters Publication 88-03, pp. 219–225. Galford, J., Auchmoody, L.R., Smith, H.C. and Walter, R.S. (1991) Insects affecting establishment of northern red oak seedlings in central Pennsylvania. USDA Forest Service General Technical Report NE NE-148, pp. 271–280. Gatto, M. and Rinaldi, S. (1987) Some models of catastrophic behaviour in exploited forests. Vegetatio 69, 213–222. Gingrich, S.F. (1967) Measuring and evaluating stocking and stand density in upland hardwood forests in the Central States. Forest Science 13, 38–53. Glenn-Lewin, D.C. and van der Maarel, E. (1992) Patterns and processes of vegetation dynamics. In: Glenn Lewin, D.C., Peet, R.K. and Veblen, T.T. (eds) Plant Succession, Theory and Prediction. Chapman & Hall, New York, pp. 11–59. Godman, R.M., Yawney, H.W. and Tubbs, C.H. (1990) Acer saccharum Marsh. Sugar maple. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 78–91. Golden, M.S. and Loewenstein, E.F. (1991) Regeneration of tree species 7 years after clearcutting in a river bottom in central Alabama. USDA Forest Service General Technical Report SE SE-70, Vol. 1, pp. 76–83. Gordon, D.R., Welker, J.M., Menke, J.W. and Rice, K.J. (1989) Competition for soil water between annual plants and blue oak (Quercus douglasii) seedlings. Oecologia 79, 533–541. Gottschalk, K.W. (1985) Effects of shading on growth and development of northern red oak, black oak, black cherry, and red maple seedlings. I. Height, diameter, and root/shoot ratio. Proceedings Central Hardwood Forest Conference V. University of Illinois, Urbana–Champaign, pp. 189–195. Gottschalk, K.W. (1987) Effects of shading on growth and development of northern red oak, black oak, black cherry, and red maple seedlings. II. Biomass partitioning and prediction. Central Hardwood Forest Conference VI. University of Tennessee, Knoxville, pp. 99–110. Gottschalk, K.W. (1990) Gypsy moth effects on mast production. Proceedings Workshop on Southern Mast Management. University of Tennessee, Knoxville, pp. 42–50. Grimm, E.C. (1984) Fire and other factors controlling the Big Woods vegetation of Minnesota in the mid-nineteenth century. Ecological Monographs 54, 291–311. Grubb, P.J. (1977) The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biological Reviews 52, 107–145. Guyette, R.P. and Cutter, B.E. (1991) Tree-ring analysis of fire history of a post oak savanna in the Missouri Ozarks. Natural Areas Journal 11(2), 93–99. Guyette, R.P. and Dey, D.C. (1995) A history of fire, disturbance, and growth in a red oak stand in the Bancroft District, Ontario. Ontario Forest Research Institute Forest Research Information Paper, Paper 119.

162

Chapter 3

Haney, A. and Apfelbaum, S.I. (1990) Structure and dynamics of Midwest oak savannas. In: Sweeney, J.M. (ed.) Management of Dynamic Ecosystems. The Wildlife Society, North Central Section, West Lafayette, Indiana, pp. 19–30. Hannah, P.R. (1987) Regeneration methods for oaks. Northern Journal of Applied Forestry 4, 97–101. Hanson, P.J. and Dixon, R.K. (1985) Allelopathic inhibition of northern red oak by interrupted fern and goldenrod. Central Hardwood Forest Conference V. University of Illinois, UrbanaChampaign, pp. 269–274. Hanson, P.J., Isebrands, J.G. and Dickson, R.E. (1987) Carbon budgets of Quercus rubra L. seedlings at selected stages of growth: Influence of light. Proceedings of the Central Hardwood Forest Conference VI. University of Tennessee, Knoxville, pp. 269–276. Harmon, M.E. (1984) Survival of trees after low-intensity surface fires in Great Smoky Mountains National Park. Ecology 65, 796–802. Harper, K.T., Wagstaff, F.J. and Kunzler, L.M. (1985) Biology and management of the Gambel oak vegetative type: a literature review. USDA Forest Service General Technical Report INT INT-179. Harrison, J.S. and Werner, P.A. (1984) Colonization by oak seedlings into a heterogeneous successional habitat. Canadian Journal of Botany 62, 559–563. Hengst, G.E. and Dawson, J.O. (1993) Bark thermal properties of selected central hardwood species. USDA Forest Service General Technical Report NC NC-161, pp. 55–75. Hilt, D.E. (1985) OAKSIM: an individual-tree growth and yield simulator for managed, even-aged, upland oak stands. USDA Forest Service Research Paper NE NE-562. Hix, D.M. (1988) Multifactor classification and analysis of upland hardwood forest ecosystems of the Kickapoo River watershed, southwestern Wisconsin. Canadian Journal of Forest Research 18, 1405–1415. Hodges, J.D. and Gardiner, E.S. (1993) Ecology and physiology of oak regeneration. USDA Forest Service General Technical Report SE SE-84, pp. 54–65. Horsley, S.B. (1991) Using Roundup and Oust to control interfering understories in Allegheny hardwood stands. USDA Forest Service General Technical Report NE NE-148, pp. 281–290. Host, G.E. and Pregitzer, K.S. (1991) Ecological species groups for upland forest ecosystems of northwestern Lower Michigan. Forest Ecology and Management 43, 87–102. Huddle, J.A. and Pallardy, S.G. (1996) Effects of long-term annual periodic burning on tree survival and growth in a Missouri Ozark oak–hickory forest. Forest Ecology and Management 82, 1–9. Isebrands, J.G., Rauscher, H.M., Crow, T.R. and Dickmann, D.I. (1990) Whole-tree growth process models based on structural-functional relationships. In: Dixon, R.K., Meldahl, R.S., Ruark, G.A. and Warren, W.G. (eds) Process Modeling of Forest Growth Responses to Environmental Stress. Timber Press, Portland, Oregon, pp. 96–112. Jarvis, P.G. (1964) Interference by Deschampsia flexuosa (L.) Trin. Oikos 15, 56–78. Johnson, P.S. (1966) Vegetational developments after logging in southern Michigan xeric oak forests. PhD dissertation, Michigan State University, East Lansing. Johnson, P.S. (1974) Survival and growth of northern red oak seedlings following a prescribed burn. USDA Forest Service Research Note NC NC-177. Johnson, P.S. (1975) Growth and structural development of red oak sprout clumps. Forest Science 21, 413–418. Johnson, P.S. (1976) Modal development of regeneration in clearcut red oak stands in the Driftless Area. Proceedings of the Central Hardwood Forest Conference I. University of Southern Illinois, Carbondale, pp. 455–475. Johnson, P.S. (1977) Predicting oak stump sprouting and sprout development in the Missouri Ozarks. USDA Forest Service Research Paper NC NC-149. Johnson, P.S. (1979) Shoot elongation of black oak and white oak sprouts. Canadian Journal of Forest Research 9, 489–494. Johnson, P.S. (1992) Oak overstory/reproduction relations in two xeric ecosystems in Michigan. Forest Ecology and Management 48, 233–248. Johnson, P.S. (1993) Perspectives on the ecology and silviculture of oak-dominated forests in the central and eastern states. USDA Forest Service General Technical Report NC NC-153. Johnson, P.S. and Rogers, R. (1984) Predicting 25th-year diameters of thinned stump sprouts of northern red oak. Journal of Forestry 82, 616–619. Johnson, P.S., Jacobs, R.D., Martin, A.J. and Godell, E.D. (1989) Regenerating northern red oak: three successful case histories. Northern Journal of Applied Forestry 6, 174–178.


163

Johnson, R.L. (1975) Natural regeneration and development of Nuttall oak and associated species. USDA Forest Service Research Paper SO SO-104. Johnson, R.L. (1981) Oak seeding – it can work. Southern Journal of Applied Forestry 5, 28–33. Johnson, R.L. and Deen, R.T. (1993) Prediction of oak regeneration in bottomland forests. USDA Forest Service General Technical Report SE SE-84, pp. 146–155. Johnson, R.L. and Krinard, R.M. (1983) Regeneration in small and large sawtimber sweetgum–red oak stands following selection and seed tree harvest: 23-year results. Southern Journal of Applied Forestry 7, 176–184. Johnson, W.C. and Webb, T. III. (1989) The role of blue jays (Cyanocitta cristata L.) in the postglacial dispersal of fagaceous trees in eastern North America. Journal of Biogeography 16, 561–571. Jørgensen, S.E. (1990) Ecosystem theory, ecological buffer capacity, uncertainty and complexity. Ecological Modelling 52, 125–133. Keeley, J.E. (1991) Seed germination and life history syndromes in the California chaparral. Botanical Review 57, 81–116. Keeley, J.E. (1992) Recruitment of seedlings and vegetative sprouts in unburned chaparral. Ecology 73, 1194–1208. Kimmins, J.P. (1987) Forest Ecology. Macmillan, New York. Kolb, T.E. and Steiner, K.C. (1990) Growth and biomass partitioning response of northern red oak genotypes to shading and grass root competition. Forest Science 36, 293–303. Korstian, C.F. (1927) Factors controlling germination and early survival in oaks. Yale University School Forestry Bulletin 19. Kotar, J., Kovach, J.A. and Locey, C.T. (1988) Field Guide to Forest Habitat Types of Northern Wisconsin. University of Wisconsin and Wisconsin Department of Natural Resources, Madison. Krajicek, J.E. (1960) Some factors affecting oak and black walnut reproduction. Iowa State Journal of Science 34, 631–634. Lamson, N.I. (1988) Role of stump sprouts in regenerating Appalachian hardwood stands. Society of American Foresters Publication 88-03, 31–37. Landsberg, J.J. (1986) Physiological Ecology of Forest Production. Academic Press, New York. Langdon, O.G. (1981) Some effects of prescribed fire on understory vegetation in loblolly pine stands. Symposium Proceedings: Prescribed Fire and Wildlife in Southern Forests. Clemson University, Charleston, South Carolina, pp. 143–153. Latham, R.E. (1992) Co-occurring tree species change rank in seedling performance with resources varied experimentally. Ecology 73, 2129–2144. Levins, R. (1966) The strategy of model building in population biology. American Scientist 54, 421–431. Lewis, R.W. (1990) Theories, speculation, and the structure of knowledge. Speculations in Science and Technology 13, 1–17. Liming, F.G. and Johnston, J.P. (1944) Reproduction in oak–hickory forest stands of the Missouri Ozarks. Journal of Forestry 42, 175–180. Loach, K. (1967) Shade tolerance in tree seedlings. 1. Leaf photosynthesis and respiration in plants raised under artificial shade. New Phytologist 66, 607–621. Loewenstein, E.F. (1992) Relationships between stem origin and growth characteristics of three bottomland hardwood species on a seven-year-old clearcut area in the Coastal Plain of Alabama. MS Thesis, Auburn University, Auburn, Alabama. Loewenstein, E.F. and Golden, M.S. (1995) Establishment of water oak is not dependent on advance reproduction. USDA Forest Service General Technical Report SRS SRS-1, pp. 443–446. Loftis, D.L. (1983) Regenerating southern Appalachian mixed hardwood stands with the shelterwood method. Southern Journal of Applied Forestry 7, 212–217. Loftis, D.L. (1988a) Regenerating oaks on high-quality sites, an update. Society of American Foresters Publication 88-03, pp. 199–209. Loftis, D.L. (1988b) Regenerating red oak in the southern Appalachians: predictive models and practical applications. PhD dissertation, North Carolina State University, Raleigh. Loftis, D.L. (1990a) Predicting post-harvest performance of advance red oak reproduction in the southern Appalachians. Forest Science 36, 908–916. Loftis, D.L. (1990b) A shelterwood method for regenerating red oak in the southern Appalachians. Forest Science 36, 917–929. Loomis, R.M. (1973) Estimating fire-caused mortality and injury in oak–hickory forests. USDA Forest Service Research Paper NC NC-94.

164

Chapter 3

Loomis, R.M. (1974) Predicting the losses in sawtimber volume and quality from fires in oak–hickory forests. USDA Forest Service Research Paper NC NC-104. Lorimer, C.G. (1981) Survival and growth of understory trees in oak forests of the Hudson Highlands, New York. Canadian Journal of Forest Research 11, 689–695. Lorimer, C.G. (1983) Eighty-year development of northern red oak after partial cutting in a mixedspecies Wisconsin forest. Forest Science 29, 371–383. Lorimer, C.G. (1989) The oak regeneration problem: new evidence on causes and possible solutions. University of Wisconsin Forest Resource Analyses No. 8. Lorimer, C.G. (1993) Causes of the oak regeneration problem. USDA Forest Service General Technical Report SE SE-84, pp. 14–39. Lorimer, C.G., Chapman, J.W. and Lambert W.D. (1994) Tall understorey vegetation as a factor in the poor development of oak seedlings beneath mature stands. Journal of Ecology 82, 227–237. Margalef, R. (1963) On certain unifying principles in ecology. American Naturalist 97, 357–374. Marks, P.L. (1974) The role of pin cherry (Prunus serotina L.) in the maintenance of stability in northern hardwood ecosystems. Ecological Monographs 44, 73–88. Marquis, D.A. and Johnson, R.L. (1989) Silviculture of eastern hardwoods. USDA Forest Service General Technical Report WO WO-55, pp. 9–15. Marquis, D.A., Ernst, R.L. and Stout, S.L. (1992) USDA Forest Service General Technical Report NE NE-96. Marquis, D.A., Eckert, P.L. and Roach, B.A. (1976) Acorn weevils, rodents, and deer all contribute to oak-regeneration difficulties in Pennsylvania. USDA Forest Service Research Paper NE NE-356. Martin, R.E. (1962) Thermal properties of bark. Forest Products Journal 13, 419–426. Maslen, P. (1989) Response of immature oaks to prescribed fire in the North Carolina Piedmont. USDA Forest Service General Technical Report SO SO-74, pp. 259–266. Matney, T.G. (1974) Oak regeneration in natural upland oak stands of West Virginia. MS thesis, West Virginia University, Morgantown. Matsuda, K. and McBride, J.R. (1986) Difference in seedling growth morphology as a factor in the distribution of three oaks in Central California. Madroño 33, 207–216. Matsuda, K., McBride, J.R. and Kimura, M. (1989) Seedling growth form of oaks. Annals of Botany 64, 439–446. McClaran, M.P. and Bartolome, J.W. (1989) Fire-related recruitment in stagnant Quercus douglasii populations. Canadian Journal of Forest Research 19, 580–585. McGee, C.E. (1987) Clearcutting in upland hardwoods: panacea or anathema? Proceedings of the Central Hardwood Forest Conference VI. University of Tennessee, Knoxville, pp. 21–29. McGee, C.E. (1988) Spring weather, canopy removal, and early budbreak threaten oak seedlings. Journal of the Elisha Mitchell Science Society 104, 108–115. McGill, D.W., Rogers, R., Martin, A.J. and Johnson, P.S. (1999) Measuring stocking in northern red oak stands in Wisconsin. Northern Journal of Applied Forestry 16, 144–150. McLemore, B.F. (1990) Cornus florida L. flowering dogwood. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 278–283. McQuilkin, R.A. (1983) Silvical factors affecting the regeneration of oaks and associated species in Pennsylvania. Proceedings of Regenerating Hardwood Stands. Pennsylvania State University, University Park, pp. 37–65. Merz, R.W. and Boyce, S.G. (1956) Age of oak ‘seedlings’. Journal of Forestry 54, 774–775. Minckler, L.S. and Woerheide, J.D. (1965) Reproduction of hardwoods 10 years after cutting as affected by site and opening size. Journal of Forestry 63, 103–107. Muller, C.H. (1951) The significance of vegetative reproduction in Quercus. Madroño 11, 129–137. Myers, S.A. (1978) Insect impact on acorn production in Missouri upland forests. PhD dissertation, University of Missouri, Columbia. Nowacki, G.J., Abrams, M.D. and Lorimer, C.G. (1990) Composition, structure, and historical development of northern red oak stands along an edaphic gradient in north-central Wisconsin. Forest Science 36, 276–292. Nyland, R.D., Abrahamson, L.P. and Adams, K.B. (1983) Use of prescribed fire for regenerating red and white oak in New York. Proceedings of the 1982 Society of American Foresters National Convention, pp. 163–167. Olson, D.F., Jr (1974) Quercus L. Oak. USDA Forestry Services Agriculture Handbook 450, 88pp.


165

Overton, R.P. (1990) Oxydendron arboreum (L.) DC Sourwood. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 497–500. Pallardy, S.G. and Rhoads, J.L. (1993) Morphological adaptations to drought in seedlings of deciduous angiosperms. Canadian Journal of Forest Research 23, 1766–1774. Pavlik, B.M., Muick, P.C., Johnson, S. and Popper, M. (1991) Oaks of California. Cachuma Press, Los Olivos, California. Platt, W.J., Glitzenstein, J.S. and Streng, D.R. (1991) Evaluating pyrogenicity and its effects on vegetation in longleaf pine savannas. Proceedings 17th Tall Timbers Fire Ecology Conference, pp. 143–161. Powell, D.S. (1976) Sprouting ability of advance reproduction of undisturbed forest stands in West Virginia. MS thesis, West Virginia University, Morgantown. Prentice, I.C. and Helmisaari, H. (1991) Silvics of north European trees: compilation, comparisons and implications for forest succession modelling. Forest Ecology and Management 42, 79–93. Rebertus, A.J., Williamson, G.B. and Moser, E.B. (1989) Longleaf pine pyrogenicity and turkey oak mortality in Florida xeric sandhills. Ecology 70, 60–70. Reich, P.B., Abrams, M.D., Ellsworth, D.S., Kruger, E.L. and Tabone, T.J. (1990) Fire affects ecophysiology and community dynamics of central Wisconsin oak forest regeneration. Ecology 71, 2179–2190. Renton, J.J., Lanasa, M.J. and Tryon, E.H. (1974) Radiography for observing wood features. Journal of Forestry 72, 292–293. Rice, C. and Struve, D.K. (1997) Seedling growth form and water use of selected oak species. Diversity and Adaptation in Oak Species, Proceedings of Working Party 2.08.05, Oct. 12–17, 1997, Genetics of Quercus, of the International Union of Forest Research Organizations. Pennsylvania State University, University Park 2, pp. 269–278. Riegel, G.M., Smith, B.G. and Franklin, J.F. (1992) Foothill oak woodlands of the Interior Valleys of southwestern Oregon. Northwest Science 66(2), 66–76. Roach, B.A. and Gingrich, S.F. (1968) Even-aged silviculture for upland central hardwoods. USDA Forest Service Agriculture Handbook 335. Rogers, R. and Johnson, P.S. (1998) Approaches to modeling natural regeneration in oak-dominated forests. Forest Ecology and Management 106, 45–54. Rogers, R., Johnson, P.S. and Loftis, D.L. (1993) An overview of oak silviculture in the United States: the past, present, and future. Annales des Sciences Forestieres 50, 535–542. Ross, M.S., Sharik, T.L. and Smith, D.W. (1986) Oak regeneration after clear felling in southwest Virginia. Forest Science 32, 157–169. Roth, E.R. and Sleeth, B. (1939) Butt rot in unburned sprout oak stands. USDA Technical Bulletin 684. Rouse, C. (1986) Fire effects in northeastern forests: oak. USDA Forest Service General Technical Report NC NC-105. Sander, I.L. (1971) Height growth of new oak sprouts depends on size of advance reproduction. Journal of Forestry 69, 809–811. Sander, I.L. (1979a) Regenerating oaks. Proceedings National Silviculture Workshop. USDA Forest Service, Washington, DC, pp. 212–221. Sander, I.L. (1979b) Regenerating oaks with the shelterwood system. Proceedings of 1979 J.S. Wright Forestry Conference. Purdue University, West Lafayette, IN, pp. 54–60. Sander, I.L. (1983) Oak regeneration in the Central States. Proceedings of the 11th Annual Hardwood Symposium of the Hardwood Research Council, pp. 18–31. Sander, I.L. (1990) Quercus rubra L. Northern red oak. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 727–733. Sander, I.L. and Clark, F.B. (1971) Reproduction of upland hardwood forests in the Central States. USDA Forest Service Agriculture Handbook 405. Sander, I.L., Johnson, P.S. and Rogers, R. (1984) Evaluating oak advance reproduction in the Missouri Ozarks. USDA Forest Service Research Paper NC NC-251. Scholz, H.F. (1955) Growth of northern red oak seedlings under variable conditions of ground cover competition. USDA Forest Service Lake States Forest Experiment Station Technical Note 430. Scholz, H.F. (1959) Further observations on seedbed scarification show benefits to northern red oak were temporary. USDA Forest Service Lake States Forest Experiment Station Technical Note 555.

166

Chapter 3

Sharpe, P.J.H. (1990) Forest modeling approaches: compromises between generality and precision. In: Dixon, R.K., Meldahl, R.S., Ruark, G.A. and Warren, W.G. (eds) Process Modeling of Forest Growth Responses to Environmental Stress. Timber Press, Portland, Oregon, pp. 180–190. Shugart, H.H. (1984) A Theory of Forest Dynamics. Springer-Verlag, New York. Smalley, G.W. (1978) Classification and evaluation of forest sites for the Interior Highlands. Proceedings of the Central Hardwood Forestry Conference II. Purdue University, West Lafayette, IN, 257 pp. Smalley, G.W. (1984) Classification and evaluation of forest sites in the Cumberland Mountains. USDA Forest Service General Technical Report SO SO-50. Smith, D.M. (1986) The Practice of Silviculture, 8th edn. John Wiley & Sons, New York. Smith, H.C. and Miller, G.W. (1987) Managing Appalachian hardwood stands using four regeneration practices – 34-year results. Northern Journal of Applied Forestry 4, 180–185. Smith, L.L. and Vankat, J.L. (1991) Communities and tree seedling distribution in Quercus rubra- and Prunus serotina-dominated forests in southwestern Pennsylvania. American Midlands Naturalist 126, 294–307. Sokal, R.R. and Rohlf, F.J. (1969) Biometry: the Principles and Practices of Statistics in Biological Research. W.H. Freeman, San Francisco. Sork, V.L. (1984) Examination of seed dispersal and survival in red oak, Quercus rubra (Fagaceae), using metal-tagged acorns. Ecology 65, 1020–1022. Sork, V.L., Stacey, P. and Averett, J.E. (1983) Utilization of red oak acorns in non-bumper crop year. Oecologia 59, 49–53. Spurr, S.H. and Barnes, B.V. (1980) Forest Ecology, 3rd edn. Ronald Press, New York. Steiner, K.C., Abrams, M.D. and Bowersox, T.W. (1993) Advance reproduction and other stand characteristics in Pennsylvania and French stands of northern red oak. USDA Forest Service General Technical Report NC NC-161, pp. 473–483. Streng, D.R., Glitzenstein, J.S. and Harcombe, P.A. (1989) Woody seedling dynamics in an east Texas floodplain forest. Ecological Monographs 59, 177–204. Sullivan, N.H. (2001) An algorithm for a landscape model of mast production. PhD dissertation, University of Missouri, Columbia. Terborgh, J. (1985) The vertical component of plant species diversity in temperate tropical forests. American Naturalist 126, 760–776. Thilenius, J.F. (1968) The Quercus garryana forests of the Willamette Valley, Oregon. Ecology 49, 1124–1133. Thor, E. and Nichols, G.M. (1973) Some effects of fires on litter, soil, and hardwood regeneration. Proceedings of the 13th Tall Timbers Fire Ecology Conference, pp. 317–329. Tiedemann, A.R., Clary, W.P. and Barbour, R.J. (1987) Underground systems of Gambel oak (Quercus gambelii) in central Utah. American Journal of Botany 74, 1065–1071. Trimble, G.R., Jr (1960) Relative diameter growth rates of five upland oaks in West Virginia. Journal of Forestry 58, 111–115. Trimble, G.R., Jr (1973) The regeneration of Central Appalachian hardwoods with emphasis on the effects of site quality and harvesting practice. USDA Forest Service Research Paper NE NE-282. Trimble, G.R., Jr and Tryon, E.H. (1966) Crown encroachment into openings cut in Appalachian hardwood stands. Journal of Forestry 64, 104–108. Tryon, E.H. and Carvell, K.L. (1958) Regeneration under oak stands. West Virginia University Agricultural Experiment Station Bulletin 424T. Urban, D.L. and Shugart, H.H. (1992) Individual-based models of forest succession. In: Glenn-Lewin, D.C., Peet, R.K. and Veblen, C.C. (eds) Plant Succession, Theory and Prediction. Chapman & Hall, New York, pp. 249–292. Van Lear, D.H. and Watt, J.M. (1993) The role of fire in oak regeneration. USDA Forest Service General Technical Report SE SE-84, pp. 6–78. Veblen, T.T. (1992) Regeneration dynamics. In: Glenn-Lewin, D.C., Peet, R.K. and Veblen, T.T. (eds) Plant Succession, Theory and Prediction. Chapman and Hall, New York, pp. 152–187. Waldrop, T.A., Buckner, E.R., Shugart, H.H., Jr and McGee, C.E. (1986) FORCAT: a single tree model of stand development following clearcutting on the Cumberland Plateau. Forest Science 32, 297–317. Walters, M.B., Kruger, E.L. and Reich, P.B. (1993) Growth, biomass distribution and CO2 exchange of northern hardwood seedlings in high and low light: relationships with successional status and shade tolerance. Oecologia 94, 7–16.


167

Walters, R.S. (1990) Site quality, fire, and herbicide effects on establishment, growth, and development of regeneration three years after partial cutting of oak stands. PhD dissertation, State University of New York, Syracuse. Ward, J.S. and Stephens, G.R. (1994) Crown class transition rates of maturing northern red oak (Quercus rubra L.). Forest Science 40, 221–237. Watt, A.S. (1947) Pattern and process in the plant community. Journal of Ecology 35, 1–22. Weigel, D.R. and Johnson, P.S. (1998) Stump sprouting probabilities for Indiana oaks. USDA Forest Service Technical Brief TB-NC TB-NC–7. Welker, J.M. and Menke, J.W. (1990) The influence of simulated browsing on tissue water relations, growth and survival of Quercus douglasii (Hook and Arn.) seedlings under slow and rapid rates of soil drought. Functional Ecology 4, 807–817. Wendel, G.W. (1975) Stump sprout growth and quality of several Appalachian hardwood species after clearcutting. USDA Forest Service Research Paper NE NE-329. Wendel, G.W. (1977) Longevity of black cherry, wild grape, and sassafras seed in the forest floor. USDA Forest Service Research Paper NE NE-375. Wendel, G.W. (1990) Prunus pensylvanica L. f. Pin cherry. USDA Forest Service Agriculture Handbook 654, Vol. 2, pp. 587–593. Whitney, G.G. (1986) A demographic analysis of Rubus idaeus and Rubus pubescens. Canadian Journal of Botany 64, 2916–2921. Whitney, G.G. (1994) From Coastal Wilderness to Fruited Plain. Cambridge University Press, Cambridge, UK. Williams, K., Davis, S.D., Gartner, B.L. and Karlsson, S. (1991) Factors limiting the establishment of a chaparral oak, Quercus durata Jeps., in grassland. USDA Forest Service General Technical Report PSW PSW-126, pp. 70–73. Williamson, G.B. and Black, E.M. (1981) High temperature of forest fires under pines as a selective advantage over oaks. Nature 293, 643–644. Will-Wolf, S. (1991) Role of fire in maintaining oaks in mesic oak maple forests. Proceedings of the Oak Resource in the Upper Midwest Conference. University of Minnesota, St Paul, pp. 27–33. Wright, S.L. (1987) Managing insects affecting oak regeneration by prescribed burning. USDA Forest Service General Technical Report SE SE-46, pp. 186–192.

4 Site Productivity

Introduction Silviculturists use the term site to refer to an area of forested land that is qualitatively characterized by its climate, soil, vegetation, or quantitatively by its productivity. The latter is usually expressed as potential wood production per unit land area per unit of time (Helms, 1998). Climate, soil, vegetation and productivity characteristics may be used singly or in combination to define site productivity. More generally, site productivity refers to the ability of a defined area to produce objects and their attributes that are required for living organisms or human society to function (Lee, 1989). Lee (1989) used this definition to refer to ‘resource site quality’. The researcher’s view of site quality may differ from the land manager’s, and Gholz (1988) pointed out that there is no universally accepted definition. The two prevalent schools of thought nevertheless consider site productivity as either: (i) a hypothetical, optimal or potential level of productivity, or (ii) an index or relative measure of actual productivity. Although production capability is sometimes called ‘yield’, production and yield are not the same (Smith, 1986). Yield usually refers to material that is usable. In the case of timber, yield refers to the amount of wood that is actually harvested and removed from the site. In contrast, production usually refers to all material that has resulted from tree growth whether or not it is harvested and removed (Zahner and Myers, 1984). 168

Both production and yield on a given site are affected by forest conditions. For a given species or species mix, conditions that commonly reduce stand yields below their maximum potential include low stocking, insect or disease damage, high grading and tree defects. Both the actual and potential production and yield are of interest to silviculturists. Site productivity influences stand development not only in its effect on growth and yield, but also on regeneration and other ecological processes. Site also influences stand development such as rates of change in numbers of trees, basal area and average tree diameter. For a given age, undisturbed stands on good sites have fewer trees, but higher basal areas and average stem diameters than do stands on poor sites. This results from the faster growth of trees on good sites. For example, oaks on poor sites (site index 50) in the Piedmont were almost 4 inches smaller in mean dbh at stand age 50 (6.7 inches dbh) than oaks growing on good sites (site index 90) (10.4 inches dbh). The poor-site stands also carried lower basal areas (94 ft2 acre1 vs. 113 ft2 acre1) and twice the number of trees (900 acre−1 vs. 450 acre−1). The amount of clear lumber that develops in an oak largely depends on its rate of height growth and thus on site quality. Rapid height growth increases clear bole length and minimizes the size of the knotty core – and thus increases tree value (Carmean and Boyce, 1973; see also Chapter 10). Site quality also affects other wood properties including hardness, shrinkage, strength and yield of cellulose (Zahner, 1970).

Site Productivity

Measures of Site Productivity Site productivity can be measured and expressed in various ways. Silviculturists commonly estimate the periodic increment of wood or timber production expressed as volume or weight increase over a year, a decade or a rotation. Units of measure include various volumetric measures1 and weight measures in pounds and tons. In this context, site productivity can be defined as ‘the productive capacity of a site, usually expressed as volume production of a given species’ (Society of American Foresters, 1995). The term ‘site quality’ is generally used when productivity is expressed as qualitative classes (e.g. poor, medium, good), or on a relative scale. Ecologists view forest productivity somewhat differently. They are often interested in accumulated biomass (dry weight of organic matter) per unit of land area and rates of increase and decrease of biomass per unit area. They also may separate biomass by plant parts such as bolewood, branches, bark, leaves, buds, roots and reproductive structures (Cannell, 1982; Kozlowski et al., 1991). Changes in biomass or weight of forest trees are used to study growth, nutrient cycling and energy flow in forests. An understanding of ecosystem production dynamics is important to both ecologists and silviculturists and consequently has been the subject of numerous workshops and symposia (e.g. Hennessey et al., 1986; Cole and Gessel, 1988). Components of productivity of interest to ecologists include: ● Gross primary productivity (GPP). GPP is the increase per unit area in dry weight of organic material produced by photosynthesis that remains in the plant plus the weight of dry matter lost by plant respiration. ● Net primary productivity (NPP). NPP is the increase per unit area in the sum of three component measures: (i) the 1Volumetric

169

increase in standing biomass, including leaves, stems, roots and reproductive structures; (ii) litterfall; and (iii) the amount of biomass consumed by animals and microbial decomposers. Annual above-ground net primary production (ANPP) may be the most robust measure of productivity because it includes virtually all commercially harvested material as a subset (although in some cases portions of root systems, or components of below-ground net primary productivity, also are harvested) (Gholz, 1988). ● Net ecosystem production (NEP). NEP is GPP minus loss of dry matter due to heterotrophic respiration of microbes and other non-photosynthetic organisms (Kozlowski et al., 1991). These components of productivity vary greatly among different kinds of oak forests (Table 4.1). The ratio of below-ground to aboveground biomass of individual oaks, and thus the spatial distribution of net production, varies among ecosystems. In Mediterranean climates, the roots of multiple-stemmed oaks of coppice origin may comprise 90% or more of total tree biomass, whereas the roots of single-stemmed (non-coppice) oaks typically make up less than 30% of tree biomass (Whittaker and Woodwell, 1968; Canadell and Rodà, 1991). For singlestemmed Holm oaks growing in Mediterranean Spain, root : shoot ratio was higher on dry sites than on mesic sites based on roots 0.4 inch and larger in diameter (Fig. 4.1). This suggests that site quality, itself, can affect the ratio of below-ground to above-ground biomass in oaks. The collective evidence indicates that the ratio of above-ground to below-ground biomass in oak forests depends on several factors including the reproductive origin of oaks and related disturbance history, stand or tree age, site quality and species composition.

units commonly used in the United States include board feet, cords and cunits. One board foot (bf) is a piece of sawn wood 1 inch (2.54 cm) thick, 12 inches (30.48 cm) wide and 12 inches long. One cord is a stacked pile of wood contained within a space measuring 4 ft (1.2 m) deep, 4 ft high and 8 ft (2.4 m) long, which equals 128 ft3 (11.9 m3). One cunit equals 100 cubic feet (2.8 m3).

170

Chapter 4

Table 4.1. Above-ground biomass and net productivity of trees in selected oak forests.a Biomassb (with foliage) Forest type, location

Net productivityb (with foliage)

Stand age (years)

Trees per hectarec

(t ha1)d

(t ha1)d

serrata/crispata oak, Japang

272.7 [276.9]

8.95 [13.75]

–

561

Post/blackjack oak, Oklahomah

174.4 [179.2]

7.33 [10.59]

80

English/sessile oak with beech, Francei

167.4 [171.0]

7.16 [10.76]

66

Mixed oak/hickory, Tennesseej

133.2 [137.9]

5.4 [10.1]

30–80

–

Northern pin oak, Minnesotak

120.8 [124.2]

5.26 [8.77]

45–50

Northern red oak, Great Smoky Mt, USAl

132.2 [135.0]

4.7 [7.5]

(mature)

Sessile/European turkey oak, Hungarym

199.5 [202.9]

3.00 [6.71

65–68

9.0 [9.4]

0.39 [1.67]

117

Mexican blue/Emory oak, Arizonan

Tree height (m)e

Basal area (m2 ha1)f

15.1

39.8

2600

–

18.3

958

22

32.1

12–25

25.8

1788

c. 15

26.5

2660

14

24.6

17.4

15.1

5.3

4.0

–

190

a

From summaries in Cannell, 1982. Includes boles, branches and bark of trees; does not include fruits, woody litterfall or understorey vegetation. c Minimum diameters (dbh) included in calculations vary among studies; 1 ha = 2.471 acres. d 1 tonne ha1 = 0.446 English ton acre1. e One metre = 3.281 ft. f 1 m2 ha1 = 4.356 ft2 acre1. g Katagiri and Tsutsumi, 1975, 1976, 1978. h Johnson and Risser, 1974. i Kestemont, 1971 (plantation). j Harris et al., 1973; Harris and Henderson, 1981. Species include chestnut, white, northern red and black oaks. k Reiners, 1972. l Whittaker, 1963. m Jakucs, 1981. n Whittaker and Niering, 1975. b

As we might expect, the NPP of oakdominated ecosystems in regions with Mediterranean or desert climates is much lower than those in more humid regions. In dry regions, oaks and other woody plants are often restricted in form to small trees or

shrubs. In a Mexican blue oak/Emory oak desert shrub community in Arizona, estimated above-ground NPP was 0.4 t ha1 year1 (Whittaker and Niering, 1975). This rate is about 3% of that of the aboveground component of a typical oak forest

Site Productivity

171

Dbh (inches)

Biomass of large roots (kg)

100

2.4

3.9

5.5

7.1

8.7

18

22

80

60

40 Xeric Mesic 20

0 6

10

14 Dbh (cm)

Fig. 4.1. Biomass of large diameter roots (≥1 cm (0.4 inch)) of single-stemmed Holm oak in northeastern Spain estimated from dbh measured 1.3 m (4.3 ft) above ground. Mean root biomass of oaks on the two sites differed significantly (P = 0.022) based on analysis of covariance. Log10 (mesic site biomass) = –1.393 + 2.451.log10 (dbh), R 2 = 0.81, n = 20; log10 (xeric site biomass) = –0.448 + 1.734.log10 (dbh), R 2 = 0.71, n = 12. For both equations, root biomass is in kilograms and dbh is in centimetres. (From Canadell and Rodà, 1991, used with permission.)

in eastern United States. However, the roots and associated structures (lignotubers) of oaks growing in such climates typically comprise 65–85% of total plant mass (Rundel, 1980). Much of the production in these ecosystems therefore may lie below ground. The accurate measurement of the biomass of roots in trees is difficult, especially in oaks, which have taproots that can grow to great soil depths. Moreover, it is even more difficult to obtain accurate measurements of root NPP because of the rapid turnover of fine roots. Reported results therefore tend to underestimate NPP of roots (Cannell, 1982). Economists often express forest site productivity as the capacity of an area to produce financial return through timber production. Financial return can be measured in several ways including net present value, internal rate of return or soil expectation value. Such economic measures of productivity often include assumptions about the initial state of the forest, duration

of the economic evaluation period, inflation rate, discount rate, management practices, management expenses and the timing of periodic expenses and income. Consequently, financial return can be estimated with great detail and associated complexity. Nevertheless, relatively simple measures such as per cent value increase or per cent volume increase are often useful for comparing site productivity among stands of similar initial condition. Because economic return is derived from forest volume, estimating site productivity in terms of yield is usually an intermediate step in estimating economic site productivity.

Relation of Site Productivity to Ecological Classification Determination of site productivity occurs within the broader context of regional factors that influence productivity. For example, climate and geology limit the upper

172

Chapter 4

range of oak site productivity. Likewise, localized topographic features such as slope steepness, slope position and aspect influence productivity. Experienced silviculturists often take for granted that these factors are related to site productivity and mentally account for them when assessing a site qualitatively. Ecological classification systems (ECS) provide a hierarchical framework for placing individual sites within a regional context. ECS provides a means for grouping similar ecosystems based on relations between the characteristics of living organisms and physical features of a site. The final product is a set of defined ecological groupings of plants and environmental factors that are repeated across the landscape at various spatial scales ranging from single stands to large regions (Chapter 1). However an ecologist’s perspective of a forest site is apt to be different from a silviculturist’s. Ecologists are likely to consider a site as a relatively uniform geographical unit characterized by certain stable combinations of physical and climatic factors, whereas foresters generally view a site as a land unit characterized by a specified productive capacity of timber or other forest products. Despite their differences, these views are complementary (Schonau, 1988). An ecological classification system provides information on the ecological context of a site. This, in turn, helps the silviculturist not only to evaluate site productivity, but to better understand and predict the response of the forest to silvicultural treatments. Likewise, localized quantitative estimates of site productivity (e.g. site index) can guide the ecologist in defining the spatial extent of ecological classification units. Killian (1984) proposed that ecological classification precede the determination of site productivity. He reasoned that the goal of site classification should be to identify the possibilities and risks to forest management and to predict potential yields, thereby assisting both short- and long-term local and regional forest planning as well as general land-use planning. That objective is more user orientated than that usu-

ally encompassed by ecological site classification, which some believe should focus on interrelations among ecosystem components, i.e. climate, physiography–geology, soils and their biota, and vegetation (Barnes, 1984). Regardless of viewpoint, ecological classification and site quality are closely linked. Whereas the term site productivity emphasizes the factors that influence tree growth, ecological classification emphasizes the factors that determine the abundance of species and the occurrence of natural groupings of plant species in relation to environment. Moreover, ecological classification usually considers both woody and herbaceous vegetation in defining ecological units. Herbaceous vegetation is particularly important because of the relatively large number of herbaceous species present and the ‘fidelity’ of some of those species to the factors that distinguish one ecological classification unit from another. The presence and/or abundance of many herbaceous species is often less affected by disturbance factors and associated changes in the successional status of a forest than are trees and other woody species. There are potential advantages of ecological classification over traditional site classification. In their development, the latter focuses narrowly on forest productivity whereas the former produces natural ecological groupings that can be used to identify ecological units within which productivity, succession, tree regeneration, and responses to silvicultural and natural disturbances are likely to be similar (Chapter 1). However, site productivity estimates that are based solely on ecological classification units are likely to be less accurate than those based on direct determination of site productivity at a specific location within the classification unit.

Productivity and Related Selfsustaining Properties of Oak Forests Silvicultural practices in oak forests of the United States have been, by agricultural standards, extensive rather than intensive.

Site Productivity

With few exceptions, oak silviculture has followed an ‘ecological’ model based on managing natural vegetation and plant propagules in place. Unlike the ‘agronomic’ model, new genetic material (e.g. genetically ‘improved’ trees) is seldom introduced; herbicides and fertilizers are used only to a limited extent if at all. Although there are exceptions, the usual objective is to control stand composition, structure, growth and quality largely through timber cutting practices. This approach has been and continues to be largely driven by the economics of oak timber production, which is characterized by low returns on investment (e.g. Dwyer et al., 1993). Even when potential economic returns from more intensive silvicultural practices are deemed acceptable, investment in such practices may be discouraged by long deferrals on returns plus associated risks of ‘crop’ damage or loss from insects, disease, drought, fire and other uncertainties, one or more of which have a high likelihood of occurrence over the relatively long 80- to 100-year production periods usually required for oaks. Moreover, there is growing social demand to manage publically owned forests for a wide range of products and values that transcend narrow timber production objectives. The control of stand composition, structure and density by cutting (timber harvesting) methods alone is not unique to oak forests, but characterizes hardwood silviculture in the United States in general. Although we have considerable knowledge of how to apply a more ‘agronomic’ or culturally intensive model to hardwood silviculture, such methods have seldom been applied outside of research studies. Contemporary oak silviculture therefore lies along the boundary of ‘natural’ and ‘unnatural’ (i.e. human-influenced) ecological systems. The resulting silviculture is therefore heavily dependent on natural ecological processes that are especially significant with respect to forest productivity and its sustainability. Forests, whether influenced by humans or not, are endowed with certain attributes that ensure a high capacity for self-sustaining productive

173

capacity (e.g. as expressed by ANPP). What effect, then, does the removal of trees from the forest have on forest site productivity?

Effects of timber harvesting on site productivity The site factors most vulnerable to irreversible change are those associated with the soil. Careless removal of trees from a site can cause soil erosion and nutrient losses. Erosion is the physical loss of soil particles caused by wind and water action. Loss of nutrients also accompanies the physical loss of soil. Because the upper soil layers, which are richest in soil nutrients, are eroded first and most severely, erosion can have a major impact on soil fertility. However, the loss of nutrients also can occur in the absence of the physical loss of soil particles when nutrient ions become dissolved in surface runoff, are lost in nutrient solutions (leachates) that percolate through and from the soil, or are lost through the removal of vegetation. Erosion is nevertheless a normal geologic process that results in some soil losses even in undisturbed old-growth forests. The estimated prehistoric rate of erosion in forested parts of southern Michigan, an area comprised of extensive oak forests, was estimated at 0.05 t acre1 year1 (Davis, 1970). In the Appalachian Mountains of West Virginia, annual soil erosion losses from hardwood forests range from 0.05 to 0.10 t acre1 year1 for both undisturbed and clearcut forests (Patric, 1976). This compares to ‘acceptable’ rates of soil loss from agricultural lands that range from 1 to 5 t acre1 year1 (Patric, 1977). Moreover, cropland is usually cultivated annually, whereas a managed oak forest is usually logged less frequently than one year in ten. The view that the absence or reduction of tree cover, by itself, causes soil erosion in forests of the eastern United States is largely unfounded (Patric 1976, 1978; Mills et al., 1987). However, to understand this issue, it is important to distinguish between effects related purely to the tem-

174

Chapter 4

porary absence of the forest canopy following timber harvesting, and associated effects related to skid trails and logging road construction. Soil and site effects associated with the removal of trees, per se, are related to the characteristics of forest soils. Four attributes of forest soils are especially important in moderating soil erosion and soil nutrient losses: (i) the forest floor, (ii) soil structure, (iii) soil infiltration rate, and (iv) the dynamic nature of the biomass of forest soils. In eastern deciduous forests, the surface of the forest floor consists of a layer of undecomposed or partially decomposed leaf litter and other organic debris. The litter layer provides protection against the kinetic energy of rainfall, preventing particle detachment and sealing of pores of the underlying mineral soil. In oak forests, this layer is replenished every year with 1–2 tons per acre of leaf fall, plus another 0.5 ton of woody litter fall (Rodin and Basilevich, 1968; Rochow, 1974). Forest cover thus maintains soil porosity by continually returning leaf litter and woody debris to the forest floor, which in turn is continually decomposing and being incorporated into the soil by soil organisms. Forest soils consequently develop and maintain a physical structure and macropore space associated with high soil porosity. High porosity, in turn, is associated with high infiltration rates and low surface runoff. Infiltration rates of 50 or more inches per hour are common in forest soils in the eastern United States, whereas rainfall intensities rarely exceed 2 inches per hour (Patric, 1978). When the water storage capacity of the soil is reached after prolonged precipitation, the water moves laterally through the pores of the soil to streams. The removal of trees from a site, by itself, does not destroy soil porosity provided the site is quickly revegetated. Soil porosity nevertheless can be greatly reduced by logging road construction and

log skidding practices that scrape, gouge or compact the surface layers of the soil. The risk of accelerated soil erosion from a managed forest thus depends largely on logging practices rather than on the temporary absence or reduction of tree cover. The burst of shrub, herb and sprout growth that occurs after overstorey removal tends to mitigate effects that might otherwise result in soil and site degradation. This flush of growth buffers disturbed forests against nutrient losses and facilitates their recovery. For example, 5–6 years after clearcutting a northern red oak stand in southwestern Wisconsin, the estimated biomass of fine roots2 was 70% larger than in an adjacent uncut stand (Yin et al., 1989). Most of this biomass was comprised of shrub and herbaceous species. In this case, not only was root biomass maintained, but temporarily increased after timber harvesting. Such responses ensure protection against both soil and nutrient losses during periods when trees are reestablishing their dominance. Timber harvesting also does not necessarily alter the amount of litter on the floor of oak forests (Carmean, 1959). Net losses of nutrients from a site nevertheless can occur when nearly all vegetation is removed and nutrients move out of the system at rates higher than they are replaced (Bormann and Likens, 1979). Such losses may occur when entire trees (boles, branches and foliage) are removed to a central location within a stand for chipping, and the nutrient-rich branches and foliage are piled along the roadside. Windrowing, i.e. pushing slash into long narrow piles, is a site preparation technique that also can remove nutrients from large fractions of the forest. If slash piles are burned, the resulting localized areas of intense heat may produce significant losses of nitrogen. Pushing slash into piles or windrows before burning often is accompanied by the redistribution of topsoil and minerals, which in turn may decrease site

2 In this study, fine roots were defined as those of 2–10 cm in diameter. The biomass of these roots, which were largely concentrated within the upper 30 cm (12 inches) of soil, fluctuated with the seasons.

Site Productivity

quality over a portion of the stand. In highly porous soils in regions of high rainfall and warm temperatures that are already low in organic matter, timber harvesting may accelerate nutrient losses from leaching, especially nitrogen. The important issue, however, may not be the amount of mineral nutrients removed from the site, per se, but how that amount compares with that available in the soil for forest regrowth after timber harvesting. This amount depends on initial soil fertility, the relative amounts of minerals in the ecosystem that occur in the soil, forest floor and above-ground biomass, plus the rate at which the soil mineral pool is replenished by decomposition of slash, nutrient deposition from the atmosphere, mineral weathering and nitrogen fixation by soil organisms. Logic would seem to indicate that, because a major portion of a forest’s biomass is in trees, timber harvesting would remove a proportionately large amount of nutrients. On the contrary, nutrient losses from timber harvesting are relatively low. For example, in an oak–hickory forest in Tennessee, the proportion of nutrients in trees was relatively small compared to that present in the entire ecosystem even though trees accounted for 67% of the biomass. Trees comprised 6% of the nitrogen, 2% of the phosphorus, 1% of the potassium and 16% of the calcium (Binkley, 1986). If only tree boles are considered, the percentages are even smaller: 1% or less of nitrogen, phosphorus and potassium, and 11% of calcium. Compared to harvesting only bolewood, whole-tree harvesting removes two to three times more nitrogen, phosphorus and potassium, and about one-third more calcium than does harvesting only bolewood. Whole-tree removal is also potentially more serious in infertile soils, where nutrient deficiencies already exist. Nutrient losses from whole-tree harvesting and the time required for their replacement, and the economics of wholetree harvesting are discussed in more detail by others (e.g. Waring and Schlesinger, 1985; Binkley, 1986).

175

Modifying site productivity through fertilization Although the nutrient capital of an oak forest is sufficient to sustain tree growth at some given rate, few forest soils provide nutrient levels that are optimal for tree growth (Smith, 1986; Kozlowski, et al., 1991). Suboptimal levels of nutrients in forest soils can be caused by poor land-use practices that preceded the establishment of an existing forest, or low natural soil fertility. The nutrients that are below optimal levels under oak stands are usually the same elements found in most commercial fertilizers: nitrogen, phosphorus and potassium. Of these, nitrogen is often the most important. Nitrogen is naturally added to soils by rainfall and can be fixed from the atmosphere by some bacteria and other organisms in the soil and in the root nodules of some forest plants. Although oak forests often have an abundant supply of nitrogen, much of it is unavailable for plant growth at any given time because it is tied up in organic matter. Unfortunately, any excess nitrogen does not accumulate, but is changed by bacteria to nitrate (NO3) and then leached from the soil by rainwater, or used by denitrifying bacteria as a source of oxygen. Phosphorus is sometimes deficient in southeastern oak forests growing on highly leached soils, wet soils and very sandy soils, and potassium is occasionally deficient in highly leached sandy soils. When fertilizers are applied to oak forests, trees often respond better to the application of other nutrients when they are applied with nitrogen. For example, calcium fertilization of poor sites in Pennsylvania increased stand volume growth of oaks by 10%, whereas nitrogen and calcium together increased growth by more than 40% (Ward and Bowersox, 1970). Application rates for nitrogen and other nutrients for oak forests should be matched to soil nutrient characteristics of the site. Although the height growth response of oaks to fertilization is uncertain, diameter growth can be increased by more than 30% (Ward and Bowersox, 1970; Graney, 1987). Responses to fertilization may last for 6 years, but largely disappear after 10 years.

176

Chapter 4

Methods of Evaluating Site Quality Site index In North America, site index is the most commonly used method of expressing forest site quality. Site index is defined as the average height of dominant or both dominant and co-dominant trees at a standard, or index, age. Tree height growth of dominant and co-dominant trees in even-aged, fully stocked forest stands is closely related to volume growth. Site index therefore is an indirect measure of site productivity because, by itself, it does not directly express productivity. It is nevertheless useful as an index of productivity because of its correlation with productivity and relative ease of measurement. Most timber yield tables for even-aged oak stands report yields by site index classes (Table 4.2). Similarly, growth and yield models often include site index as one of the variables used to predict yield (see Chapter 10). Height growth (and thus tree height at a specified age) is a useful indirect measure of site quality because it is relatively independent of stand density. Height growth of Table 4.2. Gross yields per acre by site index classes for normala even-aged oak stands in southwestern Wisconsin.b Stand age (years)

Site index (ft at base age 50) 45

55

65

gross yield (ft3)c 20 40 60 80 100 120 140 160 a

480 1050 1550 2000 2350 2650 2900 3050

850 1750 2550 3300 3900 4350 4750 4950

1150 2400 3600 4600 5500 6200 6700 7000

Normal oak stands are relatively undisturbed stands at average maximum density (Chapter 6). b From Gevorkiantz and Scholz, 1948. c Based on gross volume, excluding bark, of all trees 0.6 inch dbh and larger including tops and limbs suitable for cordwood. Species are predominantly black, white and northern red oaks.

dominant and co-dominant trees is reduced only at the extremes of stand crowding (Carmean, 1975; Lloyd and Jones, 1983; Lanner, 1985; Jones, 1986). Site index can be determined directly from observations of trees growing on the site, or it can be estimated indirectly from physical site characteristics. Direct determination of site index Direct determination of site index requires knowing the heights and ages of dominant and co-dominant trees. This information then is referenced to a set of hypothetical height growth curves, called site indexcurves, which are indexed to a common age (Fig. 4.2). For oak species in the United States, the usual index age is 50 years. A site index of 65 thus indicates that the dominant and co-dominant trees on a site attain an average height of 65 ft at 50 years. A set of site index curves therefore represents patterns of height growth of dominant and co-dominant trees for different sites. Their purpose is to relate observed heights of trees of any age to their expected heights at age 50. Any factors that cause actual tree height growth to differ from the patterns expressed in the curves used to determine site index will introduce error into the estimated site index. Consequently, it is important to ensure that the site index curves utilized are appropriate for the site being evaluated. For example, three sets of site index curves applicable to white oak differ somewhat for site index 70. Graphic comparison of these curves illustrate that the closer the observed tree age is to the index age, the less discrepancy in site index there is among the curves (Fig. 4.3). However, when site index estimates are based on trees younger than 30 years or older than 70 years, differences due to choice of site index curves can be substantial (Carmean, 1979). Accurately determining site index requires: (i) the presence of trees that are reliable indicators of site quality, and (ii) the availability of suitable site index curves

Site Productivity

140

177

100 90

120

80 70 60 80 50 40

60

Site index (ft)

Tree height (ft)

100

30 40 20 0 20

40

60 80 Tree age (years)

100

120

Fig. 4.2. Site index curves (index age 50) for black oak in the unglaciated uplands of southeastern Ohio, eastern Kentucky, southern Indiana and southern Missouri. Dashed-line curves represent values beyond the observed range. (From Carmean, 1971, 1972.)

120

100

Tree height (ft)

80

60

40 Carmean et al. (1989) Graney and Bower (1971)

20

Schnur (1937) 0 20

40

60

80

100

120

Tree age (years) Fig. 4.3. Site index 70 curves for white oak (Carmean, 1971; Graney and Bower, 1971), and a composite curve (applicable to white oak) for upland oaks in the eastern United States (Schnur, 1937). Although all three curves have an identical site index of 70 ft at index age 50, the height–age curves diverge elsewhere. Differences in estimated site index resulting from application of different curves can be substantial, especially when observed tree age differs from the index age by more than 20 years.

178

Chapter 4

that accurately characterize the height growth pattern of the trees observed (Carmean et al., 1989). Suitable trees for determining site index include free-growing, uninjured dominant and co-dominant trees. Such trees most frequently occur in even-aged, fully stocked stands that have not been high-graded by logging, heavily grazed, or otherwise damaged. To determine site index, heights and ages of dominant and co-dominant trees must be ascertained. Tree ages can be obtained from increment cores that reveal annual rings as well as patterns of diameter growth. Trees that are reliable indicators of site index form annual rings during early life that are wide and even. Trees with narrow rings formed in early life indicate suppression of growth and therefore should not be used to determine site index. Trees with forks and other bole defects may indicate earlier top breakage or dieback and therefore may be unreliable indicators of site index. Trees selected for site index determinations also should be from well-stocked even-aged stands comprised of dominant and co-dominant trees whose ages do not differ by more than 10 years. Greater differences can produce highly variable patterns of height growth, which in turn may cause large errors in site index estimation. For example, even-aged upland oak stands in the Missouri Ozarks sometimes include dominant and co-dominant trees appreciably younger or older than most of the trees in those crown classes, i.e. the predominant age class. Because young oaks tend to grow faster in height, they will indicate a higher site index than oaks in the predominant age class if a common age is assumed. Conversely, appreciably older oaks grow more slowly in height and will indicate a lower sites index than trees in the predominant age class if a common age is assumed (McQuilkin, 1975). Site index determined from individual dominant and co-dominant trees may vary even on small plots that appear to be homogeneous in soil and microsite. This variation in site index can produce problems in determining the number of site index trees required to obtain reliable site

index estimates (Carmean, 1975; Lloyd and Hafley, 1977; Lloyd, 1981). The requisite number of site trees depends on several factors, including the desired accuracy of the site index estimates, size of the stand and variation in site index. Sampling trees for site index determination thus is more complicated in large tracts with highly variable stand and site conditions. Ideally, large tracts should be stratified into smaller areas that are relatively homogeneous topographically, edaphically and ecologically. Site index then should be determined within each stratum. Even then, large errors in estimating site index can occur when site index curves are used outside their intended area of application. The height growth of oak coppice stands may differ markedly from seedling-origin stands. This is especially true during the early years of stand development when oak coppice benefits the most from the relatively large parent-tree root systems they are connected to. Large roots and the concomitant capacity of young coppice shoots to produce multiple long flushes (Fig. 2.20) together with clump density effects (Chapter 2) may mask the early expression of site quality effects. Site effects may not be apparent until stands are 20–25 years old. By age 50, the productive capacity of coppice stands in the South Carolina Piedmont was well differentiated. On the best sites (oak site index 90), volumes and above-ground biomass were nearly twice that of stands growing on the poorest sites (oak site index 50) (Zahner and Myers, 1984). The height growth curves for these coppice stands differed substantially from the conventional site index curves developed for upland oaks in the Piedmont region of the southeastern United States (Olson, 1959). The latter overestimated site index by as much as 18 ft when applied to coppice stands (Zahner et al., 1982). Site index curves thus were developed specifically for coppice-origin stands of that region (Fig. 4.4). Although stand site index is usually expressed as the average site index calculated from qualified sample trees and reported to the nearest foot, greater preci-

90

90

80

80

70

70

60

60

50

50

40

40

Site index (ft)

Tree height (ft)

Site Productivity

30 20 10

20

30

40

50

Tree age (years) Fig. 4.4. Site index curves (index age 50) for young mixed oak stands of sprout origin in the South Carolina Piedmont. (From Zahner et al., 1982, used with permission.)

sion should not be attributed to the estimate than is statistically justified. The precision of site index estimates made from site index equations was determined for white, black and scarlet oaks in Missouri. Estimates for trees 20 years younger to 20 years older than index age were precise to within 3–4 ft with 95% confidence when based on ten sample trees. Similar results were observed for oaks (chestnut, white, northern red, black and scarlet) in northwestern West Virginia (Lamson, 1980). Errors in estimating site indices from site index equations arise from three sources: (i) errors in estimating individual sample tree site indices from site index/height regressions; (ii) variation among sample tree heights within even-aged stands; and (iii) measurement errors (McQuilkin and Rogers, 1978). Methods of collecting data for deriving site index curves have evolved greatly since the first site index curves were published. The earliest curves were based on only a few plots with total height and age measurements taken from a few selected trees. These height and age data were used to calculate an average ‘guiding curve’ representing the average height growth pattern of a given species. Graphical or propor-

179

tional methods then were used to produce a set of anamorphic curves, within which each curve has the same shape. Anamorphic curves assume that the pattern of height growth is similar for all levels of site quality, and for all climates, soils and topographic conditions within the region of intended application (Carmean, 1970). This assumption is inappropriate for oaks because the shape of the height curve often varies with site quality (Carmean, 1970, 1975). More recently developed site index curves are based on the assumption that the shape of tree height growth curves can differ by site classes. Such curves can be derived through stem analysis, which requires that sample trees be felled and cut into sections to determine the progression of height growth over time. Stem analysis combined with non-linear regression analysis produces polymorphic site index curves. This is now the most widely used method for developing site index curves for species like the oaks that express polymorphic patterns of growth. For the oaks, polymorphic site index curves have largely replaced anamorphic curves. Methods for constructing site index curves have been described by others (Burkhart et al., 1981; Clutter et al., 1983; Borders et al., 1984; Biging, 1985; Avery and Burkhart, 1994). A comprehensive compilation of 127 site index curves for species in the eastern United States, including 21 sets of curves for oaks, is presented by Carmean and others (1989). Each set of curves is based on an expanded form of the Chapman–Richards non-linear function (Ek, 1971; Payandeh, 1974a, 1974b; Monserud and Ek, 1976). One variant of the model estimates height from site index and age whereas the other estimates site index from height and age. Site index comparisons among species Dominant and co-dominant trees of different species growing on the same site are likely to have different height growth patterns. For example, a comparison of site index curves for white, scarlet and red oaks

180

Chapter 4

able for the major upland oaks and associated species in the eastern United States (Fig. 4.6; Appendices 4–6). Site index for a given species is not necessarily uniformly higher or lower than for another species. For example, on good sites the yellow-poplar site index (index age 50) may be as much as 30 ft greater than that of white oak. However, as site quality decreases the difference between yellowpoplar and white oak site indexes also decreases. At site indices below 70 ft, the white oak site index may exceed that of yellow-poplar (Olson and Della-Bianca, 1959; Carmean and Hahn, 1983).

at a common site index of 70 shows that white oak height growth lags behind the other species for the first 40 years. But by age 80, white oak height surpasses the other oaks (Fig. 4.5). Consequently, separate site index curves are generally required for each species. This is evident from species comparisons in the Missouri Ozarks. There, the observed site index for white oak on a given site is about 4 ft less than that of black oak; the scarlet oak site index is about 3 ft greater than that of black oak on the same site (McQuilkin, 1974). The co-occurrence of species across a wide range of sites makes possible the development of equations that can be used to convert the observed site index of one species to the estimated site index of another species. Such site index conversions are useful when the species of interest is absent and a common site index basis for all stands or locations within stands is desired. Such conversions are facilitated by graphs or equations designed for this purpose (Nelson and Beaufait, 1956; Trimble and Weitzman, 1956; Doolittle, 1958; Olson and Della-Bianca, 1959; McQuilkin, 1974; Carmean and Hahn, 1983). They are avail-

Indirect estimation of site index from soil and topographic factors Site index curves cannot be applied to sites where suitable trees are absent or where no trees are present. To solve that problem, equations have been developed in some regions to estimate oak site index from soil, topographic and other factors. These equations typically account for 70–85% of observed variation in site index. However, they are often not used in practical applica-

120

Tree height (ft)

100

80

60

40 White oak N. red and black oaks Scarlet oak

20

0 20

40

60

80

100

120

Tree age (years) Fig. 4.5. Comparison of site index curves for scarlet oak, white oak and red/black oaks in the Central Hardwood Region. The curves shown are for site index 70 ft (index age 50). (From Carmean et al., 1989.)

Site Productivity

181

110 100

Site index (ft)

90 80 70 60 50 40

Black oak White oak Scarlet oak Northern red oak Yellow-poplar

30 Fig. 4.6. Chart for converting the site index (index age 50) of one species to another in even-aged upland oak and yellow-poplar stands in the central states. The site index of species absent from a stand can be estimated from species present by using this nomogram. Site indexes for all species are read from the vertical axis. For example, assume that height and age measurements of several dominant and co-dominant yellow-poplars indicate a site index of 94 for that species. The corresponding site index of another species is read by moving vertically downward from 94 on the yellow-poplar curve to the curve of the species of interest. At that interception point, the unknown site index is read horizontally across on the vertical axis. On this chart, site indexes for scarlet, black, northern red, chestnut and white oak corresponding to yellow-poplar site index 94 are approximately 89, 88, 86 and 83, respectively. Conversions also can be derived from equations (see Tables 4.3, 4.4 and 4.5). (From Carmean and Hahn, 1983, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

tion because they require information on one or more soil characteristics such as texture, horizon or soil thickness, and soil chemical properties that are difficult or inconvenient to measure. Also, many of the equations are derived from observed values of site index derived from anamorphic curves that may not accurately describe tree height growth in the region to which the equation applies. Even though such equations are often adequate for practical applications, they only provide an estimate of site index, which itself is only an indirect measure of forest productivity. Thus, these equations are two steps removed from direct determination of productivity (Leary, 1985). Equations for estimating site index nevertheless can provide insight into the factors that influence the height growth of oaks and therefore oak site productivity. Most of these equations include as predictors one or more topographic factors including slope posi-

tion, aspect, slope gradient and slope shape. Although these factors, by themselves, have no direct effect on tree growth, they are correlated with more directly causative factors. For example, aspect and slope gradient jointly determine the amount of solar radiation received on a slope (Frank and Lee, 1966; Swift, 1976), and thus account for factors more directly related to tree growth such as evapotranspiration, leaf temperature and fluctuation in microclimate. Aspect and slope gradient are correlated with site quality in most oak forests. The better sites usually occur on north to east aspects and the poorer sites on south to west aspects (Carmean, 1965; Hannah, 1968; Hartung and Lloyd, 1969; Graney, 1978; Auchmoody and Smith, 1979). Slope position and slope shape are also related to soil moisture. Lower slopes, for example, usually have higher site indices due to subsurface water flow from upper slopes.

Chapter 4

One predictive site index equation (Auchmoody and Smith, 1979) considers the interactive effects of slope gradient and aspect. In this equation, the effect of aspect on site index is modified by slope gradient. Site quality accordingly decreases as aspect departs from the cool and moist microenvironment in the northeast quadrant at 81° azimuth. At this most favourable azimuth, slope gradient has no effect. But as aspect changes in either direction towards the least favourable aspect at 261° in the southwest quadrant (180° from 81°), the effect of increasing slope gradient is to reduce site index due to associated increases in solar radiation (Fig. 4.7). Black oak in southern Ohio responded similarly to slope and aspect (Carmean, 1965, 1967). However, in that region, the most favourable and least favourable aspects occurred at 45° and 225°, respectively. The effects of aspect may be asymmetrical, i.e. the most favourable and least favourable aspects may not be separated by 180°. In an Ohio study, the observed opti-

mum aspect occurred at 45° azimuth whereas south, southwest and west aspects (180° to 270°) were almost equally unfavourable (Carmean, 1965, 1967).3 The occurrence of the most favourable aspect in the northeast quadrant was associated with thicker litter layers, thicker A1 horizons high in base saturation, less acidic A and upper B horizons, greater available soil nitrogen and thicker colluvial accumulations than soils on southwest aspects. Thus, soil development processes related to tree growth also appear to be correlated with variation in topography and solar radiation. Relations between aspect and observed tree growth may not be entirely consistent with the actual incident radiation received on a slope. Commonly observed relations between topographic factors and tree growth suggest that other factors can modify solar radiation effects. These factors include soil development processes that, although partially dependent on incident solar radiation, do not necessarily change linearly with solar radiation.

70

0 10

Site index (ft)

25 68

% slope

182

50 66

64

62 0

60

81

120

180

261

240

300

360

Aspect (degrees azimuth) Fig. 4.7. Relation between oak site index (index age 50), aspect and four slope gradients in upland forests in northwestern West Virginia. For comparative purposes, site index has been arbitrarily set at 70 for the most favourable aspect (81° azimuth) for each slope gradient. For each gradient, site index is minimized at 261° azimuth. (Adapted from Auchmoody and Smith, 1979). 3 Stage (1976) describes methods for mathematically identifying and specifying interactive aspect and slope gradient effects on tree growth.

Site Productivity

Like topographic factors, soil factors commonly incorporated into site index prediction equations are often associated with variables thought to be more causally related to tree growth. For example, soil texture and stone content are correlated with available moisture. Likewise, soil pH and per cent base saturation are correlated with the availability of nitrogen and other macronutrients, and soil depth or depth to an impenetrable horizon is correlated with the effective volume of the rooting zone. The effectiveness of a given variable as a predictor of productivity may vary among regions, ecosystems within regions and species. Such variability is reflected in the differences among existing sets of site index curves. The development of satisfactory site index estimation equations for hardwood sites on alluvial soils in the South has been shown to be difficult, if not unfeasible (Broadfoot, 1969). The apparent reason is the large number of interacting factors (including fluctuating water tables, aeration, and associated soil rooting space and nutrient availability problems) that influence site productivity in bottomlands. To overcome these problems, Baker and Broadfoot (1979) developed a field guide for classifying sites occupied by, or potentially suitable to, cherrybark, Nuttall, Shumard, water, willow and swamp chestnut oaks. Their method is based on a matrix of soil factors that influence tree growth in southern bottomlands. Application requires evaluating four major soil factors: (i) soil physical condition; (ii) moisture availability during the growing season; (iii) nutrient availability; and (iv) aeration. Each of these factors is further comprised of specific soilsite properties whose qualitative or quantitative properties are subdivided into three relative site quality classes: best, medium and poor. A site quality rating (SQR) is then assigned to each soil-site property and site quality class for each of the six bottomland oak species considered by the method. Site suitability and quality for a given species is then evaluated by summing its SQRs across all soil-site properties considered by the method.

183

Estimating site index from tree height and diameter Intensive forest management requires measures of productivity that are sensitive to site differences while retaining validity over time-dependent changes in the tree crop. Site index attempts to satisfy these requirements by extrapolating measured tree height at a given age to tree height at a reference age. Although site index is the most widely used method for assessing site quality in North America, the method has been criticized (Gevorkiantz and Scholz, 1944; Jones, 1969; Gholz, 1988; Avery and Burkhart, 1994). Potential problems in the application of site index include extrapolation errors, measurement errors and possible insensitivity to production expressed as volume (Stout and Shumway, 1982). It is frequently difficult to obtain accurate tree ages from increment cores and to obtain accurate heights of standing trees. Some trees are difficult to age from an increment core because the pith or growing centre of the tree bole is often missed when it is non-circular (Lamson, 1987). In slow-growing trees, growth rings are often obscure. Variability in ages among trees within the same stand also contributes to uncertain age estimates. Although any one of these factors could introduce large errors into estimating site index, age variability is potentially the most serious limitation of site index as an estimator of volume productivity. Moreover, height alone is used in the site index method, even though tree and stand volume depend on both diameter and height. Foresters nevertheless have been reluctant to discard site index for measures of productivity based on diameter growth, which although sensitive to site quality, is also strongly influenced by stand density. Studies nevertheless have shown that, across a wide range of initial stand densities, yields per unit area tend to converge with time toward a site-specific maximum value (Drew and Flewelling, 1977; Harper, 1977). This convergence is caused by physiological adjustments in the height and diameter growth of trees to changes in available growing space. When stand den-

184

Chapter 4

sity is low, diameter growth increases and height growth decreases, and conversely when stand density is high, diameter growth decreases and height growth increases (Gevorkiantz and Scholz, 1944). Because the two trends are compensatory, they tend to produce approximately the same volume in the average dominant tree. Theoretically, using both diameter and height to estimate site productivity should compensate for variation in stand density. Based on this concept, tree dbh, height and age can be used to place stands in site index classes based on tree ‘volume index’. A tree’s volume index is the product of its basal area (or squared diameter) and height. The volume index method was originally developed for mixed-oak stands in southwestern Wisconsin (Gevorkiantz and Scholz, 1944). However, the method is potentially applicable to any forest type for which site index and average heights and diameters of dominant trees in fully stocked (‘normal’) stands are known for a

range of tree ages. Such information is often provided in conjunction with published site index, stand and yield tables. From that information, site index classes based on volume index can be graphically related to the average age of dominant trees (Fig. 4.8). The volume index of any observed stand then can be referenced to the graph. To reliably estimate a stand’s volume index, at least 25 dominant trees of approximately the same age should be measured. The observed stand’s volume index is calculated by multiplying the average basal area by the average height of the sample trees. The stand’s site index class then can be determined by referencing its volume index to the reference curves. Volume index may be especially useful for estimating site quality in understocked oak stands. However, volume indices may need to be ‘corrected’ by reducing volume index when stocking is below 50% or above 100% of normal (Gevorkiantz and Scholz, 1944).

65 160

Volume index

120

55

80

50

Site index class (ft)

60

40 45 0 20

40

60

80

100

120

140

Average age of dominant oaks (years) Fig. 4.8. Site index classes (index age 50) derived from volume index and age of dominant oaks in mixed-oak stands in southwestern Wisconsin. Classes are represented by areas between the curves. A stand’s volume index is calculated by multiplying the average basal area of dominant trees (ft2) by their average height (ft). For example, by reference to the chart, 90-year-old trees with a volume index of 80 represent a site index class of 55. In this region, the maximum site index for oak is about 70 ft. (From Gevorkiantz and Scholz, 1944, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

Site Productivity

Unfortunately, the volume index method does not eliminate the problem of age variability because it requires an age measurement. There is, however, a method for estimating site quality that, in field application, does not require determining tree age (Stout and Shumway, 1982). The method is based on families of height–diameter curves formulated for site classes by species. Such curves have been derived for white, black and northern red oaks. They were generated by substituting coefficients for S and b (from Appendix 7) into the following equation for tree height in feet (H): H = 4.5 + S(1 – ebD)

[4.1]

where D is dbh in inches, and S and b are site-specific and species-specific coefficients, respectively. Equation 4.1 thus can be used to generate families of species-specific height–diameter curves associated with different site index classes. Site index estimation using this technique requires measuring the heights and diameters of several dominant or co-dominant trees and referencing the paired measurements to height–diameter site index curves (Fig. 4.9). For example, a red oak 90 ft tall with a dbh of 20 inches lies within site class 80 according to Fig. 4.9.

185

Site evaluation alternatives to site index Methods based on soil and physiography Another approach to evaluating oak sites is to place sites into discrete productivity categories or on a continuous site quality scale other than site index. Like soil-site equations for estimating site index, these methods are usually based on soil and topographic factors. A categorical site evaluation method applicable to oak forests in southern Michigan is based on: (i) soil texture; (ii) the presence of moist layers (high water table or fine textured materials) within 4–10 ft of the soil surface; (iii) slope steepness; and (iv) slope position (Gysel and Arend, 1953). High productivity is associated with sites with fine textured subsoils, gentle mid to lower slopes and moisture retentive layers within the upper 4–10 ft of the soil. The associated site evaluation scheme places sites in one of five productivity classes ranging from very poor to very good, which in turn, are associated with the average volume of dominant and co-dominant oaks (Table 4.3).

110 90 80 70

90

60 80 50 70

Site index (ft)

Tree height (ft)

100

60 50 40 6

10

14

18

22

26

30

Dbh (inches) Fig. 4.9. Site index curves (index age 50) for northern red oak in the central Appalachians based on tree height and dbh. Application requires the measurement of several dominant or co-dominant trees in each stand. (Adapted from Stout and Shumway, 1982, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

186

Chapter 4

Table 4.3. Site classification system for upland oak stands in southern Michigan.a

Texture of subsoil Fine (loams, clay loams and clays)

Medium (sandy loams and loamy sands)

Coarse (sands)

Position of moist layers in substrata

Topography

Highc High

Flatd Rollinge

High

Hillyf

Low Low

Flat Rolling

Low

Hilly

High High

Flat Rolling

Low Low

Flat Rolling

Low

Hilly

High High

Flat Rolling

Position on slope

Site class

Average volume of dominant/ co-dominant oaks [ft3 (bd ft)]b

— Upper Middle Lower Bottom Upper Middle Lower

Very good Good Good Good Very good Medium Good Good

54 (210) 35 (130) 35 (130) 35 (130) 54 (210) 24 (70) 35 (130) 35 (130)

— Upper Middle Lower Bottom Upper Middle Lower — Middle

Medium Medium Medium Good Very good Poor Medium Good Good Good

24 (70) 24 (70) 24 (70) 35 (130) 54 (210) 17 (43) 24 (70) 35 (130) 35 (130) 35 (130)

— Upper Middle Lower Bottom Upper Middle Lower — Middle

Very poor Very poor Poor Medium Good Very poor Poor Good Good Good

12 (25) 12 (25) 17 (43) 24 (70) 35 (130) 12 (25) 12 (25) 17 (43) 17 (43) 17 (43)

a From

Gysel and Arend, 1953. Average volume of 80-year-old black oaks and northern red oaks. c For the fine-textured subsoil class, ‘high’ refers to the presence of a fine-textured (clayey) subsoil; for other subsoil classes, ‘high’ refers to the presence of a water table within 4–10 ft of the soil surface. d Slope steepness of 5% or less. e Slopes are of moderate length with relatively broad ridges and valleys. f Slopes are relatively steep with narrow ridges and valleys. b

The topographic site coefficient (TSC) represents a continuously scaled relative measure of forest site productivity applicable to the Driftless Area of southwestern Wisconsin and adjacent southeastern Minnesota, northeastern Iowa and northwestern Illinois (Johnson, 1975). It can be used to assess site productivity on upland soils ranging in texture from sandy loam to silt loam. TSC integrates into a single value the effects of soil depth, slope position and aspect on

soil moisture and therefore on tree growth. The index value is based on soil plus parent material depth to bedrock (from a minimum of 10 to a maximum of 50 inches) weighted by an index of average growing season soil moisture associated with slope position and aspect (azimuth). TSC is scaled from 0.1 (poorest sites) representing sites that occupy southwest-facing upper slopes on thin soils, to 1.0 (best sites) that occupy low northeastfacing slopes on deep soils (Fig. 4.10).

Slope position

Site Productivity

NE N

Aspect SE NW

187

Soil depth (inches) ≤10

S SW

20

30

≥50

40

Lower

Middle Upper

Level

3.0

2.5 2.0 1.5 1.0 [cos(azimuth-45)]+2

0.1

0.2

Poor

0.3 0.4 0.5 0.6 0.7 0.8 Topographic site coefficient Medium

0.9

1.0

Good

Site quality Fig. 4.10. Topographic site coefficient (TSC) in relation to aspect, slope position and soil depth. To find TSC, locate aspect on the transformed azimuth or compass scale. Then proceed upwards until the appropriate slope position is intersected. Use the LEVEL line for level topography and all slopes less than 15%. Next, proceed horizontally to the right until the appropriate soil depth is intersected. Finally, proceed downward and read the corresponding TSC value. (From Johnson and Rogers, 1982, used with permission.)

TSC thus provides an empirical method for assessing forest site quality where it is impossible to obtain site index because of the absence of trees or the lack of suitable trees for its determination. Potential applications include the assessment of site quality in young clearcuts and other harvested areas. TSC also has been used as a predictor of early stand development after timber harvesting (Johnson, 1976). In addition, the method can be used to assess the suitability of forest or non-forest sites for tree planting and for predicting the growth of natural reproduction. For example, the height growth of northern red oak stump sprouts (Fig. 2.28) and planted hardwoods in clearcuts in southwestern Wisconsin is related to TSC (Johnson, 1975; Johnson and Rogers, 1980, 1982, 1984, 1985). Although the exact relation between TSC and site index is unknown, TSC values span the approximate northern red oak site index

range of 45–70 ft based on the site index curves of Gevorkiantz (1957). In addition to the methods mentioned above, information on oak forest productivity in relation to physiography and soil taxa is contained in many county and regional soil survey manuals published by the USDA Natural Resources Conservation Service (formerly USDA Soil Conservation Service). A site classification system for the south-central part of the Central Hardwood Region was developed largely from such information. The classification system is applicable to the northern Cumberland Plateau, central ridge and valley, and interior low plateau highland rim sections, which lie within portions of the broadleaved forest, oceanic province (221a, Fig. 1.2) and the broadleaved forest, continental province (221b, Fig. 1.2) (Smalley, 1979, 1982, 1984, 1986). This system has much in

188

Chapter 4

common with the ecological classification systems discussed in Chapter 1. The system divides the landscape into ‘landtypes’ based on commonalities in geology, physiography, soils and vegetation. Landtype descriptions are provided and each is rated with respect to forest productivity (site index and mean annual growth of forest stands), plant competition, seedling mortality, equipment limitations, erosion and windthrow hazards, and tree species desirability. Biophysical methods Biophysical methods relate primary environmental variables such as air temperature, precipitation, relative humidity and incident solar radiation to the productive potential of forests. In can be argued that such variables represent the underlying forces that are causally related to variation in forest productivity. Accordingly, productivity estimated from such variables would not be subject to the vagaries of changing stand conditions as are the more common empirical indices like site index. Because site index, by definition, is tree height at a specified age, events that alter the pattern of tree height growth may diminish the accuracy and therefore the usefulness of site index. Although foresters assume site index to be constant for a given forest site, certain silvicultural practices such as thinning and fertilizing can affect height growth and therefore the estimate of site index. Biophysical methods of site evaluation attempt to relate forest productivity to factors more directly related to productivity. The biophysical approach is based on the hypothesis that the capability of the land to produce wood depends on the localized characteristics of the continuum representing the movement of water through the soil and plant, and into the atmosphere. Any one of three continuum elements (soil–plant–atmosphere) can potentially limit growth and thus site productivity. The approach assumes that the quantity of soil water consumed in forest transpiration is directly proportional to

wood production (Czarnowski, 1964). For a given species, transpiration rate is dependent on the rate of soil water absorption and the leaf–atmosphere relationships that control water loss. Soil depth, moisture content and texture are factors that determine soil water availability, while variation in soil temperature mediates absorption rates at any given moisture level. However, the driving force for transpiration occurs at the leaf surface and is dependent on vapour pressure differences between leaf and air. For a given ambient vapour pressure, transpiration depends on leaf temperature. Consequently, during periods of intense radiation, even trees growing in moist soils can be severely water stressed if transpiration exceeds absorption. The relation between these two processes thus provides a biophysical basis for explaining differences in site productivity such as those that occur between north- and south-facing slopes. The extent to which these processes are related to (and thus are predictable from) more easily measured soil and topographic factors forms the basis for most other methods of forest site evaluation. For the same species, differences in productivity between two sites with similar soils but different topographies can be substantial. For example, basal area growth of northern red oak in the Appalachian Mountains of West Virginia was more than three times greater on north-facing slopes than on south-facing slopes even though the physical and chemical properties of the soils were similar. The difference in productivity therefore was largely ascribed to differences in radiant and thermal energy regimes (Lee and Sypolt, 1974). Unfortunately, data for relating site productivity to primary environmental variables are difficult to obtain. One difficulty is the measurement of key environmental variables such as relative humidity over sufficiently long periods (greater than 1 year). Moreover, correlations between above-ground net primary production (ANPP) and single or even multiple envi-

Site Productivity

ronmental factors are extremely variable. Because of these problems, it is difficult to estimate the direct effect of environmental variables on carbon allocation and fixation and to model short-term ANPP for develop-

189

ing a site quality measure that has predictive and interpretative power (Gholz, 1988). Examples of such ‘physiological process’ models of ANPP driven by environmental factors are presented by Gholz (1988).

References Auchmoody, L.R. and Smith, H.C. (1979) Oak soil–site relationships in northwestern West Virginia. USDA Forest Service Research Paper NE NE-434. Avery, T.E. and Burkhart, H.E. (1994) Forest Measurements. McGraw-Hill, New York. Baker, J.B. and Broadfoot, W.M. (1979) Site evaluation for commercially important southern hardwoods. USDA Forest Service General Technical Report SO SO-26. Barnes, B.V. (1984) The ecological approach to ecosystem classification. Proceedings of the Symposium Site and Productivity of Fast-growing Plantations (IUFRO), pp. 69–89. Biging, G.S. (1985) Improved estimates of site index curves using a varying-parameter model. Forest Science 31, 248–259. Binkley, D. (1986) Forest Nutrition Management. John Wiley & Sons, New York. Borders, B.E., Bailey, R.L. and Ware, K.S. (1984) Slash pine site index from a polymorphic model by joining (splining) nonpolymorphic segments with an algebra difference method. Forest Science 30, 411–423. Bormann, F.H. and Likens, G.E. (1979) Pattern and Process in a Forested Ecosystem. Springer-Verlag, New York. Broadfoot, W.M. (1969) Problems in relating soil to site index for southern hardwoods. Forest Science 15, 354–364. Burkhart, H.E., Cao, Q.V. and Ware, K.D. (1981) A comparison of growth and yield prediction models for loblolly pine. Virginia Polytechnic Institute and State University, School of Forestry and Wildlife Resources. Canadell, J. and Rodà, F. (1991) Root biomass of Quercus ilex in a Mediterranean forest. Canadian Journal of Forest Research 21, 1771–1778. Cannell, M.G.R. (comp.) (1982) World Forest Biomass and Primary Production Data. Academic Press, London. Carmean, W.H. (1959) Litter weight not reduced following clearcutting of poor-site oak stands. Journal of Forestry 57, 208–209. Carmean, W.H. (1965) Black oak site quality in relation to soil and topography in southeastern Ohio. Soil Science Society of America Proceedings 1965, pp. 308–312. Carmean, W.H. (1967) Soil survey refinements for predicting black oak site quality in southeastern Ohio. Soil Science Society of America Proceedings 1967, pp. 805–810. Carmean, W.H. (1970) Site quality for eastern hardwoods. USDA Forest Service Research Paper NE NE-144, pp. 36–56. Carmean, W.H. (1971) Site index curves for black, white, scarlet and chestnut oaks in the Central States. USDA Forest Service Research Paper NC NC-62. Carmean, W.H. (1972) Site index curves for upland oaks in the Central States. Forest Science 18, 109–120. Carmean, W.H. (1975) Forest site quality evaluation in the United States. Advances in Agronomy 27, 209–269. Carmean, W.H. (1979) Soil-site factors affecting hardwood regeneration and growth. In: Holt, H.A. and Fischer, B.C. (eds) Regenerating Oaks in Upland Hardwood Forests. Purdue University, West Lafayette, Indiana, pp. 61–74. Carmean, W.H. and Boyce, S.G. (1973) Hardwood log quality in relation to site quality. USDA Forest Service Research Paper NC NC-103. Carmean, W.H. and Hahn, J.T. (1983) Site comparisons for upland oaks and yellow-poplar in the Central States. Journal of Forestry 81, 736–739.

190

Chapter 4

Carmean, W.H., Hahn, J.T. and Jacobs, R.D. (1989) Site index curves for forest tree species in the eastern United States. USDA Forest Service General Technical Report NC NC-128. Clutter, J.L., Fortson, J.C., Pienaar, L.V., Brister, G.H. and Bailey, R.L. (1983) Timber Management: A Quantitative Approach. John Wiley & Sons, New York. Cole, D.W. and Gessel, S.P. (1988) Forest Site Evaluation and Long-term Productivity. University of Washington Press, Seattle. Czarnowski, M.S. (1964) Productive Capacity of Locality as a Function of Soil and Climate with Particular Reference to Forest Land. Louisiana State University Press, Baton Rouge. Davis, M.B. (1970) Erosion rates and land use history in southern Michigan. Geological Society of America Abstracts 2, 533. Doolittle, W.T. (1958) Site index of scarlet and black oak in relation to southern Appalachian soil and topography. Forest Science 3, 114–124. Drew, J.T. and Flewelling, J.W. (1977) Some recent Japanese theories of yield–density relationships and their application to Monterey pine plantations. Forest Science 23, 517–534. Dwyer, J.P., Dey, D.C. and Kurtz, W.B. (1993). Profitability of precommercially thinning oak stump sprouts. Northern Journal of Applied Forestry 10, 179–183. Ek, A.R. (1971) A formula for white spruce site index curves. University of Wisconsin Forestry Research Note 161. Frank, E.C. and Lee, R. (1966) Potential solar beam irradiation on slopes: tables for 30° to 50° latitude. USDA Forest Service Rocky Mountain Forest and Range Experiment Station RM RM-18. Gevorkiantz, S.R. (1957) Site index curves for red oak in the Lake States. USDA Forest Service Lake States Forest Experiment Station Technical Note 485. Gevorkiantz, S.R. and Scholz, H.F. (1944) Determining site quality in understocked oak forests. Journal of Forestry 42, 808–811. Gevorkiantz, S.R. and Scholz, H.F. (1948) Timber yields and possible returns from the mixed-oak farmwoods of southwestern Wisconsin. USDA Forest Service Lake States Forest Experiment Station Publication 521. Gholz, H.L. (1988) Problems in the biophysical determination of forest site quality. In: Cole, D.W. and Gessel, S.P. (eds) Forest Site Evaluation and Long Term Productivity. University of Washington Press, Seattle, pp. 12–21. Graney, D.L. (1978) Site quality relationships for the oak–hickory forest type. Proceedings of 1978 Society of American Foresters National Convention, pp. 339–343. Graney, D.L. (1987) Ten-year growth of red and white oak crop trees following thinning and fertilization in the Boston Mountains of Arkansas. USDA Forest Service General Technical Report SO SO-42, pp. 445–449. Graney, D.L. and Bower, D.R. (1971) Site index curves for red and white oaks in the Boston Mountains of Arkansas. USDA Forest Service Research Note SO SO-121. Gysel, L.W. and Arend, J.L. (1953) Oak sites in southern Michigan: their classification and evaluation. Michigan State University Technical Bulletin 236. Hannah P.R. (1968) Estimating site index for white and black oaks in Indiana from soil and topographical factors. Journal of Forestry 66, 412–416. Harper, J.L. (1977) Population Biology of Plants. Academic Press, London. Harris, W.F. and Henderson, G.S. (1981) In: Reichle, D.E. (ed.) Dynamic Properties of Forest Ecosystems. Cambridge University Press, London, pp. 658–661. Harris, W.F., Goldstein, R.A. and Henderson, G.S. (1973) Analysis of forest biomass pools, annual primary production and turnover of biomass for a mixed deciduous forest watershed. In: IUFRO Biomass Studies. University of Maine, College of Life Sciences and Agriculture, Orono, pp. 43–64. Hartung, R.E. and Lloyd, J. (1969) Influence of aspect on forests of the Clarksville soil in Dent County, Missouri. Journal of Forestry 67, 178–182. Helms, J.A. (ed.) (1998) The Dictionary of Forestry. Society of American Foresters, Bethesda, Maryland. Hennessey, T.C., Dougherty, P.M., Kossuth, S.V. and Johnson, J.D. (1986) Stress Physiology and Forest Productivity: Proceedings of the Physiology Working Group of the Society of American Foresters. Kluwer Academic, Boston. Jakucs, P. (1981) In: Reichle, D.E. (ed.) Dynamic Properties of Forest Ecosystems. Cambridge University Press, London, 586pp.

Site Productivity

191

Johnson, F.L. and Risser, P.G. (1974) Biomass, annual net primary production, and dynamics of six mineral elements in a post oak–blackjack oak forest. Ecology 55, 1246–1258. Johnson, P.S. (1975) Growth and structural development of red oak sprout clumps. Forest Science 21, 413–418. Johnson, P.S. (1976) Eight-year performance of interplanted hardwoods in southern Wisconsin oak clearcuts. USDA Forest Service Research Paper NC NC-126. Johnson, P.S. and Rogers, R. (1980) Predicting growth of individual stems within northern red oak sprout clumps. Proceedings of the Central Hardwood Forest Conference III. University of Missouri, Columbia, pp. 420–439. Johnson, P.S. and Rogers, R. (1982) Hardwood interplanting in the upper Mississippi Valley. Proceedings of the Hardwood Regeneration Conference. University of Minnesota, St Paul, pp. 90–109. Johnson, P.S. and Rogers, R. (1984) Predicting 25th-year diameters of thinned stump sprouts of northern red oak. Journal of Forest 82, 616–619. Johnson, P.S. and Rogers, R. (1985) A method for estimating the contribution of planted hardwoods to future stocking. Forest Science 31, 883–891. Jones, E.P., Jr (1986) Slash pine plantation spacing study – age 30. USDA Forest Service General Technical Report SE SE-43, pp. 45–49. Jones, J.R. (1969) Review and comparison of site evaluation methods. USDA Forest Service Research Paper RM RM-51. Katagiri, S. and Tsutsumi, T. (1975) The relationship between site condition and circulation of nutrients in forest ecosystems. III Aboveground biomass and nutrient contents of stands. Journal of the Japanese Forestry Society 57, 412–419. Katagiri, S. and Tsutsumi, T. (1976) The relationship between site condition and circulation of nutrients in forest ecosystems. IV The amount of mineral nutrient returned to forest floor. Journal of the Japanese Forestry Society 58, 79–85. Katagiri, S. and Tsutsumi, T. (1978) The relationship between site condition and circulation of nutrients in forest ecosystems. V The differences in nutrient circulation between stands located in upper part of slope and lower part of slope. Journal of the Japanese Forestry Society 60, 195–202. Kestemont, P. (1971) Productivité primaire des taillis simples et concept de nécromasse. In: Durigneaud, P. (ed) Productivity of Forest Ecosystems. UNESCO, Paris, pp. 271–279. Killian, W. (1984) Site classification and mapping: principles and trends. Proceedings of the IUFRO Symposium Site and Productivity of Fast Growing Plantations 1, pp. 51–68. Kozlowski, T.T., Kramer, P.J. and Pallardy, S.G. (1991) The Physiological Ecology of Woody Plants. Academic Press, San Diego, California. Lamson, N. (1980) Site index prediction tables for oak in northwestern West Virginia. USDA Forest Service General Technical Report NE NE-462. Lamson, N.I. (1987) Estimating northern red oak site-index class from total height and diameter of dominant and codominant trees in central Appalachian hardwood stands. USDA Forest Service Research Paper NE NE-605. Lanner, R.M. (1985) On the insensitivity of height growth to spacing. Forestry Ecology and Management 13, 143–148. Leary, R.A. (1985) Interaction Theory in Forest Ecology and Management. Martinus Nijoff, Dordrecht, Netherlands. Lee, R.G. (1989) The concept and measurement of multi-resource site quality. USDA Forest Service General Technical Report NC NC-135, pp. 32–39. Lee, R. and Sypolt, C.R. (1974) Toward a biophysical evaluation of forest site potential. Forest Science 20, 145–154. Lloyd, F.T. (1981) How many tree heights should you measure for natural Atlantic Coastal Plain loblolly site index? Southern Journal of Applied Forestry 5, 180–183. Lloyd, F.T. and Hafley, W.L. (1977) Precision and the probability of misclassification in site index estimation. Forest Science 23, 493–499. Lloyd, F.T. and Jones, E.P., Jr (1983) Density effects on height growth and its implications for site index prediction and growth projection. USDA Forest Service General Technical Report SE SE24, pp. 329–333. McQuilkin, R.A. (1974) Site index prediction table for black, scarlet, & white oaks in southeastern Missouri. USDA Forest Service Research Paper NC NC-108.

192

Chapter 4

McQuilkin, R.A. (1975) Errors in site index determination caused by tree age variation in even-aged oak stands. USDA Forest Service Research Note NC NC-185. McQuilkin, R.A. and Rogers, R. (1978) A method for determining the precision of site index estimates made from site index prediction functions. Forest Science 24, 289–296. Mills, W.L. Jr, Fischer, B.C. and Reisinger, T.W. (1987) Upland hardwood silviculture: a review of the literature. Purdue University Agricultural Experiment Station Bulletin 527. Monserud, R.A. and Ek, A.R. (1976) Site index curves and equations for several northern hardwood forest species. University of Wisconsin School of Natural Resources Bulletin R2771. Nelson, T.C. and Beaufait, W.R. (1956) Studies in site evaluation for southern hardwoods. Society of American Foresters Proceedings, pp. 67–70. Olson, D.F., Jr (1959) Site index curves for upland oak in the southeast. USDA Forest Service Southeastern Forest Experiment Station Research Note 125. Olson, D.F., Jr and Della-Bianca, L. (1959) Site index comparisons for several tree species in the Virginia–Carolina Piedmont. USDA Forest Service Station Paper SE SE-104. Patric, J.H. (1976) Soil erosion in the eastern forest. Journal of Forestry 74, 671–677. Patric, J.H. (1977) Soil erosion and its control in eastern woodlands. Northern Logger and Timber Processor 25, 4–5, 22–23. Patric, J.H. (1978) Harvesting effects on soil and water in the eastern hardwood forest. Southern Journal of Applied Forestry 2, 66–73. Payandeh, B. (1974a) Formulated site index curves for major timber species in Ontario. Forest Science 20, 143–144. Payandeh, B. (1974b) Nonlinear site index equations for several Canadian timber species. Forestry Chronicle 47, 194–196. Reiners, W.A. (1972) Structure and energetics of three Minnesota forests. Ecological Monographs 42, 71–94. Rochow, J.J. (1974) Litter fall relations in a Missouri forest. Oikos 25, 80–85. Rodin, L.E. and Basilevich (1968) World distribution of plant biomass. In: Eckardt, F.E. (ed.) Functioning of Terrestrial Ecosystems at the Primary Production Level Proceedings of Copenhagen Symposium. UNESCO, Liège, Belgium, pp. 45–52. Rundel, P.W. (1980) Adaptations of Mediterranean-climate oaks to environmental stress. USDA Forest Service General Technical Report PSW PSW-44, pp. 43–54. Schnur, G.L. (1937) Yield, stand, and volume tables for even-aged upland oak forests. USDA Technical Bulletin 560. Schonau, A.P.G. (1988) Problems in using vegetation or soil classification in determining forest site quality. In: Cole, D.W. and Gessel, S.P. (eds) Forest Site Evaluation and Long Term Productivity. University of Washington Press, Seattle, pp. 3–11. Smalley, G.W. (1979) Classification and evaluation of forest sites on the southern Cumberland Plateau. USDA Forest Service General Technical Report SO SO-23. Smalley, G.W. (1982) Classification and evaluation of forest sites on the Mid-Cumberland Plateau. USDA Forest Service General Technical Report SO SO-38. Smalley, G.W. (1984) Classification and evaluation of forest sites in the Cumberland Mountains. USDA Forest Service General Technical Report SO SO-50. Smalley, G.W. (1986) Classification and evaluation of forest sites on the Northern Cumberland Plateau. USDA Forest Service General Technical Report SO SO-60. Smith, D.M. (1986) The Practice of Silviculture, 8th edn. John Wiley & Sons, New York. Society of American Foresters (1995) Silviculture Terminology. Society of American Foresters, Bethesda, Maryland. Stage, A.R. (1976) An expression for the effect of aspect, slope, and habitat type on tree growth. Forest Science 22, 457–460. Stout, B.B. and Shumway, D.L. (1982) Site quality estimation using height and diameter. Forest Science 28, 639–645. Swift, L.W., Jr (1976) Algorithm for solar radiation on mountain slopes. Water Resources Research 12, 108–112. Trimble, G.R., Jr and Weitzman, S. (1956) Site index studies of upland oaks in the northern Appalachians. Forest Science 2, 162–173. Ward, W.W. and Bowersox, T.W. (1970) Upland oak response to fertilization with nitrogen, phosphorous, and calcium. Forest Science 16, 113–120.

Site Productivity

193

Waring, R.H. and Schlesinger, W.H. (1985) Forest Ecosystems: Concepts and Management. Academic Press, Orlando, Florida. Whittaker, R.W. (1963) Net production of heath balds and forest heaths in the Great Smoky Mountains. Ecology 46, 176–182. Whittaker, R.H. and Niering, W.A. (1975) Vegetation of the Santa Catalina Mountains, Arizona. V Biomass, production and diversity along the elevation gradient. Ecology 56, 771–790. Whittaker, R.H. and Woodwell, G.M. (1968) Dimension and production relations of trees and shrubs in the Brookhaven Forest, New York. Journal of Ecology 56, 1–25. Yin, X., Perry, J.A. and Dixon, R.K. (1989) Fine-root dynamics and biomass distribution in a Quercus ecosystem following harvest. Forest Ecology and Management 27, 159–177. Zahner, R. (1970) Site quality and wood quality in upland hardwoods: theoretical considerations of wood density. Proceedings of 3rd North American Forestry and Soils Conference, pp. 477–497. Zahner, R. and Myers, R.K. (1984) Productivity of young Piedmont oak stands of sprout origin. Southern Journal of Applied Forestry 8, 102–108. Zahner, R., Myers, R.K. and Churchill, L.A. (1982) Site index curves for young oak stands of sprout origin. Clemson University Forestry Bulletin 35.

5 Development of Natural Stands

Introduction Stands are the basic units of forest management. Although they can vary widely in area, oak stands typically range from 2 to 40 acres. By definition, a stand covers a relatively homogeneous area with respect to vegetation, soil and site quality. Stand boundaries are usually delineated by a combination of stand characteristics including age structure (even-aged or uneven-aged), predominant tree size (seedling, sapling, pole or sawlog) and species composition. Site homogeneity is usually evaluated through the use of site evaluation (Chapter 4) and ecological classification methods (Chapter 1).

Forest Canopy Layers The canopy of a mature oak stand can be divided into two broadly defined layers: overstorey and understorey. The overstorey consists of a main canopy and a subcanopy. The main canopy includes the upper layers of tree crowns, which intercept most of the sunlight (Fig. 5.1). Beneath the main canopy there is often a subcanopy of sapling-size trees (1–5 inches dbh), and beneath that an understorey layer that includes tree reproduction, shrubs and herbaceous plants. Understorey species are usually shade tolerant because only about 1–5% of the sunlight received by the main canopy reaches the forest floor.

The understorey is nevertheless the domain of advance tree reproduction including oak seedlings and seedling sprouts. For silvicultural purposes, it is convenient to define an upper size limit for tree reproduction. In the Central Hardwood Region, for example, 1.5 inches dbh is commonly used to define this limit; larger trees are thus defined as members of the overstorey. However, this limit was probably set more by mensurational convention rather than by biological considerations. Overstorey trees can be further categorized by their relative canopy position, or crown class. Four crown classes are generally recognized: dominant, co-dominant, intermediate and overtopped (or suppressed) (Fig. 5.1). The crowns of dominant trees extend partially above the general level of the main canopy where they receive light both from above and from the sides. These trees are the tallest in a stand, and they usually have the largest crowns. The crowns of co-dominant trees receive full sunlight directly from above but little from the sides. These are among the taller trees in the stand, and their crowns define the upper level of the main canopy. The crowns of intermediate trees occupy the lower part of the main canopy. They are among the smaller trees in the stand, and receive light only on limited portions of crown tops. The crowns of overtopped trees lie completely below the main canopy in the subcanopy. Little direct sunlight is received by any part of the crowns of overtopped trees, which are often flat-topped and irregular in shape. 194

Development of Natural Stands

195

D

D

C C

C I

I

I

Main canopy

I

Gap

Overstorey O

O

O

O

O

Subcanopy Understorey

Fig. 5.1. Forest canopy layers can be subdivided into overstorey (including the main canopy and subcanopy), and understorey (including tree reproduction, shrubs and herbaceous vegetation). Associated tree crown classes are: D = dominant, C = co-dominant, I = intermediate, O = overtopped (suppressed). The capacity of oak reproduction (especially seedling sprouts) to capture canopy gaps depends on their root size at the time the canopy gap is created, their inherent growth rate (which varies among oak species), gap size, competition from other vegetation and other factors.

The way trees are spatially arranged is called stand structure. Stand structure can be described by various qualitative and quantitative attributes of forests. Crown class is an example of a qualitative attribute that is useful in describing a tree’s relative crown position. By definition, crown classes infer that trees in forests are vertically stratified. However, the concept of stand structure also encompasses the areal distribution of trees including their ages and sizes, and the distribution of other stand components such as herbaceous vegetation and coarse woody debris (Helms, 1998). Although age and size structure are related to vertical stratification, they can also be quantitatively defined by frequency distributions of tree ages or sizes. The frequency distribution of tree ages is used to distinguish even-aged from uneven-aged stands and to measure other age-specific properties of stands. Stand size structure is often described by the frequency distribution of tree diameters; this is usually referred to simply as the diameter distribution. Because the diameter distribution of a stand changes as the stand develops, it is a useful diagnostic for assessing stand development. Stand structure in an even

broader context can include how closely packed the trees are. Various measures of stand density and stocking have been developed for quantifying this aspect of stand structure (Chapter 6). Stands also vary in species composition, i.e. the relative proportions of species present. Collectively, the various measures of stand structure and species composition can provide the silviculturist with essential information about the current state of a stand and, by extension, silvicultural options for its management. Central to an understanding of these options is an understanding of forest disturbance events and the related ecological processes and patterns of stand development.

Disturbance Forests continually change as a result of events originating from inside and outside the forest. These events are ubiquitous and in a broad sense can be termed ‘disturbances’. Many forest disturbances are ecologically and silviculturally important because they affect the current structure and composition of a stand as well as its

196

Chapter 5

future development. Disturbances have been variously defined as ‘any relatively discrete event in time that disrupts ecosystem, community, or population structure and changes resources, substrate availability, or the physical environment’ (White and Pickett, 1985); ‘as a discrete, punctuated killing, displacement, or damaging of one or more individuals (or colonies) that directly or indirectly creates an opportunity for new individuals (or colonies) to become established’ (Sousa, 1984); or as ‘a force that kills at least one canopy tree’ (Runkle, 1985). Disturbances increase the resources available to surviving trees in the vicinity of the disturbance. The type, size, severity and frequency of disturbances greatly affect how oak stands develop.

Disturbance type Forest disturbances can be classified as exogenous or endogenous. Exogenous disturbances originate from forces external to the stand and may damage or kill vigorous as well as unhealthy trees. Examples include disturbances caused by weather, fire and human activity including silviculture. Endogenous disturbances originate within the stand, and tree death and treefall are primary causes. It is not always apparent which type of disturbance has occurred. For example, stresses resulting from the competition between trees (an endogenous factor) may predispose a tree to a fatal insect defoliation or windthrow (exogenous factors). The ecological impact of a disturbance may depend more on its size and severity than its origin (White and Pickett, 1985). Exogenous and endogenous factors represent endpoints of disturbance size that can range from the creation of extensive forest canopy openings with abrupt boundaries to small changes in stand structure associated with the toppling of a single overstorey tree (White and Pickett, 1985). Although exogenous factors typically produce larger disturbances than endogenous factors, endogenous events nevertheless can have substantial localized impacts. For example, when a large, senescent tree falls

it may shear the crowns of surrounding trees, topple or crush other trees in the line of fall, disturb the soil at points of impact, and uplift a mound of soil in its root mass (Bormann and Likens, 1979).

Disturbance size and frequency of occurrence Disturbances of one type or another occur frequently during the life of every oak stand, and each one changes the structure and composition of the stand to some degree. Reice (1994) suggested that to be in recovery from their last disturbance is the normal state for most natural plant communities, and that they rarely attain a state of equilibrium. This is certainly true of oak forests, which owe their very existence to disturbance. Most are therefore in a state of recovery and transition related to their disturbance history. Disturbances vary greatly in their area, severity and frequency, and in associated ecological responses (Sousa, 1984). Even seemingly small disturbances such as animal browsing or leaf litter fires may significantly alter a forest if they are widespread and frequent. Disturbances that create large openings in the forest canopy are of particular interest to ecologists and silviculturists because such openings significantly alter the course of stand development. Many silvicultural practices fall into this category, and silvicultural treatments can be thought of as disturbances that are designed to control future stand composition and structure. From a silvicultural perspective, forest disturbances can be categorized according to their relative size and impact on stand development: (i) gap-scale disturbances; (ii) incomplete stand-scale disturbances; and (iii) stand-initiating disturbances. Gap-scale disturbances are the smallest and can be endogenous or exogenous in origin. The other two categories result from exogenous forces. The impact of a disturbance on a stand should be evaluated in the context of stand structure and composition. A disturbance of a given type


and size will usually elicit a different response from a young stand than from a mature stand or among stands that differ in species composition. Gap-scale disturbances occur when a single tree or a small group of trees are lost from the main canopy. The resulting canopy gaps increase light and soil moisture available to trees within the gap and to trees adjacent to the gap. When stands are well stocked and pole-size or smaller, canopy gaps created by the death of small trees are quickly filled through crown expansion of the surrounding trees. During later stages of stand development, gaps created by the death or harvest of individual trees often are larger and thus may persist for decades. Silvicultural methods designed to create gap-scale disturbances include some types of thinning (Chapter 7), and single-tree and group selection methods (Chapter 8). During a century of even-aged oak stand development, more than 95% of the trees initially present will die. Most of those deaths are caused by inter-tree competition and consequent self-thinning of stands (Chapter 6). As a result, canopy gaps are a common occurrence in oak stands. Although trees of all crown classes die from self-thinning, mortality rates are lower for dominant and co-dominant trees than for trees in subordinate crown classes. For example, in New England stands, 9% and 33% of northern red oaks in dominant and co-dominant crown classes, respectively, died between stand ages 25 and 55. In contrast, 67% and 90% of trees in intermediate and suppressed classes, respectively, died over the same period (Ward and Stephens, 1994). When an individual tree dies, it leaves a canopy gap that is roughly proportionate in area to its basal area. The vacated space is subsequently captured by the expanding crowns of surrounding trees and/or by tree reproduction. Larger gap-scale disturbances occur when small clusters of trees die at the same time, or when individual trees with large crowns die. As gap size increases, the more abrupt and extensive are the associated stand changes. The size of the gap, rate of crown

197

expansion of surrounding trees, and the growth rate of trees within the gap all influence future stand composition and structure (Hibbs, 1982). Incomplete stand-scale disturbances are larger and more disruptive to the stand than gap-scale disturbances. They are of exogenous origin and caused by events that affect a large proportion of a stand’s area. An incomplete stand-scale disturbance removes enough overstorey trees to initiate the development of a new age class of trees within the stand. The resulting stand age structure is often irregular (i.e. the tree age classes occupy unequal areas), and there may be large variation in the size and number of openings in the canopy. Silvicultural practices that can create incomplete standscale disturbances include seed tree, diameter limit, shelterwood and group selection methods. Fires of sufficient intensity to kill patches of overstorey trees also can result in incomplete stand-scale disturbances. The largest and most severe disturbances are those that initiate a new stand. These disturbances remove most or all of the overstorey. Some residual canopy trees may remain, but the net result is the creation of a new, even-aged stand. High wind, severe fire, and other destructive events that are an acre or larger can cause stand-initiating disturbances. Silvicultural treatments such as clearcutting and shelterwood removal harvests (Chapter 7) are also stand-initiating disturbances. A disturbance of this type resets stand age to or close to zero. In evenaged stands, stand age is thus a measure of the time elapsed since the previous standinitiating disturbance. In managed forests, silvicultural treatments are typically implemented at 10–30 year intervals. Gap-scale disturbances within a stand generally occur with greater frequency than that and are anticipated in silvicultural prescriptions. However, standinitiating disturbances or incomplete stand-scale disturbances are infrequent events that are generally not anticipated in a managed stand except as a prescribed timber harvest. Silvicultural prescriptions nevertheless can be designed to reduce the negative consequences of extensive natural

198

Chapter 5

disturbances such as severe fire, wind or insect damage, which inevitably affect some managed stands. When a disturbance creates an opening of any size in the forest, it increases the availability of resources in and around the opening. Woody vegetation responds in three ways. First, the crowns of trees surrounding the opening grow laterally to capture the space. Although the rate of lateral crown expansion depends on many factors, oaks expand at rates from 6 to 12 inches per year in eastern forests (Fig. 3.22; Trimble and Tryon, 1966; Hibbs, 1982). Crowns of northern red oaks often expand more rapidly than crowns of co-occurring species. In general, crowns of mature trees tend to expand more slowly than young, vigorous trees. And mature oaks in the superior crown classes may not expand at all (Sampson, 1983). Large trees in mature stands nevertheless leave large gaps when they die. As stands mature, there are more opportunities for such gaps to form, and those that persist for decades may facilitate the establishment of oak and other tree species in the understorey. A second response to gap formation is an increase in the height growth of preestablished tree reproduction and subcanopy trees within the gap. This response, in turn, can facilitate the successional replacement of canopy dominants by trees that have persisted in intermediate and suppressed crown classes and that have the capacity to opportunistically expand into canopy gaps. Gap formation accordingly can potentially accelerate the recruitment of shade-tolerant trees and even some fastgrowing, shade-intolerant trees into the overstorey. Whether or not such a replacement process occurs varies with the species composition and the physiological capacity of subordinate canopy layers to capture and hold canopy gaps. A third response to canopy gap formation is the establishment of new trees (and other vegetation). If the openings are large enough and persist long enough, trees that become established in a gap after its formation may eventually grow into the overstorey. In a hemlock–mixed hardwood

forest, seedlings (including oaks) required a canopy opening with a diameter at least half the height of the surrounding trees to facilitate growth into the overstorey (Hibbs, 1982). In smaller openings the crown expansion of surrounding trees usually closed the gap before new trees reached the main canopy. Not only does a canopy gap close laterally with time (Fig. 3.22), the minimum area required to sustain a tree within the gap increases as the tree itself increases in size. A tree therefore must reach the level of the main canopy before its own minimum growing space requirements exceed the gap’s diminishing size. Thus, recruitment of new trees within a gap depends not only on the species, initial size, vigour and growing space requirements of trees within the gap, but also on the size, crown expansion rate and spatial arrangement of trees surrounding the gap. Disturbances that eliminate oaks in the main canopy without also eliminating shade-tolerant trees in the sub-canopy may accelerate succession towards dominance by shade-tolerant species (Abrams and Scott, 1989; Abrams and Downs, 1990; Abrams and Nowacki, 1992). However, moderate overstorey disturbances sometimes favour the oaks. For example, prolonged drought reduced numbers of yellow-poplars, which consequently benefited competing oaks in an Ohio stand (Hilt, 1985). Where oaks and yellow-poplar co-occur, yellow-poplar typically overtops the oak within 10 years of a stand-initiating disturbance (Loftis, 1983, 1990; Beck and Hooper, 1986; Weigel and Johnson, 1999). Consequently, when oaks compete with yellow-poplar, the oaks usually lose. Exceptions occur when a disturbance such as drought or fire selects against yellow-poplar and alters the usual course of succession. In the absence of timber harvesting, the average rate of canopy disturbance across a wide range of temperate forest conditions is about 1% per year and ranges from about 0.5 to 2% (Runkle, 1985). The average time between successive replacement of trees in the main canopy is therefore about 100 years (range 50–200 years). This is consistent with the observed longevity of oaks.


Although individual oaks may live 400 years or longer, attaining such ages is rare. Even in old-growth oak forests, only a small fraction of trees survive more than 200 years. The distribution of disturbances varies in both time and space. When a disturbance occurs, it temporarily reduces the likelihood of a subsequent disturbance of the same type. For example, high winds that remove the most susceptible trees from a stand usually also reduce the likelihood of additional wind damage for several years thereafter. Likewise, periodic fires that eliminate the most fire-sensitive trees and reduce fuels decrease the likelihood of intense fire damage for several years thereafter. Large or severe disturbances therefore tend to occur at infrequent intervals and are followed by years with below-average rates of disturbance. Suppression of disturbances (e.g. through fire suppression) may increase the likelihood of a severe disturbance in the future. Frequent low-intensity disturbances that remove some trees may increase the growth rate and vigour of surviving trees, and also may reduce the number and impact of subsequent disturbances (Waring and Schlesinger, 1985). However, this is not always the case. Certain combinations of events (e.g. windthrow followed by fire) may create conditions (e.g. increased fuels) capable of producing catastrophic events. Just as the size and species composition of a stand are affected by a disturbance, the reverse is also true. The area, spatial pattern and intensity of a natural disturbance may be modified by the size and species composition of the vegetation and by the landform where it occurs (Reice, 1994). For example, certain species and certain topographic positions are especially susceptible to wind and fire damage. Temporal and spatial variation in disturbance events also tends to be interrelated (Runkle, 1985). During years when the total disturbed area is relatively large, disturbances are likely to be spatially clustered. This happens either because a few disturbances cover an exceptionally large area or because a large number of smaller disturbances occur within a

199

fixed area, and many disturbed areas are adjacent to one another. As stands change over time, their susceptibility to disturbance also may change (White and Pickett, 1985). Small trees are usually more susceptible to fire damage than large trees. Large trees with welldeveloped crowns are more vulnerable to wind damage than small trees. Changes in species composition that occur during the course of stand development also may alter a stand’s susceptibility to disturbance by agents such as fire, flooding, insects, disease or silvicultural activity.

Development of Even-aged Stands An even-aged stand consists of a cohort of trees comprising a single age class. This usually means that tree ages differ by no more than 20 years. In silvicultural applications, a more specific definition is sometimes preferred. For example, Smith (1986) defined an even-aged stand as one where the difference in age between the youngest and oldest trees does not exceed 20% of the rotation (Chapter 7). Although the biological age of an oak (i.e. its age from germination) is most accurately expressed by the number of annual rings just below its root collar (Chapter 3), stand age is conventionally defined by the average age of tree boles. This age usually corresponds to the number of years since the previous standinitiating disturbance. Even-aged stands progress through a relatively predictable series of developmental stages until the next stand-initiating disturbance or incomplete stand-scale disturbance occurs. Defining these stages is useful in understanding the development of oak forests even though the duration of each stage and the accompanying changes in stand structure, density and species composition may differ from stand to stand. Although various terms have been used to define the stages of stand development (e.g. Bormann and Likens, 1979), we herein follow the terminology of Oliver and Larson (1996) as modified by Oliver (1997). Four stages have been defined: (i) the stand initi-

200

Chapter 5

ation stage; (ii) the stem exclusion stage; (iii) the understorey reinitiation stage; and (iv) the complex stage (Fig. 5.2). The complex stage of development also has been called the old-growth stage (Oliver, 1981; Oliver and Larson, 1996). The progression of an even-aged oak stand through the various stages of development is accompanied by changes in stand structure and often by changes in species composition. Changes in stand size structure include the accumulation of biomass (Bormann and Likens, 1979), vertical stratification of tree crowns and correlated changes in diameter distributions. For some regions, these changes are generalized in published stand tables that list by site classes the expected numbers of oaks and other species

Stand initiation stage

Stem exclusion stage

by tree diameter and age classes in wellstocked stands (e.g. Schnur, 1937; Gevorkiantz and Scholz, 1948). Such tables thus represent models of average expected change in stand structure associated with different stages of stand development. Unlike changes in stand size structure, which follow a similar progression as stands age (Fig. 5.2), changes in species composition often vary from stand to stand, even within similar ecological settings. Species composition in even-aged oak stands largely depends on the size and initial species composition of tree reproduction and propagules present at the time of the previous stand-initiating disturbance. The trees and the other vegetation present at that time develop into the new

Understorey reinitiation stage

Complex stage

Fig. 5.2. Stages of stand development that occur after a major disturbance that destroys all or most of the parent stand. Stand initiation stage: immediately after the disturbance, pre-established reproduction grows rapidly and new trees and other plants appear. In oak stands this stage typically lasts 10–20 years. Stem exclusion stage: no new trees appear and many die from crowding. Trees are well stratified into crown classes by the end of this period. This stage usually begins after the stand reaches 10 years of age and concludes before age 70. Understorey reinitiation stage: tree reproduction becomes re-established under the maturing overstorey. Re-establishment of trees in the understorey is facilitated by the death of individual trees in the main canopy. Canopy gaps are of sufficient size and frequency to significantly increase light on the forest floor. This stage typically begins after age 50 and concludes before age 120. Complex stage: natural mortality of large overstorey trees produces irregular canopy gaps and accelerates the recruitment of reproduction and sub-canopy trees into the overstorey and main canopy, respectively. This stage marks the transition from an even-aged stand to an uneven-aged stand. Oak forests typically require 100 years or longer to reach the complex stage of development. The stated durations for the four stages of stand development are representative of oak forests of the eastern United States and assume that no significant stand-scale disturbances occur. Actual durations of stages of development vary with species composition, site productivity and other factors. (Adapted from Oliver and Larson, 1996; Oliver, 1997.)


oak stand. Differences among tree species in shade tolerance, longevity and growth potential influence how the species composition of the stand will change over time. Changes in composition are also influenced by disturbances such as extreme weather events, disease, animal browsing and human activity. Often these events are only predictable probabilistically. Consequently, young stands of similar composition and structure may follow different developmental trajectories and thus differ greatly in composition and structure at maturity. Structural and compositional changes in forests are attributes of the broader process of ecological succession. Although ecological succession is generally regarded as the natural change in the composition, structure and function of an ecosystem, successional concepts are also applicable and central to silviculture. In this context, silviculture is about directing ecological succession. Ecological succession is of two types: primary and secondary. Primary succession involves long-term ecosystem changes, usually beginning with bare ground devoid of vegetation, and often spanning thousands of years. Related changes in vegetation are associated with changes in the parent material, soil and other factors that are interactively modified through time. Secondary succession involves shorterterm changes occurring after a disturbance. Secondary successions therefore usually begin with vegetation, plant propagules and other organisms already in place. Thus, there is a direct connection between ecological succession and silvicultural practice, which seeks to design and create disturbances that direct secondary succession in specific ways. The study of secondary forest succession has traditionally focused on changes occurring over relatively long periods and in the absence of significant exogenous disturbance. However, one objective of silviculture is to anticipate and control successional changes in stand structure and composition that result from stand disturbances. In turn, such disturbances result from endogenous events associated with natural stand development (e.g. the death

201

of individual canopy trees), from natural exogenous events (e.g. wildfire or weather) or from human intervention (e.g. silviculture). Silvicultural prescriptions are thus applied at various stages of stand development with the intent of directing the course of change in stand structure and species composition in specific ways. Before prescribing silvicultural treatments, it is therefore important first to understand how even-aged stands progress through the stages of structural development in the absence of major exogenous disturbances.

The stand initiation stage The development of an even-aged stand begins with the stand initiation stage. In oak forests, this stage may last up to 20 years. In temperate regions it is characterized by a brushy mass of woody vegetation comprised of thousands of trees and shrubs per acre often mixed with a luxuriant growth of vines and herbaceous plants (Gingrich, 1971) (Fig. 5.3A). It is also a period of rapid change with intense competition among trees and other plants for growing space. Standing biomass is small relative to later stages of development, but the rate of biomass increase is high. During this stage, the quantity of dead biomass is often larger than during other stages of stand development. This is due to the standinitiating disturbance itself, which (except for fire) usually leaves a large residue of tree boles and branches on the forest floor (Bormann and Likens, 1979; Jenkins and Parker, 1997; Spetich et al., 1999). During the stand initiation stage, gaps in the new vegetative cover may persist for a decade or longer as new trees and other vegetation become established. New tree seedlings and herbaceous vegetation initially require little growing space, and numerous small openings in the developing forest vegetation provide ‘safe sites’ for their establishment (Harper, 1977). These are places where seeds find the necessary conditions for germination and growth free from predators, competitors and pathogens. Changes in the number, species and size of

202

Chapter 5

A

B

C

D

Fig. 5.3. The four stages of stand development representative of oak stands in the Ozark Highlands of Missouri: (A) stand initiation stage; (B) stem exclusion stage; (C) understorey reinitiation stage; and (D) complex stage. (USDA Forest Service, North Central Research Station photographs.)

trees during the stand initiation stage are difficult to predict accurately. This is due to the numerous and essentially random

events that influence the timing and spatial patterns of seed dispersal, germination and seedling survival. Stand development


during this period is subject to great natural variation, and predictions of stand development during this stage are often specified qualitatively (Johnson and Deen, 1993) or probabilistically (Chapter 3; also see Johnson and Sander, 1987; Loftis, 1990; Dey, 1991; Dey et al., 1996). Although the composition of the future stand is strongly influenced by the species initially present at this stage of development, the relative dominance of the various species usually changes as the stand matures. Early theories of forest succession (e.g. Clements, 1916) proposed that each species modifies the site to make it more favourable for the establishment and growth of succeeding species. However, studies of mixed oak–hardwood forests in New England have shown that most of the trees that form the overstorey originated within one or two decades after a major disturbance (Oliver and Stephens, 1977; Oliver, 1978). Oaks that attain membership in the main canopy are usually established before the stand-initiating disturbance occurs (Chapter 2). Species that are shade tolerant, longlived and capable of growing to large size may persist in subordinate canopy positions for decades. They may gradually make their way into the main canopy by opportunistically responding to gap-size disturbances. At stand initiation, the presence of 100 seedlings or seedling sprouts per acre of a late-successional species might seem insignificant. However, if that population persistently captures canopy gaps, over time it may eventually dominate the main canopy. Different species that become established at approximately the same time therefore may exert dominance at different stages of stand development (Egler, 1954; Oliver and Larson, 1996; also see Chapter 3). The overstorey of a mature even-aged oak stand is largely comprised of species present at the end of the stand initiation stage. During the two stages of stand development that follow the stand initiation stage, few if any new trees are added to the overstorey. Consequently, the composition of an even-aged stand at the end of the

203

stand initiation stage is a good indicator of the composition of the overstorey of the future stand. However, a species’ relative abundance at this stage is often a poor indicator of its future importance. During the ensuing stem exclusion stage, species composition usually shifts toward the species that are best adapted to the site and able to attain and persist in dominant and co-dominant crown positions. Dominance by oaks at the end of the stand initiation stage therefore does not, in itself, ensure their continued dominance.

The stem exclusion stage In temperate regions, crown closure in oak stands has usually occurred by the beginning of the second decade after a stand-initiating disturbance. By that time, trees have stratified into well-defined crown classes and natural mortality has changed the initially clumped spatial distribution of trees to a more random distribution (Rogers, 1983). This stage of stand development is termed the stem exclusion stage because few, if any, new stems are added to the population of overstorey trees (Oliver, 1980; Oliver and Larson, 1996) (Figs 5.2 and 5.3B). Mortality rates are high, especially among trees in intermediate and suppressed crown classes. The combined growth, competition and mortality of trees during this stage produce spatial adjustments in the main canopy that lead to full or nearly full crown closure and a corresponding full utilization of growing space (Gingrich, 1967, 1971). It is usually not until after the stem exclusion stage begins that growth and yield tables or other predictive models are applicable (Chapter 10). During this and subsequent stages, patterns of stand development and species differentiation are more predictable than during the stand initiation stage. Dense populations of saplings (trees 1–5 inches dbh) and pole-size trees (5–10 inches dbh) dominate the stem exclusion stage. During this stage, the expanding crowns of dominant and co-dominant trees quickly fill the space vacated by dead and dying

204

Chapter 5

oaks to eventually grow taller than subcanopy competitors (e.g. flowering dogwood) that may initially overtop oaks. Persistence thus involves negative impacts on two or more populations, but one population emerges the ‘winner’ because it persists and is less negatively affected than the other. However, the persistence advantage of one species over another often changes across environmental gradients such as the moisture gradient. The diameter distribution of an evenaged stand changes continually as the stand develops. For example, in 10-yearold upland oak stands of the eastern United States, diameter distributions form reverse J-shaped curves comprised of thousands of trees per acre (Fig. 5.4). By the time stands reach a mean diameter of 8 inches dbh and in the absence of disturbance, a bell-shaped diameter distribution comprised of a few hundred trees per acre has formed (Schnur, 1937). On sites of average quality, this occurs at about stand age 85, which is well into the understorey reinitiation stage of development. In the absence of exogenous disturbance,

trees of the inferior crown classes. Even if tree reproduction is present in the understorey, it is usually unable to capture the small, transient canopy gaps that occur during this stage. Little light reaches the forest floor and understorey vegetation decreases as the herbaceous and low woody vegetation that flourished during the stand initiation stage die from suppression beneath the rising level of the overstorey. During the stem exclusion stage, oaks can sustain a position of dominance in three ways: (i) through inherently faster growth than competitors; (ii) through an initially superior crown position; or (iii) through persistence. Persistence involves survival and continued growth of a species when the population of competitors fares less well under the same conditions. For example, proportionately more oaks may survive drought than faster-growing and potentially long-lived competitors such as red maple or yellow-poplar (Hilt, 1985). Oaks are relatively persistent following disturbance events such as fire in uplands or flood scouring in bottomlands. In some ecosystems, persistence also may allow

800 10* 700 Trees per acre

3000

Trees per acre

600 20 500

2000 1000

400 0 300 30

200

40

100

1

5 9 13 Dbh (inches)

17

50 60 80

*Stand age

0 2

4

6

8 10 12 Dbh (inches)

14

16

Fig. 5.4. Diameter distributions of trees for a time-series of even-aged upland oak stands on average sites in the eastern United States. Inset: The reverse J-shaped diameter distribution resulting from summing the number of trees in each dbh class across all even-aged diameter distributions (for 10-year age classes from 10 to 100). (Adapted from Schnur, 1937.)


numbers of trees decline with time through the process of self-thinning (Chapter 6). The development of a bell-shaped diameter distribution is not a universal characteristic of even-aged oak stands. Because oak stands are usually mixtures of species, forest canopies in later stages of development tend to stratify vertically by species. This stratification results from species’ differences in shade tolerance, longevity, growth rate and maximum attainable size. Slower growing, shade-tolerant trees may persist for decades in subordinate crown classes (Oliver, 1980; Hibbs, 1983). Differential rates of establishment and growth among species in even-aged stands may produce reverse J-shaped and other types of diameter distributions often associated with uneven-aged stands. However, the correlation between tree age and diameter is often low – and near zero in truly even-aged stands. Diameter distributions, by themselves, therefore are not reliable indicators of age distributions, especially in stands comprised of several species (Oliver, 1980; Loewenstein, 1996; also see Chapter 8). The development of an Arkansas sweetgum–red oak stand illustrates the transition from the stand initiation stage to the stem exclusion stage of stand development (Johnson and Krinard, 1983, 1988). This bottomland stand in the Southern Hardwood–Pine Region originated from a timber harvest that left only a few seed trees per acre. Unmerchantable trees were killed and areas influenced by seed trees were excluded from the analysis. Consequently the results essentially describe the development of a stand after clearcutting. Estimated site index for sweetgum was 100 ft at 50 years. Numbers of trees and stand basal area increased for the first 15 years (Fig. 5.5A). By age 18, the declining number of trees and increasing basal area marked the onset of the stem exclusion stage (Fig. 5.5B). The transition from stand initiation stage to stem exclusion stage is also apparent from the simultaneous changes in basal area and numbers of trees (Fig. 5.5C). Between stand ages 3 and 9, numbers of

205

trees rapidly increased as trees filled the growing space vacated by the harvested parent stand. By age 15, the new stand had attained its maximum number of trees. Thereafter, the trajectory reversed direction, i.e. basal area continued to increase but numbers of trees decreased. This change in direction also marked the beginning of the stem exclusion stage and the onset of competition-induced self-thinning (Chapter 6). Despite the decreasing numbers of trees, the stand continued to increase in basal area after age 15. Although the basal area of undisturbed stands usually increases until stands are in the complex stage of development, the periodic rate of basal area increment slows with increasing stand age. This slowing is indicated by the smaller annual increments of basal area that occurred after age 15 (Fig. 5.5B and C). It is during the stem exclusion stage that oaks must capture growing space from their competitors if they are to dominate the later stages of stand development. When numbers of trees decline, it is usually the smallest trees that die first. The vacated growing space and thus the greatest basal area increment usually accrues to large trees of superior crown classes. The time-dependent change in the relation between basal area and numbers of trees also can vary among species in the same stand. For example, sweetgum and river birch followed nearly identical trajectories through stand age 9 (Fig. 5.5D). Because river birch declined in both numbers of trees (after age 9) and in basal area (after age 18), it was nearly excluded from the stand by age 29. Although the trajectory for American hornbeam was similar, it did not decline in numbers of trees until later, which may be related to its greater shade tolerance. Although numbers of red oaks also declined after age 18, their survival rate was greater than that of other species. The combination of rapid growth and a low mortality rate thus enabled the oaks to attain a dominant position in the stand even though they were relatively few in number. Nevertheless, after 29 years of stand development, sweetgum and other non-oaks accounted for more basal area and trees per acre than the oaks.

206

Chapter 5

2500

160 B

A 140

1500

1000 Stand initiation stage


Basal area (ft2 acre–1)

Trees per acre

2000 120 100


80


60 40

500 20 0

0 0

5

10

15

20

25

0

30

5

10

160

(18)

100

(15)

80 60 (9)

40

30 Red (29) oaks

(29)

20

River birch

1000

(18)

American hornbeam

(18)

(18) (18)

(29)

(9)

(6)

0 500

(18)

(29)

(3) 0

D

Other major species

(6)

0

30

Sweetgum (29)

40

10

20

25

50

(23)



60

C

(29)

120

20

Stand age (years)

Stand age (years)

140

15

1500

2000

2500

Trees per acre

0

200

400

600

Trees per acre

Fig. 5.5. Change in number and basal area of trees ≥1 inch dbh in an even-aged sweetgum–red oak stand in southeastern Arkansas during its first 29 years of development. (A) During the first 15 years, the increasing number of trees ≥1 inch dbh is characteristic of the stand initiation stage of development. After age 15, the decreasing number of trees ≥1 inch dbh marks the stem exclusion stage of development. (B) Basal area increased through both stages of development. (C) Change over time in basal area and number of trees for all species combined. Plotted lines trace changes from stand age 3 to 29 (stand ages are shown in parentheses). (D) Change over time in basal area and number of trees for five species groups. All species begin near the origin at stand age 3. The plotted age sequence through 29 years is the same as shown in (C). Red oaks include cherrybark, water and willow oaks. (Adapted from Johnson and Krinard, 1988.)

Stocking charts are commonly used to quantify the degree of tree crowding (and thus the utilization of growing space) in relation to basal area and trees per acre

(Chapter 6). Although stocking charts are used primarily to evaluate stand density at a single point in time, they are also useful in illustrating stand development through


time (e.g. Fig. 5.5C and D). Accordingly, they can be used to show how stands change in composition and structure, and in the proportionate utilization of growing space by different species. For example an even-aged sweetgum–red oak stand in a southern bottomland maintained a reverseJ diameter distribution during its first 29 years of development (Fig. 5.6). But over time the shape of the diameter distribution gradually flattened as average tree diameter increased and as numbers of trees in the smaller diameter classes decreased. After the stand entered the stem exclusion stage, the shape of the diameter distribution of the oaks departed from the overall diameter distribution. By stand age 23, the largest oaks in the stand formed a bell-shaped diameter distribution, whereas smaller oaks maintained a reverse J-shaped distribution (Fig. 5.6). Oaks larger than 6 inches dbh formed a ‘hump’ in the diameter distribution, which in previous years formed a reverse J-shape. By age 29, the red oaks thus diverged into two distinct populations: a population of main canopy trees and a population of sub-canopy trees. The structures of these two sub-populations appear destined to diverge further as the largest oaks continue to increase in dominance and the smaller oaks become increasingly suppressed beneath the rising level of the main canopy. Changes in species composition commonly occur after oak stands enter the stem exclusion stage and trees begin to differentiate into well-defined crown classes. Such changes are reflected in differences in proportions of species across diameter classes. Although after stand age 15 the total number of trees per acre decreases with increasing stand age and mean stand diameter (Fig. 5.5A and C), the number of trees of any given species may increase or decrease as a proportion of all trees as the stand matures. For example, sweetgum and other non-oaks were the only trees in the largest diameter classes (≥11 inches dbh) at stand age 18, whereas the largest trees in the red oak group (cherrybark, water and willow oaks) were all less than 7 inches dbh at the same age. However, by stand age 29 red

207

oaks represented half the trees in the largest (14 inches) dbh class. Although total numbers of trees decreased by nearly 50% between stand ages 18 and 29, numbers of oaks decreased by only 13%. Because of the rapid diameter growth of the largest oaks, they were among the largest trees by age 29. Thus, it was not until the stand was more than 10 years into the stem exclusion stage that the oaks emerged as a major component of the main canopy. Although sweetgum still dominated the stand after 29 years, the red oaks had increased in dominance at the expense of other species. Oak stands on droughty sites or sites affected by recurrent, low-intensity disturbances (e.g. oak savannas maintained by periodic burning) may not experience a stem exclusion stage. In oak savannas, periodic burning maintains a relatively open canopy, which in turn maintains relatively high light intensities in the understorey. Although oak and other tree reproduction usually accumulate under these conditions, their recruitment into the overstorey is inhibited by burning unless there is a fire-free period of sufficient duration to permit their recruitment into the overstorey (Chapter 9). Under these conditions, recruitment of oaks into the overstorey is limited more by the disturbance regime than by competition or light.

The understorey reinitiation stage During the understorey reinitiation stage, tree reproduction becomes re-established beneath the parent stand (Figs 5.2 and 5.3C). In oak forests, this reproduction usually becomes a major component of the new stand that develops after the next stand-initiating disturbance. Many factors influence which species become established in the understorey and consequently which species are likely to dominate after a stand-initiating disturbance. In oak forests, light and soil moisture rank among the most important of these factors (Fig. 3.7). This is also the stage of development when oak stands usually attain economic maturity.

208

Chapter 5

All trees

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

1600

Age 3

120

1200

90

800

60

400

30

0 1600

120

Age 6

90

800

60

400

30

0 1600

0 120 Age 9

1200

90

800

60

400

30

0 1600

Age 15

0 120

1200

90

800

60

400

30

0 1600

Age 18

0 120

1200

90

800

60

400

30

0 1600

Age 3

0

1200

Age 23

0 120

1200

90

800

60

400

30

0

0 120

1600 Trees per acre

Red oaks

Age 29

1200

90

800

60

400

30

Age 6

Age 9

Age 15

Age 18

Age 23

Age 29

0

0 0 2 4 6 8 10 12 14 16 Dbh (inches)

0 2 4 6 8 10 12 14 16 Dbh (inches)

Fig. 5.6. Change in diameter distributions in an even-aged sweetgum–red oak stand in southeastern Arkansas during its first 29 years of development. Until age 15, the stand was in the stand initiation stage of development and thereafter was in the stem exclusion stage. By age 23, the red oaks formed two distinct populations: large trees (6–10 inches dbh) occupying the main canopy, and smaller trees occurring in the sub-canopy. The number of trees is plotted over the midpoints of 2-inch dbh classes. Red oaks include cherrybark, water and willow oaks. (Adapted from Johnson and Krinard, 1988.)


Consequently, it is a critical period from a silvicultural perspective. Compared to the stem exclusion stage, trees in the main canopy are larger and fewer in number during the understorey reinitiation stage. At this stage of stand development, some of the main canopy trees periodically produce large quantities of seed. Large crowns are important determinants of acorn production (Chapter 9) and thus the establishment of oak seedlings (Fig. 3.8). Moreover, when large trees die they create larger canopy openings than those created during earlier stages of stand development. The crowns of large main canopy trees also expand more slowly than they did during earlier stages of stand development. Consequently, canopy gaps remain open for longer periods, and this often results in light intensities near the forest floor that are sufficient for seedling establishment and growth. The consequent spatial heterogeneity of the main canopy creates spatial variation in the amount of light reaching the forest floor. This produces microenvironments favourable for the establishment and growth of oak and other tree reproduction. Established trees in subordinate crown positions also benefit from the increased growing space when large overstorey trees die. In oak forests, the onset of the understorey reinitiation stage typically occurs one or two decades before the end of a silvicultural rotation, i.e. stand age at final harvest, which commonly ranges from 80 to 120 years in the eastern United States. Progression to this stage of development may be accelerated by silvicultural actions or by natural disturbances. The establishment of oak reproduction during the understorey reinitiation stage is usually a prerequisite to successful oak regeneration (Chapter 3). The tree reproduction, shrubs and herbaceous vegetation that develop during this period set the stage for regeneration cuttings. If the stand is not regenerated through timber harvesting or other disturbances, some of the reproduction established during this stage of development may eventually grow into the overstorey as the stand approaches the next

209

stage of development – the complex stage. The successional replacement of oaks by more shade tolerant species is one of the most pervasive problems associated with oak silviculture in mesic and hydric forests. The understorey reinitiation stage is therefore a critical time for intervening silviculturally if the objective is to maintain or increase the proportion of oak in the future stand (Chapter 7). However, not all oak forests are successional to non-oaks. In the Ozark Highlands and similar xeric oak forests in the eastern United States, the successional displacement of oaks is limited by the inability of other hardwoods to persist in the superior crown classes beyond the stand initiation stage. Although non-oaks on these sites may aggressively fill canopy gaps immediately after disturbance, their dominance is usually shortlived. In the Ozark Highlands, differences among species in their competitive capacity are reflected in their probabilities of attaining an intermediate-or-better crown class after clearcutting. For a given initial (preharvest) basal diameter, these probabilities are higher for oaks than for other hardwoods 15 years after cutting (Fig. 5.7A). By that time, oaks dominate the stands and species are stratified into well-defined crown classes (Fig. 5.7B). This outcome is the result of the collective influence of initial floristics, overstorey inhibition and competitive sorting processes (Chapter 3) that control secondary succession in oak-dominated ecosystems such as the Ozark Highlands. Oak forests in other ecoregions may develop differently. An example is provided by a composite analysis of four mixed oak–hardwood stands in Connecticut during the stem exclusion and understorey reinitiation stages (Ward et al., 1999). When first measured, the stands were 25 years old, compositionally similar, and oaks comprised 40% of the basal area. Other major species included red maple, yellow birch, black birch and white ash. Mean site index was 67 ft at age 50. At age 25, the stands were well into the stem exclusion stage of development. During the next 70 years, numbers of trees decreased

210

Chapter 5

0.8 A

P

Oaks

0.6

Hickories and Blackgum

0.4

Sassafras

0.2

Flowering dogwood

0 0

0.4

0.8

1.2

1.6

2

2.4

Preharvest basal diameter of reproduction (inches) B O O

H

O O

BG O

O S

BG

D

BG

S D

S

S D O

O = Oaks H = Hickories BG = Blackgum S = Sassafras D = Flowering dogwood Fig. 5.7. (A) The estimated probability (P) that 15 years after overstorey removal a genet of advance reproduction (i.e. an individual seedling or seedling-sprout) will occupy an intermediate-or-greater crown class in upland oak stands in the Ozark Highlands of Missouri. Size of advance reproduction (basal diameter) at the time of clearcutting or other major disturbance is a key determinant of future stand composition. Estimates of P are based on logistic regression equations for average sites. (B) Vertical stratification of species after 15 years illustrates the probabilities in (A). Oaks are predominant in the upper crown classes while non-oaks are largely relegated to the sub-canopy. The four species groups shown represent the most frequently occurring hardwood species within this ecological region. Oaks are predominantly black, white, scarlet and post oaks. (From authors’ data.)

from 1400 to 600 acre−1 while basal area simultaneously increased from 70 to 120 ft2 acre1 (Fig. 5.8A and B). Drought com-

bined with defoliation by gypsy moth and canker worm reduced stand basal area between ages 55 and 65, which in turn


1600

130 A

B Basal area (ft2 acre–1)

1400 Trees per acre

211

1200 1000 800 600

120 110 100 90 80 70 60

400 20

40

60

80

100

20

40

Stand age (years)

60

80

100

Stand age (years)

130

50 C

D

(95) Red oaks

(95)

120

(85)

100 (55) 90



40 110

(75) (35)

(65)

80

(55) (95) Birches

30

20

(55)

Maples

(25)

(25)

(95)

(25) 60 500

(25)

White oaks

10

70

(95) (25)

0 700

900

1100

Trees per acre

1300

1500

0

100

200

300

400

Trees per acre

Fig. 5.8. Seventy-year change in number and basal area of trees ≥0.5 inches dbh in unthinned oak–mixed hardwood stands in Connecticut. The values shown are averages of four stands of similar composition and structure. (A) Number of trees. Stands were in the stem exclusion stage of development until age 55. Between ages 65 and 75, the number of trees increased due to recruitment into canopy openings resulting from tree mortality caused by gypsy moth and canker worm defoliation in the previous decade (ages 55–65). (B) Stand basal area. Oaks comprised 37% of stand basal area at stand age 25 and 45% by age 95. The decrease in basal area between ages 55 and 65 was caused by tree mortality. (C) Trajectories over time in the relation between basal area and number of trees for all species combined. Stand age at each point is shown in parentheses. (D) Trajectories for four species groups. Canopy gaps created between ages 55 and 65 accelerated the recruitment of new trees into the overstorey between ages 65 and 75 and marked the beginning of the understorey reinitiation stage of stand development. The number of birches and maples increased after canopy gaps were formed. Basal area of red oaks declined temporarily after age 55, but then increased from age 65 to 95. By age 95, red oaks dominated that stand in terms of basal area, but comprised fewer than 46 trees per acre. White oaks declined in both number of trees and basal area. The red oaks include northern red, black, and scarlet oaks; white oaks include white and chestnut oaks; birches include yellow, black and paper birches; maples include red and sugar maples. (Adapted from Ward et al., 1999; additional data courtesy of Jeffrey S. Ward, Connecticut Agricultural Experiment Station, used with permission.)

212

Chapter 5

accelerated the establishment and ingrowth of reproduction into the overstorey. The total number of overstorey trees increased in the following decade. Defoliation by gypsy moth and elm spanworm occurred episodically after age 65, but those events did not reduce total stand basal area. Both the absolute and relative basal area of the white oak group decreased markedly between stand ages 55 and 65 after insect defoliation and drought. Although those disturbances were not sufficient to initiate a new stand, they did alter species composition by creating conditions that increased numbers of maples and birches. Decreases in basal area and subsequent increases in numbers of trees between ages 55 and 75 are evident from the trajectory of change for all species combined (Fig. 5.8C). Trajectories for individual species-groups reveal a sharp increase in numbers of birches and maples between ages 65 and 75 following the decrease in oak basal area a decade earlier (Fig. 5.8D). The diameter distribution of all trees combined retained a reverse J-shape for the entire 70-year observation period (Fig. 5.9). Like the bottomland oak stand in Fig. 5.6, the shape of the diameter distribution flattened as the stand aged. The number of stems in the smallest diameter class decreased by 600 per acre over 70 years; numbers of trees in the largest diameter classes increased by 10–15 per acre over the same period. At stand age 25, the diameter distribution of the oaks was similar to the distribution of all species combined. But between ages 35 and 65, the oak diameter distributions became increasingly bell-shaped (Fig. 5.9). During that time, oaks decreased from 320 to 55 per acre and remained at approximately 50 per acre until age 95. Although the largest oaks increased in size, few smaller oaks grew into the overstorey. The range of oak diameters also increased over time. By age 95, oaks were the largest trees present and occurred in all diameter classes. Numbers of oaks in the smallest diameter classes increased modestly after age 65 (understorey reinitiation stage), but

birch, maple and beech were 4–12 times more abundant than the oaks in the small diameter classes. Although numbers of birches and maples per acre continually increased over time as a proportion of the total number of trees, the oaks collectively showed the greatest relative increase in basal area. White oak declined in both numbers of trees and basal area following the disturbances occurring between ages 55 and 65. The diameter distribution of oaks present at the time of stand establishment thus formed a continually changing population in terms of numbers of trees and basal area. Although the changes in the size distribution of the oak component of the stand are consistent with expectations for unthinned oak forests based on normal stand tables (Fig. 5.4; Schnur, 1937), they differ fundamentally from the overall diameter distribution of the composite stand (Fig. 5.9). This and the preceding example of stand development (Fig. 5.6) illustrate how species composition and stand structure of oak forests change over time and how those changes can differ among oak forests with different species mixtures, site characteristics and disturbance histories. In both examples, oaks were not the predominant species at the start of the stem exclusion stage of stand development. However, by persistently capturing growing space, the largest oaks were able to maintain rapid diameter growth and increase in basal area relative to other species present. Although reverse J-shaped diameter distributions are most often associated with uneven-aged forests, they also occur in certain even-aged forests – including those beyond the stem exclusion stage of development. These stands often originate following the disturbance of stands comprised of both shade tolerant and shade intolerant species. In subsequent stages of stand development, reverse J-shaped diameter distributions may evolve through the ingrowth of shade tolerant reproduction into the overstorey. Meanwhile, the less tolerant oaks may develop a bell-shaped diameter distribution embedded within the whole-stand diameter distribution.


Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

1027 500

All trees Age 25

400

182 80

213

All oaks Age 25

60

300

40

200

20

100 0 500

Age 35

400

0 80

Age 35

60

300

40

200

20

100 0 500

Age 55

400

0 80

Age 55

60

300

40

200

20

100 0 500

Age 65

400

0 80

Age 65

60

300

40

200

20

100 0 500

Age 75

400

0 80

Age 75

60

300

40

200

20

100 0 500

Age 85

400

0 80

Age 85

60

300

40

200

20

100 0 500

Age 95

400

0 80

Age 95

60

300

40

200

20

100

0

0 0

4

8 12 16 20 24 Dbh (inches)

0

4

8 12 16 20 24 Dbh (inches)

Fig. 5.9. Seventy-year change in diameter distributions in unthinned oak–mixed hardwood stands in Connecticut. Values are based on the average of four stands of similar composition and structure. The left-hand column of graphs shows all trees and the right-hand column shows the oaks. Stands were in the stem exclusion stage of development until age 55. Basal area decreased between ages 55 and 65 due to mortality from gypsy moth and canker worm defoliation. During the next decade, the number of trees per acre increased as new trees became established after the reduction in stand density. Trees per acre are plotted over the midpoints of 2-inch dbh classes. (Adapted from Ward et al., 1999; additional data courtesy of Jeffrey S. Ward, Connecticut Agricultural Experiment Station, used with permission.)

214

Chapter 5

Although the diameter distribution changes for the entire stand are somewhat predictable and are useful for describing patterns of stand development, they explain little about the interactive relations between oaks and other species. Predicting the outcome of the competitive struggle between oaks and associated species is especially important when the co-occurring species differ greatly in growth rate and silvical characteristics. Determining the silvicultural prescription that assures the desired representation of oaks in the future stand is a common silvicultural problem in tending young, even-aged stands comprised of oaks mixed with other species. An appropriate silvicultural action requires an understanding of how species interact within shared environments. Large differences in rates of height growth of two desirable, long-lived species growing on the same site can create difficulties in maintaining both species. In young stands, species’ differences in height growth may not always be apparent from the current stand structure because those differences may not yet be expressed. And in older stands, the disturbance history that produced the height differences among species may not be evident. Comparing the site index of one species with another is one way to assess the ecological and silvicultural significance of species’ height growth differences (Chapter 4). Differences in site index as small as 5 ft may ultimately result in the suppression of the slower-growing species, assuming both species originate about the same time. Such differences, in turn, may limit silvicultural options. Large growth-rate differences between an oak and a co-occurring species may be silviculturally problematic if the oak is the slower-growing species and both species are to be grown to sawlog size, or if the primary objective is to sustain long-term acorn production. However, site index comparisons, by themselves, do not always predict how co-occurring species will interact. Virtually all oak stands eventually become vertically stratified by species, and even for the same initial species mix, patterns of height stratification may differ. The development of various species mixtures

provides examples of how vertical stratification can differ among ecosystems. For example, the early pattern of height growth of co-occurring northern red oak, black birch and red maple in young even-aged stands in New England was similar. Not until the end of the third decade did northern red oak emerge as the dominant species (Oliver, 1978) (Fig. 5.10A). A different pattern of stratification occurs in the mixed forest type that is transitional between the Allegheny Plateau and the Northern Hardwood Region to the north. There, forests are commonly, but only temporarily, dominated by northern red oak. The oaks and other intolerant and mid-tolerant species retain their bell-shaped distribution as they develop (Stout, 1991). In contrast, sugar maple and other tolerant species form reverse J-shaped diameter distributions as their reproduction accumulates in the understorey, which ultimately leads to the successional displacement of the oaks. In co-occurring cherrybark oak and sweetgum stands in minor river bottoms in Mississippi, it takes about 40 years for the oak to dominate the sweetgum (Clatterbuck and Hodges, 1988). The oak thereafter grows faster when directly competing with sweetgum than when not competing with sweetgum (Fig. 5.10B). This effect may be partially related to differences in the crown shape of the two species. Whereas oak crowns are wider at the top than the bottom (excurrent in shape), sweetgum crowns are the opposite (decurrent in shape). Because the height growth of the two species continues to diverge with age, there is minimal interference between sweetgum crowns and oak crowns. The persistence of the sweetgum also promotes long, clear boles on the oaks. Cherrybark oaks growing in competition with sycamore do not fare as well. The extremely rapid height growth and somewhat excurrent crown of the sycamores quickly suppress the oaks unless 20 ft or more separates the two species (Fig. 5.10C). Similar interspecific relations occur when scarlet and white oaks grow in competition with yellow-poplar (Fig. 5.10D). Although these examples are generally consistent with


A

70 Northern red oak

50

120 Black birch

Red oak

40 Red maple

30 20

Close spacing

B

Black birch Red maple

Tree height (ft)

Tree height (ft)

60

Sweetgum (close spacing)

Wide spacing

Cherrybark oak

Oak (close spacing) 100

Cherrybark oak Sweetgum

80

Nearest competitor

Oak (wide spacing) 60

10

215

>18ft

0 0

10 20 30 40 50 60

40 10 20 30 40 50 60 70 Stand age (years)

Stand age (years)

Dominant

60

Codominant

20 0

Yellow-poplar stand 1 100 stand 2

Sycamore Cherrybark oak

Intermediate

Suppressed

0 4 8 12 16 20 24 Stand age (years)

Tree height (ft)

Tree height (ft)

Sycamore Cherrybark 80 oak

40

120

C Crown class

100

Stand 1

80 60 40 20

0 10 20 30 40 Distance (ft)

D Stand 2 Yellowpoplar

Scarlet oak Scarlet oak stand 1

White oak

White oak stand 1

0 1900 1940 1980 1920 1960 Year of origin

Fig. 5.10. Height growth and vertical stratification of oaks in mixed-species stands. (A) Even-aged stands of northern red oak, red maple and black birch in New England. Red oaks emerge as dominants 20–30 years after disturbance. (Adapted from Oliver, 1978.) (B) Even-aged stands of cherrybark oak and sweetgum in minor river bottoms in central Mississippi. Close spacing between sweetgum and oak accelerates the height growth of oaks; wide spacing results in slower height growth and shorter clear-bole length of oaks. (Adapted from Clatterbuck and Hodges, 1988.) (C) Plantation-grown cherrybark oak and American sycamore in a minor river bottom in Arkansas. Close spacing between oaks and sycamores results in suppression or reduced height growth of oaks. (Adapted from Oliver et al., 1990.) (D) Representative trees in two natural stands of mixed oak and yellow-poplar of unspecified age structure in the North Carolina Piedmont. In this region, yellow-poplar outgrows oaks where yellow-poplar site index exceeds 65 ft. (Adapted from O’Hara, 1986, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

site index comparisons between species, site index alone does not explain the magnitude of such species-specific interactions.

The complex stage In the absence of timber harvesting or other exogenous disturbances that eliminate the overstorey, even-aged stands progress toward the complex stage (Oliver, 1997)

(Figs 5.2 and 5.3D). This stage evolves as the consequence of the natural mortality of large overstorey trees that create canopy gaps that occur at irregular times and locations within a stand. Because these gaps are often large, crown expansion of trees adjacent to a gap is often insufficient to fill the gap. This lag in crown closure allows sub-canopy trees and established reproduction to fill gaps. As new canopy gaps occur, they are filled by new age classes of trees.

216

Chapter 5

As this gap-filling process continues in the absence of a stand-initiating disturbance, an uneven-aged stand eventually evolves. The complex stage of stand development includes old-growth forests, which was the original term applied to this stage (Oliver and Larson, 1996). However, there is a basis for distinguishing between the two. All old-growth forests are complex, but not all complex forests are old-growth. Definitions of old-growth oak forests are usually based on overstorey age, stand disturbance history and structural characteristics such as the presence of old trees, snags and down wood (see Chapter 9; also Meyer, 1986; Parker, 1989; Martin, 1992; Batista and Platt, 1997; Greenberg et al., 1997; Murphy and Nowacki, 1997; Tyrell et al., 1998). The various definitions of old-growth all assume that human influences in such forests have been minimal. However, older second-growth forests (managed and unmanaged) may have complex structures without meeting the strict definition of old growth.

Development of Uneven-aged Stands As an even-aged stand advances towards the complex stage of development, it gradually evolves into a multi-aged population as a result of stand maturation and associated gap formation and filling. Canopy gaps become more numerous until the stand forms a mosaic of old trees and gaps filled with younger trees of various ages and species. As trees refill the gaps, standwide diameter distributions gradually change. In the absence of a stand-initiating disturbance, the overall diameter frequency distribution may change from bell-shaped to irregularly shaped or reverse J-shaped. Regardless of the shape of the diameter distribution, the normal evolution of stand structure from even- to uneven-aged eventually produces an uneven-aged collection of dispersed, even-aged groups of trees, each occupying a relatively small area. Over time, the various tree age classes become visually indiscernible.

The resulting uneven-aged stand structure can be generalized from the diameter distributions of a time series of even-aged stands consisting of 10-year age classes with each class comprising an equal area (Fig. 5.4). If numbers of trees in each age class are summed across all diameter classes, numbers of trees in successively larger diameter classes decrease exponentially (Fig. 5.4 inset). This type of diameter distribution therefore could result from either a collection of even-aged stands or from a single uneven-aged stand. However, in the latter, the even-aged spatial units are so small and intermingled that they are largely indistinguishable. During the complex stage of development, differences in shade tolerances and growth rates among the oaks themselves may produce gradual changes in the proportions of oaks among the crown classes and thus corresponding shifts in species composition. For example, through the stem exclusion stage, the relatively shadetolerant and slow growing white oak tends to lag behind co-occurring red oaks in capturing growing space. However, during the understorey reinitiation and complex stages of development, the white oaks that persist in the inferior crown classes gradually ascend to canopy dominance as canopy gaps are formed (Shifley et al., 1995; Spetich, 1995). Although old-growth forests comprised of oaks mixed with other species often form reverse J-shaped diameter distributions, the oaks themselves may not conform to that distribution. Structural and compositional changes during the complex stage largely depend on differences in the rates at which co-occurring species of tree reproduction are recruited into the overstorey and how persistent they are there. Where non-oaks fill most of the canopy gaps created during the complex stage, those species will eventually predominate among the smaller dbh classes. Depending on site quality and other factors, they may or may not ascend to dominance, inhibit oak regeneration or successionally displace the oaks. In the absence of disturbances that reinitiate the establishment and


217

eter distributions in this ecosystem and their silvicultural maintenance largely depend on the species composition and density of stands (Chapter 8).

development of oak reproduction, a bellshaped diameter distribution of oaks embedded within an overall reverse Jshaped distribution nevertheless is a harbinger of the impending successional displacement of the oaks (Fig. 5.11). In the Ozark Highlands, crown stratification among the oaks persists into the complex stage (Shifley et al., 1995). Large numbers of oaks in the smaller diameter classes, even in old-growth stands, reflect the oak’s permanence in these forests (Fig. 5.12A and B). In mature, relatively undisturbed second-growth forests, diameter distributions approach a reverse J-shape (Fig. 5.12C and D). For all species combined, the reverse J-shape is even more pronounced because of the high density of permanent subcanopy species such as flowering dogwood and blackgum (Fig. 5.12 inset graphs). Except during the stand initiation stage, the non-oaks are largely relegated to the sub-canopy (Dey et al., 1996). The specific characteristics of diam-

Disturbance–Recovery Cycles When forest disturbance is limited to gapscale events, stand development follows the sequence illustrated in Fig. 5.2. However, stand-initiating events that eliminate all or most of the overstorey can occur during any stage of stand development. These events return stands to the stand initiation stage of development. In contrast, incomplete stand-scale disturbances may eliminate only a portion of the overstorey and leave significant numbers of trees standing. Although incomplete stand-scale disturbances change the stage of stand development, they do not return the stand to the stand initiation stage. Rather, they create a mixed stage of stand

80 3 Trees per acre (oaks)

Trees per acre (all species)

70 60 50 40

2

1

0 30

4

8 12 16 20 24 28 32 36 Dbh class (inches)

20 10 0 4

8

12

16

20

24

28

32

36

Dbh class (inches) Fig. 5.11. The diameter distribution of trees in Spitler Woods, an old-growth forest in central Illinois. The bell-shaped (normal) diameter distribution of the oaks (inset) is obscured by the reverse J-shaped distribution of all tree species. This mesic forest is dominated by sugar maple, oaks and American basswood. The oaks (white, northern red, black, chinkapin, bur and shingle oaks) comprise 33% of stand basal area but only 6% of trees. Other important species include elms, hickories, hackberry, Ohio buckeye, black walnut and white ash. (Author’s data.)

Chapter 5

50 Red oaks

30

Oak stocking = 69%

White oaks

20

B

400

Stocking = 101%

300 200

2

6 10 14 18 22 26 Dbh class (inches)

10

20

400 Stocking = 103%

300 200 100 0 2


10

0 2

6

10

14

18

22

26

0 2

6

10

14

18

22

26

50 C

40

Trees per acre (total)

50 Trees per acre (oaks)

Oak stocking = 71%

30

100 0

40


Trees per acre (oaks)

40


50 A

D

400

Oak stocking = 64%

30

300

Stocking = 103%

200

400

Oak stocking = 71%

30

100

20

40


218

300

Stocking = 87%

200 100

0 2 6 10 14 18 22 26 Dbh class (inches)

10

20

0 2


10

0

0 2

6

10

14

18

22

26

Dbh class (inches)

2

6

10

14

18

22

26

Dbh class (inches)

Fig. 5.12. Diameter distributions of oaks and all trees (inset graphs) in four relatively undisturbed oak forests of the Ozark Highlands of Missouri. (A) A 120-acre old-growth stand with trees up to 200–250 years old; human disturbance has been limited to periodic fires occurring more than 40 years ago. (B) A 330-acre old-growth stand with trees up to 140 years old (a few exceed 200 years). There is little evidence of disturbance during the last 40 years. (C) A second-growth 4000-acre forest dominated by 70- to 90-year-old oaks with some shortleaf pine. Stands were disturbed by fire, grazing and logging 45 years prior. (D) A second-growth 3000-acre forest disturbed by fire, grazing and logging 45 years prior. Intensity/recentness of disturbance (increasing from A to D) is reflected in the increasing number of oaks in the smaller diameter classes. All four forests form mosaics of xeric to xero-mesic associations dominated by various mixtures of oaks, hickories and shortleaf pine growing on well- to excessively drained cherty soils. Red oaks include black, scarlet, southern red and northern red oaks; white oaks include white, post and chinkapin oaks. Important subordinate species include flowering dogwood, blackgum, sassafras, elms and red maple. (From Shifley et al., 1995.)

development. Mixed-stage stands resulting from natural events often form mosaics of younger trees developing in large canopy openings interspersed with patches of older trees. They often form irregularly spaced tree populations of variable size and age structure. Mixed-stage stands also can result from various types of timber harvesting.

Stands in the mixed stage of development are distinguishable from stands in other stages by: (i) the spatial scale of disturbance, which is greater than gap scale, and (ii) stand density, which is below average maximum density (see Chapter 6). Relatively young trees also may predominate in stands during the mixed stage. Low density, highly disturbed stands of all descriptions therefore fall into


the mixed stage of development. Examples of stands in the mixed stage include oak savannas (Chapter 9), stands resulting from indiscriminate timber harvests (e.g. ‘high-grading’), as well as some designed silvicultural practices. The latter include heavily thinned stands, low-density shelterwoods and stands managed by group selection or single tree selection at low stand densities (Chapters 7 and 8). Oak forests in the mixed stage of development are ubiquitous because the events that create them are so common. The complete spectrum of forest developmental stages thus includes the mixed stage plus the four stages previously defined. Collectively they represent points in a potentially endless series of disturbance–recovery cycles initiated by standscale and gap-scale disturbances. These cycles follow specific sequences determined by the developmental stage of the stand at the time of disturbance and the type and spatial scale of disturbance (Fig. 5.13). Disturbance and recovery cycles provide a conceptual framework for ecological process and silvicultural practice. Silviculture involves the planned application of controlled disturbances to create and maintain (at least temporarily) specific species compositions, age distributions and size distributions. An ecological understanding of disturbance–recovery cycles is central to achieving silvicultural objectives. However, silvicultural control of stand composition and structure is often complicated by unpredictable natural disturbance events that lie beyond the direct control of the silviculturist. Even when forests are intensively managed, unwelcome disturbances are a part of silvicultural reality. Frequent stand-initiating disturbances can maintain a stand in the stand initiation stage indefinitely. Such disturbance regimes were common and extensive in many regions of North America before European settlement. For example, in the prairie–forest border region of southwestern Wisconsin, plant communities described as ‘oak scrub’ failed to develop beyond the sapling stage because of frequent wildfires that recurred for centuries (Marks, 1942; Grimm, 1984; Chumbley

219

et al., 1990). As long as these fires persisted, the stands of oak scrub persisted. The fires arrested succession, and thus maintained stands in the stand initiation stage of development. But by the early 20th century, wildfires were controlled and the oak scrubs quickly developed into closed canopy forests (Marks, 1942; Curtis, 1959). An incomplete stand-scale disturbance will transform most stands into the mixed stage of development. Exceptions are young, even-aged stands that are in the stand initiation stage of development. They remain in the stand initiation stage of development following incomplete standscale disturbances because the existing trees and the new trees recruited following the disturbance are within the same age class. If a stand is already in the mixed stage of development, frequent incomplete stand-scale disturbances (such as the fires that historically maintained oak savannas) may keep it in that state indefinitely. Not all stands progress through the stages of development according to the illustrated sequence (Fig. 5.13). For example, oak chaparral and other shrub-like oak communities may not conform to the five-stage model presented. Disturbances are of many types and intensities, and follow many temporal–spatial patterns. Even distinguishing between gap-scale and stand-scale disturbances is artificial because real-world disturbance scales are continuous. Disturbances also may affect only some parts of a stand, or two types of disturbances (e.g. windthrow and disease-induced mortality) may affect different parts of a stand simultaneously. Some forests are also more resilient than others, i.e. they have a greater capacity to return to their previous state after disturbance. Consequently, oak forests differ in their disturbance recovery rates and in related changes in stand structure and species composition. The 35-year history of an oak–hickory stand in the Ozark Highlands of Missouri provides an example of disturbance and recovery in that ecoregion. The stand is part of a 160,000-acre forest that has been managed by the single tree selection system since exploitative timber harvesting reduced aver-

220

Chapter 5



Mixed stage

Understorey reinitiation stage

Complex stage

Key: Type of disturbance

Stand-scale: stand-initiating

Stand-scale: incomplete

Gap-scale

Fig. 5.13. Types of forest disturbances and their relation to stages of stand development and disturbance–recovery cycles. Stands in all developmental stages are subject to gap-scale disturbances originating from the natural mortality of individual trees (dotted arrows). As average tree size (and thus gap size) increases with increasing stand age, stands progress towards the uneven-aged state (complex stage) via the gap-wise replacement of main canopy trees by sub-canopy trees and reproduction. However, stand-scale disturbances can occur at any stage of stand development. When stand-initiating disturbances eliminate all or most of the overstorey, stands abruptly return to the stand initiation stage (solid arrows). Disturbances that eliminate only a fraction of the overstorey, but leave a significant number of trees standing, are termed incomplete stand-scale disturbances. These disturbances produce a mixed stage of stand development comprised of both new and older trees that develop into multi-tiered stands of irregular age structure (dashed arrows). The mixed stage may advance to the complex stage in the absence of further stand-initiating or incomplete stand-scale disturbances, return to the stand initiation stage after a stand-scale disturbance, or remain in the mixed stage as a result of further incomplete stand-scale disturbances.


age forest-wide stand densities to low levels in the early 1950s (Loewenstein et al., 1995). The recovery of this forest is illustrated by data from four 0.2-acre plots in a single stand that was inventoried every 5 years from 1957 to 1992. During that time, numbers of trees ≥5 inches dbh doubled and stand basal area increased from 33 to 73 ft2 acre1 (Fig. 5.14A and B). This increase occurred even though 17 ft2 of basal area was harvested between 1972 and 1977. During this 35-year period, oaks maintained a nearly constant proportion (48%) of total basal area. Changes in stand structure and species composition are described by changes in basal area and numbers of trees per acre for all species combined and for individual species or groups (Fig. 5.14C and D). Although the number of trees and basal area were reduced by the 1972–1977 timber harvest, the stand quickly resumed its preharvest trajectory of change (Fig. 5.14C). However, within the overall pattern of stand development, there was great variation among species in recovery responses. Whereas numbers of trees and basal areas of all species increased between 1957 and 1972, the 1972–1977 timber harvest reduced the number of large black and scarlet oaks proportionately more than other species. Black and scarlet oak basal areas nevertheless remained relatively constant from 1977 to 1992 (Fig. 5.14D). In contrast, the white oaks and shortleaf pine increased in basal area relative to black and scarlet oaks. The continual increase in the basal area of white oak on the study plots paralleled forestwide changes that occurred during the same period (Loewenstein, 1996; Wang, 1997). Patterns of recovery from disturbance were also evident from changes in diameter distributions. Even though the distributions of all species combined retained a reverse Jshape during the 35-year period, the distributions of the oaks did not (Fig. 5.15). From 1957 to 1972, the oak diameter distributions formed two or more peaks. Oaks in the smallest (6 inch) dbh class nevertheless were more numerous than in any other class throughout the observed period. For the oaks, reverse J-shaped distributions became increasingly prominent with time.

221

Over time, numbers of white oaks increased in the smaller diameter classes and by 1992 white oak dominated the 6–8 inch dbh classes (Fig. 5.15). This change reflects white oak’s shade tolerance and its capacity in this ecosystem to sustain rates of ingrowth into successively larger diameter classes in numbers sufficient to maintain a reverse J-shaped diameter distribution. The reduction in stand density to 47 ft2 of basal area per acre between 1972 and 1977 was an important factor in sustaining white oak ingrowth and thus in perpetuating an uneven-aged stand structure. The less shade tolerant black and scarlet oaks decreased in abundance in the smaller diameter classes although they retained their earlier representation in the larger diameter classes. Stand density, diameter distributions and species composition in this uneven-aged stand thus changed continually during the 35year period. The forest’s resilience is reflected in its rapid rate of recovery from exploitative timber harvests in the 1950s and a silviculturally designed harvest in the 1970s. Since then, oaks have maintained their dominance, stand basal area has continued to increase and a reverse J-shaped diameter distribution consistent with sustaining an uneven-aged stand structure has developed. Comprehensive analyses of this forest have similarly indicated that continued application of the single tree or group selection method is likely to sustain the oak’s dominance (Loewenstein, 1996; Wang, 1997). However, successful application of the method in this and ecologically similar forests will require reducing stocking to about 50 ft2 of basal area per acre every 15–20 years in order to sustain adequate rates of oak recruitment into the overstorey (Larsen et al., 1999). In effect, this requires using silvicultural control of stand structure and density to suspend stands in the mixed stage of development. From this it should not be inferred that oak forests are generally amenable to uneven-aged silviculture. On the contrary, the intrinsic oak regeneration characteristics of an ecosystem largely determine its capacity to sustain uneven-aged oak populations and thus its suitability to uneven-aged management (see Chapters 3 and 8).

222

Chapter 5

80

180 A

B Basal area (ft2 acre–1)

Trees per acre

160 140 120 100

70

60

50

40

80 1955

1965

1975

1985

30 1955

1995

1965

1975

Year

1995

Year

40

80 C

D

(92) Basal area (ft2 acre–1)


1985

70 (72) (87)

60

(82)

(67)

50

(77)

(62) 40

Shortleaf pine (72)

120

140

Trees per acre

160

180

(57)

(57) (72)

10

White oaks

Hickories

0 100

(92)

Black and scarlet oaks (92)

20

(57)

30

(72)

30

(57) 80

(92)

0

20

40

60

80

Trees per acre

Fig. 5.14. Change in number and basal area of trees ≥5 inches dbh during a 35-year period in an oak stand managed by the single-tree selection method in the Ozark Highlands of Missouri. The stand is dominated by black, scarlet and white oaks; black oak site index ranges from approximately 55 to 65 ft at an index age of 50 years. The stand density was greatly reduced by an exploitative harvest prior to 1957 which placed the stand in the mixed stage of development (Fig. 5.13). This sequence of stand development illustrates recovery from that disturbance and response to a designed silvicultural disturbance between 1972 and 1977. Number of trees (A) and basal area (B) increased during all intervals except 1972–1977 when a timber harvest removed some trees 15 inches dbh and larger. (C) Trajectories over time in the relation between basal area and number of trees for all species combined. The plotted line traces changes from 1957 to 1992 (year shown in parentheses). The stand was in the mixed stage of development during the 35-year period, and low stand basal area facilitated recruitment of trees into the overstorey. (D) Trajectories for the four species groups. The timber harvest reduced the number and basal areas of black/scarlet oaks and shortleaf pine. Although drought-induced mortality after 1977 further reduced the number and basal area of black and scarlet oaks, by 1992 values were similar to those in 1962. Both the number and basal area of white oaks (white and post oaks combined) continually increased over time. (Author’s analysis; data courtesy of Pioneer Forest, Salem, Missouri.)


Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

Trees per acre

All trees

223

Oaks 1957

1957

60

60

40

40

20

20

0

White/post oak Black/scarlet oak

0 1962

1962

60

60

40

40

20

20

0

0 1967

1967

60

60

40

40

20

20

0

0 1972

1972

60

60

40

40

20

20

0

0 1977

1977

60

60

40

40

20

20

0

0 1982

1982

60

60

40

40

20

20

0

0 1987

1987

60

60

40

40

20

20

0

0 1992

1992

60

60

40

40

20

20

0

0 6

10 14 18 22 Dbh (inches)

26

6

10


26

Fig. 5.15. Thirty-five-year change in diameter distributions in an uneven-aged oak forest in the Ozark Highlands of Missouri managed by the single-tree selection method. The diameter distributions for all species (left-hand column of graphs) maintained a reverse J-shape throughout the period. However, the number of trees in the smaller diameter classes continually increased during the period. By 1992, white oaks dominated the 6- to 10-inch dbh classes. The number of black and scarlet oaks in the smaller diameter classes increased between 1957 and 1972 but thereafter declined. (Data courtesy of Pioneer Forest, Salem, Missouri.)

224

Chapter 5

References Abrams, M.D. and Downs, J.A. (1990) Successional replacement of old-growth white oak by mixed mesophytic hardwoods in southwestern Pennsylvania. Canadian Journal of Forest Research 20, 1864–1870. Abrams, M.D. and Nowacki, G.J. (1992) Historical variation in fire, oak recruitment, and post-logging accelerated succession in central Pennsylvania. Bulletin of the Torrey Botanical Club 119, 19–28. Abrams, M.D. and Scott, M.L. (1989) Disturbance-mediated accelerated succession in two Michigan forest types. Forest Science 35, 42–49. Batista, W.B. and Platt, W.J. (1997) An old-growth definition for southern mixed hardwood forests. USDA Forest Service General Technical Report SRS SRS-9. Beck, D.E. and Hooper, R.M. (1986) Development of a southern Appalachian hardwood stand after clearcutting. Southern Journal of Applied Forestry 10, 168–172. Bormann, F.H. and Likens, G.E. (1979) Pattern and Process in a Forested Ecosystem. Springer-Verlag, New York. Chumbley, C.A., Baker, R.G. and Bettis, E.A. III. (1990) Midwestern holocene paleoenvironments revealed by floodplain deposits in northeastern Iowa. Science 249, 272–274. Clatterbuck, W.K. and Hodges, J.D. (1988) Development of cherrybark oak and sweet gum in mixed, even-aged bottomland stands in central Mississippi, USA. Canadian Journal of Forest Research 18, 12–18. Clements, F.E. (1916) Plant succession: An analysis of the development of vegetation. Carnegie Institute Washington Publication 242. Curtis, J.T. (1959) The Vegetation of Wisconsin. University of Wisconsin Press, Madison. Dey, D.C. (1991) A comprehensive Ozark regenerator. PhD dissertation, University of Missouri, Columbia. Dey, D.C., Ter-Mikaelian, M., Johnson, P.S. and Shifley, S.R. (1996) Users guide to ACORn: a comprehensive Ozark regeneration simulator. USDA Forest Service General Technical Report NC NC180. Egler, F.E. (1954) Vegetation science concepts I. Initial floristic composition. A factor in old-field vegetation management. Vegetatio 4, 412–417. Gevorkiantz, S.R. and Scholz, H.F. (1948) Timber yields and possible returns from the mixed-oak farmwoods of southwestern Wisconsin. USDA Forest Service Lake States Forest Experiment Station Publication 521. Gingrich, S.F. (1967) Measuring and evaluating stocking and stand density in upland hardwood forests in the Central States. Forest Science 13, 38–53. Gingrich, S.F. (1971) Management of young and intermediate stands of upland hardwoods. USDA Forest Service Research Paper NE NE-195. Greenberg, C.H., McLeod, D.E. and Loftis, D.L. (1997) An old-growth definition for western and mixed mesophytic forests. USDA Forest Service General Technical Report SRS SRS-16. Grimm, E.C. (1984) Fire and other factors controlling the Big Woods vegetation of Minnesota in the mid-nineteenth century. Ecological Monographs 54, 291–311. Harper, J.L. (1977) Population Biology of Plants. Academic Press, London. Helms, J.A. (ed.) (1998) The Dictionary of Forestry. Society of American Foresters, Bethesda, Maryland. Hibbs, D.E. (1982) Gap dynamics in a hemlock-hardwood forest. Canadian Journal of Forest Research 12, 522–527. Hibbs, D.E. (1983) Forty years of forest succession in central New England. Ecology 64, 1394–1401. Hilt, D.E. (1985) Where has all my yellow-poplar gone? Northern Journal of Applied Forestry 2, 67–69. Jenkins, M.A. and Parker, G.R. (1997) Changes in down dead wood volume across a chronosequence of silvicultural openings in southern Indiana forests. USDA Forest Service General Technical Report NC NC-188, pp. 162–169. Johnson, P.S. and Sander, I.L. (1987) Quantifying regeneration potentials of Quercus forests in the Missouri Ozarks. USDA Forest Service General Technical Report NC NC-120, Vol. 1, pp. 377–385. Johnson, R.L. and Deen, R.T. (1993) Prediction of oak regeneration in bottomland forests. USDA Forest Service General Technical Report SE SE-84, pp. 146–155.


225

Johnson, R.L. and Krinard, R.M. (1983) Regeneration in small and large sawtimber sweetgum–red oak stands following selection and seed tree harvest: 23-year results. Southern Journal of Applied Forestry 7, 176–184. Johnson, R.L. and Krinard, R.M. (1988) Growth and development of two sweetgum–red oak stands from origin through 29 years. Southern Journal of Applied Forestry 12, 73–78. Larsen, D.R., Loewenstein, E.F. and Johnson, P.S. (1999) Sustaining recruitment of oak reproduction in uneven-aged stands in the Ozark Highlands. USDA Forest Service General Technical Report NC NC-203. Loewenstein, E.F. (1996) An analysis of the size- and age-structure of a managed uneven-aged oak– forest. PhD dissertation, University of Missouri, Columbia. Loewenstein, E.F., Garrett, H.E., Johnson, P.S. and Dwyer, J.P. (1995) Changes in a Missouri Ozark oak hickory forest during 40 years of uneven-aged management. USDA Forest Service General Technical Report NE NE-197, pp. 159–164. Loftis, D.L. (1983) Regenerating red oak on productive sites in the southern Appalachians: a research approach. USDA Forest Service General Technical Report SE SE-24, pp. 144–150. Loftis, D.L. (1990) Predicting post-harvest performance of advance red oak reproduction in the southern Appalachians. Forest Science 36, 908–916. Marks, J.B. (1942) Land use and plant succession in Coon Valley, Wisconsin. Ecological Monographs 12, 114–133. Martin, W.H. (1992) Characteristics of old-growth mixed mesophytic forests. Natural Areas Journal 12, 127–135. Meyer, J. (1986) Management of old growth forests in Missouri. Missouri Department of Conservation Habitat Management Series 3. Murphy, P.A. and Nowacki, G.J. (1997) An old-growth definition for xeric pine and pine–oak woodlands. USDA Forest Service General Technical Report SRS SRS-7. O’Hara, K.L. (1986) Developmental patterns of residual oaks and oak and yellow-poplar regeneration after release in upland hardwood stands. Southern Journal of Applied Forestry 10, 244–248. Oliver, C.D. (1978) The development of northern red oak in mixed stands in Central New England. Yale University School of Forestry and Environmental Science Bulletin 91. Oliver, C.D. (1980) Even-aged development of mixed-species stands. Journal of Forestry 78, 201–203. Oliver, C.D. (1981) Forest development in North America following major disturbances. Forest Ecology and Management 3, 153–168. Oliver, C.D. (1997) Hardwood forest management in the United States: Alternatives for the future. Proceedings of the Annual Hardwood Symposium. National Hardwood Lumber Association, Memphis, Tennessee, 25, pp. 45–58. Oliver, C.D. and Larson, B.C. (1996) Forest Stand Dynamics. John Wiley & Sons, New York. Oliver, C.D. and Stephens, E.P. (1977) Reconstruction of a mixed species forest in central New England. Ecology 58, 562–572. Oliver, C.D., Clatterbuck, W.K. and Burkhardt, E.C. (1990) Spacing and stratification patterns of cherrybark oak and American sycamore in mixed, even-aged stands in the southeastern United States. Forest Ecology and Management 31, 67–79. Parker, G.R. (1989) Old-growth forests of the central hardwood region. Natural Areas Journal 9, 5–11. Reice, S.R. (1994) Nonequilibrium determinants of biological community structure. American Scientist 82, 424–435. Rogers, R. (1983) Spatial pattern and growth in a Missouri oak–hickory stand. PhD dissertation, University of Missouri, Columbia. Runkle, J.R. (1985) Disturbance regimes in temperate forests. In: Pickett, S.T.A. and White, P.S. (eds) The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, San Diego, California, pp. 17–33. Schnur, G.L. (1937) Yield, stand, and volume tables for even-aged upland oak forests. USDA Technical Bulletin 560. Shifley, S.R., Roovers, L.M. and Brookshire, B.L. (1995) Structural and compositional differences between old-growth and mature second-growth forests in the Missouri Ozarks. USDA Forest Service General Technical Report NE NE-197, pp. 23–36. Smith, D.M. (1986) The Practice of Silviculture, 8th edn. John Wiley & Sons, New York. Sousa, W.P. (1984) The role of disturbance in natural communities. Annual Review of Ecology and Systematics 15, 353–391.

226

Chapter 5

Spetich, M.A. (1995) Characteristics and spatial pattern of old-growth forests in the Midwest. PhD dissertation, Purdue University, West Lafayette, Indiana. Spetich, M.A., Shifley, S.R. and Parker, G.R. (1999) Regional distribution and dynamics of coarse woody debris in Midwestern old-growth forests. Forest Science 45, 302–313. Stout, S.L. (1991) Stand density, stand structure, and species composition in transition oak stands of northwestern Pennsylvania. USDA Forest Service General Technical Report NE NE-148, pp. 194–206. Trimble, G.R., Jr and Tryon, E.H. (1966) Crown encroachment into openings cut in Appalachian hardwood stands. Journal of Forestry 64, 104–108. Tyrell, L.E., Nowacki, G.J., Crow, T.R., Buckley, D.S., Nauertz, E.A., Niese, J.N., Rollinger, J.L. and Zasada, J.C. (1998) Information about old growth for selected forest type groups in the eastern United States. USDA Forest Service General Technical Report NC NC-197. Wang, Z. (1997) Stability and predictability of diameter distributions in a managed uneven-aged oak forest. PhD dissertation, University of Missouri, Columbia. Ward, J.S. and Stephens, G.R. (1994) Crown class transition rates of maturing northern red oak (Quercus rubra L.). Forest Science 40, 221–237. Ward, J.S., Anagnostakis, S.L. and Ferrandino, F.J. (1999) Stand dynamics in Connecticut hardwood forests, the old series plots (1927–1997). Connecticut Agricultural Experiment Station Bulletin 959. Waring, R.H. and Schlesinger, W.H. (1985) Forest Ecosystems: Concepts and Management. Academic Press, Orlando, Florida. Weigel, D.R. and Johnson, P.S. (1999) Planting red oak under oak/yellow-poplar shelterwoods: A provisional prescription. USDA Forest Service General Technical Report NC NC-210. White, P.S. and Pickett, S.T.A. (1985) Natural disturbance and patch dynamics: An introduction. In: Pickett, S.T.A. and White, P.S. (eds) The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, San Diego, California, pp. 3–13.

6 Self-thinning and Stand Density

Introduction Self-thinning is the natural process whereby numbers of trees per unit area decrease as average tree size increases over time. It is a process intrinsic not only to oak forests but to all forest and plant communities whose composition and structure are influenced by competition for growing space. Whereas self-thinning is a process, the term stand density refers to various expressions of the absolute or relative amounts of an attribute of tree populations (e.g. numbers of trees or stand basal area) per unit land area. As might be expected, the two concepts are closely connected. Together, they rank among the most important concepts in forest ecology and silviculture.

Self-thinning The principle of self-thinning is most easily described by the temporal changes that occur in the numbers of trees in undisturbed even-aged stands. However, selfthinning also occurs in uneven-aged stands. According to this principle, the finite growing space of a stand is occupied by progressively fewer trees as average tree size increases with stand age. Trees at a competitive disadvantage die from crowding and suppression as stands approach a limiting number of trees of a given average size that can coexist within an area. As stands reach the stem exclusion stage of development (Chapter 5), tree crowns

expand to fill the available growing space. Crown expansion continues until an upper limit of tree crowding is reached. Thereafter, stands follow a relatively predictable course of density-dependent tree mortality as numbers of trees per unit area decrease with increasing average tree size. It is generally assumed that the combined effects of crown expansion and tree mortality are compensatory so that canopy closure is always maintained except in the presence of ‘irregular’ mortality. The latter may be caused by such factors as air pollution, high winds, flooding, epidemic insect and disease outbreaks, and other factors.

Reineke’s model Reineke’s model for defining average maximum stand density expresses the negative relation between number of trees per unit area and average stand diameter in undisturbed, even-aged stands (Reineke, 1933). Plotting the logarithm of number of trees over the logarithm of mean stand diameter produces a straight line. The relation is given by: log(N) = a + b[log(D)],

[6.1]

where N is number of trees per unit area, D is the diameter (dbh) of the average tree, and a and b are constants for a given species or group of species, where the constant b defines the slope of the line. The non-linear analogue of Equation 6.1 is given by: N = aDb

[6.2] 227

228

Chapter 6

The relation has often been used to describe the average maximum limits of stand density and, by extension, to provide a relative measure or index of stand density (Reineke, 1933). Similar models of self-thinning based on the relation between numbers of trees and tree height are used in European forestry. Such models can be empirically derived by regression analysis and other statistical methods (Weller, 1987b) using data from temporary or permanent field plots from undisturbed stands encompassing a wide range of average stand diameters within a given forest type. Data from permanent plots with repeated measurements are preferred because periodic mortality is actually observed, which reduces assumptions about the self-thinning process (Zeide, 1987). Stands selected to define a line or limit of average maximum stand density should be at or near the upper limits of stand density with respect to their

average diameter. The resulting line showing number of trees per acre by mean stand dbh is sometimes interpreted as a self-thinning line, or line of 100% relative density (Fig. 6.1). The line provides a useful definition of the upper limits of stand density because the number of trees per unit area and mean dbh are highly correlated. Reineke (1933) postulated that the coefficient b, which determines the slope of the self-thinning line (Equations 6.1 and 6.2), assumes a value close to 1.605 for all tree species. For even-aged, upland oak forests in the eastern United States dominated by white, black, scarlet and chestnut oaks, the estimated slope coefficient was 1.5 based on temporary plot data from undisturbed stands ranging in site index from 50 to 80 ft (Fig. 6.1). Data from permanent plots in similar oak stands in the Central Hardwood Region produced a self-thinning line with a slope coefficient of 1.57 (Fig. 6.1).

3000

Trees per acre

2000

1000 600 400

200 From Schnur From Gingrich

100 60 2

3

4

6

8

12

16

20

Average dbh (inches) Fig. 6.1. Self-thinning lines (lines of average maximum stand density) for even-aged upland oak forests in the eastern United States based on the Reineke model. The dotted line is derived from Schnur’s (1937) stand table for mixed upland oak stands widely distributed across the eastern United States. The solid line is derived from Gingrich’s (1971) stand table for mixed upland oak stands in the Central Hardwood Region. The arrow from a hypothetical disturbed stand represented by the dot on the graph illustrates a typical trajectory of convergence with the Gingrich-based self-thinning line. Trajectories of stands below the self-thinning line generally move slightly downward from left to right. The downward trend results from competition-induced mortality, which occurs even in stands below average maximum density. After convergence with the self-thinning line and in the absence of further disturbance, the trajectory continues along the self-thinning line. The slope coefficient for the Schnur-based self-thinning line is 1.50 and for the Gingrich-based line is 1.57.

Self-thinning and Stand Density

Over time, stands lying below the selfthinning line will grow and move towards the line. On approaching the self-thinning line, stand development trajectories converge with the self-thinning line. Stand development then proceeds along the line from upper left (younger stands) to lower right (older stands) (Fig. 6.1). However, density-dependent mortality does not occur among trees with equal probability; it is concentrated among the suppressed trees. The overall rate of mortality thus is greatest during the stem exclusion stage of stand development, which is when a large proportion of trees succumb to suppression.

The 3/2 rule Another approach to defining the self-thinning line is based on the relation between average total plant biomass and number of plants per unit area in single-species populations undergoing density-dependent mor-

229

tality (Yoda et al., 1963). The power function model, similar to Equation 6.2, is used to describe the relation. However, in this case the model expresses the relation between average plant dry weight (biomass), w, and number of plants per unit area (N) such that: w = aNb

[6.3]

where a and b are usually estimated by regression from experimental data or field observations. Alternatively, the relation can be expressed as total plant weight (W) per unit area by: W = aNb

[6.4]

The 3/2 power relation is displayed with N on the horizontal axis, in contrast to Reineke’s model, where N is displayed on the vertical axis. Self-thinning for the 3/2 power relation thus graphically proceeds from lower right (younger stands) to upper left (older stands) along the self-thinning line (Fig. 6.2).

Mean dry weight of tree bole (kg)

600 200 100 50 Self-thinning line 20 10 5 3 1

0.3 100

200

500

1000

2000

4000

7000

Trees per acre Fig. 6.2. A self-thinning line (line of average maximum stand density) for normally stocked even-aged upland oak forests in the eastern United States based on the relation between average dry weight of tree bole (inside bark) and number of trees per acre. The relation is conceptually similar to Reineke’s model (Fig. 6.1), but differs in format. The arrow from a hypothetical disturbed stand represented by the dot on the graph illustrates a possible trajectory of convergence with the self-thinning line. As stand biomass increases and the number of trees per acre decreases over time, the stand trajectory moves upward and to the left along the self-thinning line. (Adapted from Schnur, 1937.)

230

Chapter 6

Numerous studies have shown that b, the slope coefficient in Equation 6.3, approximates 1.5 (and equivalently, b = 0.5 in Equation 6.4) for many plant species including herbs, shrubs and trees (Yoda et al., 1963; Harper, 1977; Miyanishi et al., 1979; White, 1985; Weller, 1987a). The relation consequently has become known variously as the ‘3/2 power law of self-thinning’, the ‘self-thinning rule’ and the ‘3/2 rule’. However, the relation is herein referred to as a rule rather than a law because of its demonstrated lack of generality (Sprugel, 1984; Weller, 1987a; Zeide, 1987; Norberg, 1988; Lonsdale, 1990) and the absence of a supporting theory (Hutchings, 1983). Discrepancies between observed slope values and 1.5 nevertheless have been interpreted as experimental error because of the coefficient’s presumed generality (Miyanishi et al., 1979; White, 1981). The rule also is purported to be independent of environmental factors (Yoda et al., 1963; White and Harper, 1970), and to be applicable to species mixtures as well as single-species stands (Westoby, 1984; White, 1985). The 3/2 rule can be interpreted geometrically. The rule assumes that plant weight, w, is proportional () to plant volume, v, which in turn is proportional to any linear plant dimension on which volume depends raised to the third power: v w = aN b

[6.5]

If we select crown diameter (Cd) as a linear dimension of interest, then Cd 3 aN b. To conceptualize the relation geometrically, it is convenient to consider Cd 3 proportional to a cylinder representing the ‘exclusive space’ of a tree (Norberg, 1988). Then Cd2, which is proportional to the cylinder’s cross-sectional area, can be used to represent crown area. Further, the cylinder’s height is assumed proportional to crown diameter. This three-dimensional space conceptually envelopes the tree, extending downward from the top of the crown to its corresponding ‘exclusive ground area’ and into its soil space. The volume of exclusive space also can be viewed as a hexagonal column, which conceptually allows for sym-

metrical packing of trees without producing crown overlap or unoccupied area as in circular crown areas (Fig. 6.3). The explanatory value of this simplified geometric view and its relation to the 3/2 rule is apparent from the geometric relation between the volume of a cylinder and its diameter (i.e. volume is equal to the cylinder’s squared diameter raised to the 3/2 power). Note that this relation only holds when the cylinder’s height is proportional to its diameter. To satisfy the geometric analogy for the 3/2 power rule, the thinning-rule model must provide a measure of Cd2. Such a measure is given by N, the number of trees per unit area (Equation 6.3). Because the reciprocal of N represents the area occupied by the average tree, N is related to crown area and thus crown diameter (Cd 2). For every unit increase in crown area, the exclusive space (volume) of a tree increases by 3/2. So, given a finite amount of growing space, the number of trees (N) in that space must decrease at a rate of 3/2 per unit increase in crown area. To conform to this geometric model, however, a tree must maintain the same height-todiameter ratio during self-thinning (Fig. 6.3). Accordingly, the various tree structures, including bole and crown, must remain proportionately similar during selfthinning (Yoda et al., 1963). Such constancy of proportions is known as isometry or geometric similarity (McMahon and Bonner, 1983; Norberg, 1988). Under the 3/2 power rule, coefficient b (Equation 6.3) is assumed to be 3/2 for all species. In contrast, coefficient a varies among species and determines the intercept (or elevation) of the thinning lines. This coefficient has been termed the ‘packing constant’ (Norberg, 1988) because it reflects the proportion of space occupied by plant biomass and the average plant biomass per unit of ground area. Coefficient a thus increases with increasing density, or packing, of plant parts within a tree’s exclusive space. Coefficient a also has been theorized to be related to the mass of mechanical tissue (bolewood in the case of trees) required to support a unit area of canopy (Givnish, 1986).


231

D

H

Fig. 6.3. The ‘exclusive space’ of closely packed trees. D is the diameter of the ‘exclusive ground area’ associated with each tree’s crown area and H is tree height. The thinning rule theory implies that the ratio of D : H remains constant throughout stand development. (Redrawn from Norberg, 1988, with permission from the University of Chicago.)

Relation between Reineke’s model and the 3/2 rule The diameter of a tree raised to some power between 2 and 3 equals its volume. For many species, the value of the exponent has been shown to be near 2.5 (Yoda et al., 1963). Reineke’s model and the 3/2 rule therefore are related by the approximate relation between average tree volume, v, and tree diameter, D, where v = D 2.5

[6.6]

From Equation 6.2 it then follows that D 2.5 N 2.5/b

[6.7]

When b = 1.605 in Equation 6.1 (Reineke’s postulated constant), the relation becomes: v = N1.56

[6.8]

where the exponent approximates 3/2 (Zeide, 1985). In postulating a constant of 1.605, Reineke was implying, intentionally or not, that the relation between a

tree’s growing space and its diameter is not constant (i.e. not isometric). To be consistent with the assumption of constancy of tree proportions (isometry) inherent in the 3/2 power rule, the Reineke model must assume a slope constant of 2.

Oak forests and the 3/2 rule Knowledge of the limiting relation between numbers of trees and volume per tree in oak stands is silviculturally useful. The relation can be used as a standard against which other stands can be measured. The 3/2 power rule attempts to describe this limiting relation in general terms for a wide range of plant communities. But to what extent do oak forests conform to the 3/2 power rule? Evaluation of the rule can be divided into two questions. First, does the mathematical form of the model (i.e. the power function) adequately express the relation? Second, is the slope coefficient of 3/2 universally

232

Chapter 6

applicable? If the answer to the first question is no, then the second becomes irrelevant. However, rejection of 3/2 as a universal slope coefficient does not, by itself, negate the utility of the model form. Evaluating conformance of oak forests to the 3/2 power rule is complicated by three factors. First, oak yield tables are the only comprehensive sources of information. But yield tables seldom report total tree biomass or volume; most yield tables include only bole volume, which must substitute for bole biomass. Moreover, even if bole mass were measured directly, it is unlikely to be a constant fraction of total tree mass (Sprugel, 1984). Although information on the allocation of biomass to below- and above-ground portions of oaks is sketchy, there is evidence that allocation varies greatly with site quality and between trees of coppice and non-coppice origin (Canadell and Rodà, 1991). Second, most oak stands are comprised of a mixture of species. This confounds the effects of competition within and among species, which have fundamentally different explanations in relation to the thinning rule (Zeide, 1985; Norberg, 1988; Weller, 1989). Variation in wood density among species can also introduce error into coefficient estimates. Third, yield tables available to evaluate self-thinning relations for oak have been smoothed by hand-fit curves or other unspecified methodologies. As such, they represent models, not data (Weller, 1987a). This obscures variation in the original data and may introduce other possible errors (Lonsdale, 1990). Despite these problems, it may be of interest to evaluate the 3/2 rule in relation to existing yield tables. Based on three oak yield tables for the midwestern and eastern United States, the line representing the relation between bole volume and number of trees produced slope coefficients that ranged from 1.7 to 2.2 (Fig. 6.4). Those estimates were based on substituting bole volume for w in Equation 6.3. The yield table values fit the power function with negligible error (Fig. 6.4), which indicates that their developers deemed this function to adequately describe the self-thinning line.

The tenuous conclusions from the above empirical evidence indicate that self-thinning lines in oak forests: (i) show varying proximity to the theoretical 3/2 slope; and (ii) are nominally, if not statistically, more negative than 3/2. Other investigators have concluded that self-thinning lines based on yield tables are likely to have slopes steeper than 3/2 because only portions of trees (boles) are represented, and because they include only trees above some minimum size (Harper, 1977; Lonsdale, 1990). The latter factor may explain, in part, why the yield tables for the Connecticut oak stands (Fig. 6.4C), limited to stems 2 inches dbh and larger, produce a steeper thinning line than either the upland oak or the Wisconsin yield tables, which included stems 0.6 inch dbh and larger (Fig. 6.4A and B). More recent data based on permanent plot data from unthinned upland oak stands representing three site classes in the oak–hickory region produced self-thinning lines that were quite different from those derived from earlier yield tables (Gingrich, 1971). The newer data suggest that each site class produces a separate concavedownward curve when both axes are transformed to logarithmic scale (Fig. 6.5). This pattern is consistent with those derived from Douglas-fir yield tables based on longterm observations from permanent sample plots (Curtis, 1982; Zeide, 1987). However, it is possible that at least part of the concave pattern could be caused by incomplete utilization of growing space in younger oak stands (e.g. stands with less than 500 ft3 per acre of volume and more than 1200 trees per acre in Fig. 6.5). Yield tables for English oak stands in England also show that the self-thinning line plotted on log–log scale does not form a straight line. The slope coefficient for fully stocked stands up to 100 years old is 1.48, which approximates to the hypothetical value of 3/2 (White and Harper, 1970). However, in stands between 100 and 150 years old, the coefficient shifts to 1. This shift indicates that bole volume per unit of ground area remains constant for English oak with continued tree growth


100

100 B Average tree volume (ft3)

Average tree volume (ft3)

A

10

b = –3/2 1

10

b = –3/2 1

b = –1.68

b = –1.83 0.1 100

233

1000 Trees per acre

10,000

0.1 100

1000 Trees per acre

10,000

100 Average tree volume (ft3)

C

10

b = –3/2

1

b = –2.16 0.1 100

1000 Trees per acre

10,000

Fig. 6.4. Self-thinning lines for oak stands derived from published yield tables compared to the 3/2 thinning rule line. (A) From yield tables for even-aged upland oak stands in the central and eastern states. (From Schnur, 1937.) (B) From yield tables for mixed hardwood stands dominated by northern red oak in southwestern Wisconsin. (From Gevorkiantz and Scholz, 1948.) (C) From yield tables for mixed oak stands in Connecticut. (From Frothingham, 1912.) Heavy lines are fit by linear regression to table values (shown by dots); b is the slope coefficient. Each regression fit is based on the model: log10(v) = a + log10[b(N )] where v is the volume of the average trees, N is stand density (trees acre−1 and log10 is the logarithm to the base 10. Regression estimates are averaged across the site classes given in each yield table.

and self-thinning (Harper, 1977; Norberg, 1988). Accordingly, the volume of bolewood lost to mortality would be compensated by the gap-filling of survivors. A slope of 1 also could indicate stagnating height growth in old stands, or root competition and soil physical constraints on root expansion (Norberg, 1988). The emerging evidence from permanent plot data collectively indicates that the thinning line as expressed by the 3/2 thinning rule does not form a straight line

over the life of a stand. If we accept that evidence, the power function equation generally, and thus the 3/2 slope coefficient specifically, cannot realistically describe self-thinning in oak stands. Moreover, there is evidence that the relation is not independent of site effects (Zeide, 1987). Conformity of tree growth to the slope coefficient of 3/2 requires that trees maintain geometric similarity and constant proportions, or isometry, among their various components as they grow. Alternatively,

234

Chapter 6

Stand volume (ft3 acre–1)

4000 2000 1000 500

200 100 60 100

Site index 55 ft Site index 65 ft Site index 75 ft –3/2 rule

200

300

500

1000 1500

3000

Trees per acre Fig. 6.5. Stand volume in relation to stand density for three site index classes in unthinned (normal) stands in upland hardwoods dominated by mixed oaks in the Central Hardwood Region. The stands represent mixtures of black, white, scarlet and chestnut oaks. The dotted line is the theoretical self-thinning line (0.5 from Equation 6.4) assumed by the thinning rule model; other lines represent table values based on models derived from permanent plots. (Adapted from Gingrich, 1971, Table 1.)

elastic similarity occurs when tree components proportionately change with increasing tree size (allometry) (McMahon and Bonner, 1983). Constant ratios among tree dimensions such as crown diameter and bole diameter in relation to changing overall tree size thus represent geometric similarity for those dimensions. For a species or species group to intrinsically conform to the 3/2 rule, ratios among tree components must remain constant (isometric) as tree size increases. A test of isometry is provided by the ratio of crown diameter to dbh in relation to changing tree size. These relations are defined by tree-area equations for oaks in stands at average maximum density. The equations indicate that the crown diameter : dbh ratio for forest-grown trees declines from about 30 : 1 for trees of small diameter to about 17 : 1 for trees of large diameter (Fig. 6.6d and e). For open-grown trees, ratios of crown diameter to dbh vary even more (Fig. 6.6a–c). In either case, small oaks have proportionately more crown area than large oaks. Such proportionate changes in tree dimensions are not consistent with geomet-

ric similarity and hence the 3/2 rule. In fact, bole diameters must increase at a proportionately faster rate than crown area (and correlatively crown mass) to prevent trees from collapsing under their own weight (McMahon, 1973). True isometry in oaks or any other tree species therefore is unlikely. Moreover, predictable changes in crown diameter : dbh ratios, by themselves, provide an alternative basis for defining a self-thinning line and corresponding measures of relative density, as discussed later in this chapter. Self-thinning consequently appears to be heavily influenced by tree geometry, which is continually changing to meet requirements for structural resistance to bole breakage as crown mass increases (McMahon, 1973). Moreover, trees grow with great physical plasticity to take advantage of their changing competition environment (Sorrensen-Cothern et al., 1993). For example, one side of a tree crown may expand into the gap created by the death or removal of one of its neighbours, resulting in an expanded but nonsymmetrical crown. Trees that survive


45

a b

40 Crown diameter : dbh ratio

235

35

c d e

30

25

20

15 0

4

8

12 16 Dbh (inches)

20

24

Fig. 6.6. Crown diameter: dbh ratios in relation to dbh. (a) Open-grown pin oak. (From Krajicek, 1967.) (b) Open-grown upland oaks and hickories. (From Krajicek et al., 1961.) (c) Open-grown bur oak. (From Ek, 1974.) (d) Northern red oak in stands at average maximum density. (From McGill et al., 1991.) (e) Upland oaks and hickories in stands at average maximum density. (From Gingrich, 1967.)

self-thinning acquire new resources (space, light, soil moisture and nutrients) as a consequence of spatial adjustments resulting from the death of neighbouring trees and differential growth rates among competing survivors. Associated with these newly acquired resources are changes in the allocation of growth to the various parts of trees, which thereby influence their proportions, shape and competitive relations with neighbours (Sorrensen-Cothern et al., 1993). The capacity of oaks to fill irregular canopy spaces, and thus their conformity to allometric growth, may be further reinforced by their upward spreading (decurrent) crowns and weak apical control of lateral branching. Despite apparent limitations of the 3/2 self-thinning rule to describe the underlying geometric relations for trees in evenaged oak stands, the general pattern of rapidly decreasing numbers of trees with increasing mean size is well established. Self-thinning formulae such as the 3/2 rule and Reineke’s model describe standlevel changes in the number of trees with

increasing tree size. These formulae are useful expressions of a relatively predictable process that has practical silvicultural value for defining limits of stand density in relation to average tree size. In turn, those limits can be used as a standard or index for expressing the relative density of any other stand. This leads to the subject of stand density and how it can be measured and expressed in oak forests.

Stand Density and Stocking Terminology Silviculturists are often interested in three related measures of stand density: absolute density, relative density and stocking. These measures can be used to describe a stand relative to some standard of comparison or to some condition that meets a silvicultural objective. Silvicultural decisions are often based on such measures of stand density, and the desired condition of the stand after treatment is usually described

236

Chapter 6

by these measures. Although the terms absolute density, relative density and stocking have not always been used consistently, some general conventions and definitions have been established (Walker, 1956; Bickford et al., 1957; Gingrich, 1964; Husch et al., 1982; Ernst and Knapp, 1985). Absolute density (or simply density in common usage) is a quantitative, objective measure of one or more physical characteristics of a forest stand expressed per unit area. Measures of absolute density are expressed quantitatively as a tree count, area, volume or mass. Ecologists usually use the term density to refer exclusively to the number of individuals per unit area. In forestry, however, the term can refer to any of several measures of site occupancy, including number of trees, basal area or volume per unit area. Measures of density are usually restricted to trees larger than some minimum size, usually expressed as a minimum dbh. Specifying this minimum size is important because absolute density usually differs with differences in the minimum measured tree size. Measures of relative density provide additional information by comparing an absolute density to a reference value. An example of a measure of relative density is the ratio of the number of trees of a given species per acre to the total number of trees per acre. Expressed as a percentage, this value has long been used by ecologists to define the relative density of a species within a specified area. Forest regeneration and growth can be greatly affected by relative stand density. Consequently, silviculturists have developed various methods of expressing relative density. Virtually all silvicultural definitions of relative density involve ratios. For example, Reineke’s (1933) stand density index provides a reference line describing the maximum number of trees per acre for stands of a given mean dbh (Fig. 6.1). The maximum number of trees decreases rapidly as the mean stand diameter increases. For any given stand, the observed number of trees and mean dbh can be used to compute the ratio (or percentage) of the number of the maximum

(reference) number of trees indicated by the stand density index relation (Reineke, 1933; Schnur, 1937). Other common measures of relative oak stand density are based on tree–area ratios or stocking per cent (Chisman and Schumacher, 1940; Gingrich 1967; Ernst and Knapp, 1985). In their application to oak forest types in North America, most measures of relative density are designed to compare one or more absolute measures of stand density to a standard. The standard is often based on an observed maximum absolute density for undisturbed natural stands at a comparable stage of development, but it may be based on other limits or reference conditions. For example, crown competition factor (CCF) estimates stand density relative to a minimum tree crown area per acre below which trees do not fully utilize available growing space (Krajicek et al., 1961). The method of French ‘normes’ compares the observed number of trees and the mean height of dominant trees to both a maximum density and the minimum number of trees necessary to maintain dbh growth below 2 mm (0.08 inch) per year, which by European standards is considered most desirable for veneer production (Oswald, 1982). Other measures of relative density that generally have not been applied to oak forests, but could be, include: Curtis’s (1982) relative density index (references observed basal area per acre to that of an undisturbed stand with the same quadratic mean diameter); Wilson’s (1946) relative spacing index (references the observed number of trees per acre to the number of trees in an undisturbed stand having the same dominant height); and Drew and Flewelling’s (1977) relative density index (references number of trees per unit area to the volume of the average tree). The latter method is analogous to the graphical format for expressing the 3/2 power rule discussed earlier. Comprehensive reviews of measures of relative density include those by Curtis (1970) and Stout and Larson (1988). Stocking is a subjective term used to describe the adequacy of any observed level of stand density with respect to a silvicultural goal (Bickford et al., 1957;


Gingrich, 1964). The terms overstocked, understocked and fully stocked are used to describe stocking adequacy relative to a specified silvicultural goal. Accordingly, a stand may be overstocked (too dense) for one silvicultural objective and fully (i.e. appropriately) stocked for another, or may be overstocked at one age and understocked at another. In contrast to the term stocking, the term stocking per cent is a measure of relative density specifically associated with the Gingrich-style stocking diagram. This diagram combines measures of absolute and relative density into a single graphical format (Gingrich, 1967). Stocking per cent is a widely used measure of stand density in North American oak silviculture. It is based on the relation between tree size and associated growing space requirements discussed later in this chapter. The word stocking is often used incorrectly to refer to stocking per cent (a measure of relative density). This sometimes creates confusion because, as discussed later, full stocking is synonymous with complete utilization of growing space, which covers a wide range of stocking percentages on the Gingrich stocking diagram. Normal stocking is a term used to describe undisturbed even-aged stands that are at or near maximum density for their age. Normally stocked stands are characterized by a lack of gaps in the forest canopy and a relatively uniform spacing between stems. Basal area and cubic foot volume are at or near their maximum for a given stand age and site quality. Normally stocked stands (sometimes simply called normal stands) usually are identified subjectively based on these criteria. Observations of the number, basal area and volume of trees per acre in normally stocked stands across a wide range of stand age and site quality classes have been used to develop normal yield tables. These tables specify the expected maximum basal area and maximum cubic foot volume for unmanaged stands of a given age and site class. In addition to their application to yield estimation, the tabulated values can be used as reference conditions to estimate the relative density of other stands.

237

Maximum and minimum growing space There are limits to the amount of growing space a tree of a given bole diameter can occupy. Although this may seem self-evident, the concept is central to quantifying stand density and stocking per cent in oak stands. The actual amount of space that a tree occupies is difficult to measure because it includes crowns and roots that overlap in three dimensions with other trees. Fortunately, for many silvicultural purposes, a tree’s growing space can be adequately estimated as a circular area, or tree area, representing the crown. In this context, tree area is interpreted geometrically as a tree’s area of influence or potential influence concentric to the tree bole; it is also highly correlated with dbh. Estimates of the maximum area that a tree of a given dbh can occupy are usually developed from crown and dbh measurements of open-grown trees. In contrast, estimates of the minimum area that a tree requires are usually developed from measurements of tree diameters in normally stocked stands. Trees that are open-grown throughout their lives develop the largest crowns possible for their dbh and species. Consequently, open-grown trees have often been used to estimate the maximum area a tree of given species and dbh can occupy. There is a high correlation between bole diameter and crown area of open-grown trees. This relation has led to the development of equations for estimating the crown areas of opengrown trees from dbh for various oaks and associated species in several regions in the eastern United States. The results have shown that the relation between maximum crown width and bole diameter is often linear or nearly linear (Krajicek et al., 1961; Krajicek, 1967; Ek, 1974). An example is the maximum crown width equation applicable to oaks and hickories in the Central Hardwood Region, which is given by: CWmax = 3.12 + 1.829D

[6.9]

where CWmax is the estimated crown width (ft) of an open-grown upland oak or hickory, and D is tree dbh (inches) (Krajicek et al.,

238

Chapter 6

1961). Assuming tree crowns are circular, squaring both sides of Equation 6.9 and multiplying by π/4 defines maximum crown area (CAmax) in relation to dbh so that: CAmax = 7.645 + 8.965D + 2.627D2 [6.10] CAmax therefore is the approximate circular crown area (ft2 in vertical projection) of an open-grown upland oak or hickory. Maximum crown width equations also have been derived for other species and regions (Table 6.1). An exponent in the diameter term of some equations indicates non-linearity in the relation. As in the derivation of Equation 6.10, equations in Table 6.1 can be similarly expressed as crown area. Graphical presentation of equations facilitates comparisons among species. For example, open-grown black walnut trees have larger crowns than oaks and hickories for a given diameter, whereas shortleaf pines have smaller crowns. The maximum crown width of sugar maple

may be larger or smaller than that of oaks and hickories, depending on dbh (Fig. 6.7). Assuming that maximum crown width equations adequately express the maximum amount of above-ground growing space that a tree of a given diameter can occupy, we can estimate the fewest trees of a given dbh required to completely occupy an acre, i.e. 43,560/CAmax. Alternatively, the maximum tree area for all the trees on any acre can be calculated by summing their individual maximum crown areas (Equation 6.10 and Table 6.1). When the sum of the maximum crown areas equals the area of an acre (43,560 ft2), the stand is said to have a maximum tree–area ratio (TARmax) of 100%. This represents the condition where tree–area satisfies the minimum requirements for full utilization of growing space given that tree crowns are, for their dbh, maximally extended. Maximum tree–area ratio (TARmax) therefore is a relative measure of stand density that defines the percentage of an area

Table 6.1. Equations for estimating open-grown crown widths from dbh for oaks and some commonly associated species.

Species (location) American elm (Wisconsin) American basswood (Wisconsin) Black cherry (Wisconsin) Black oak (Wisconsin) Black walnut (unspecified) Black walnut (Wisconsin) Bur oak (Wisconsin) Green ash (Wisconsin) Jack pine (Quebec) Loblolly pine (unspecified) Northern red oak (Wisconsin) Oaks and hickories (Iowa) Pin oak (unspecified) Red maple (Wisconsin) Shagbark hickory (Wisconsin) Shortleaf pine (Missouri) Sugar maple (Wisconsin) Sugar maple (Eastern US) Sweetgum (unspecified) White oak (Wisconsin) a Crown

Maximum crown widtha 2.829 + 3.456D 0.8575 0.135 + 3.703D 0.7307 0.621 + 7.059D 0.5441 4.504 + 2.417D 4.873 + 1.993D 4.901 + 2.480D 0.942 + 3.539D 0.7952 4.755D 0.7381 2.036 + 1.736 4.78 + 1.56D 2.850 + 3.782D 0.7968 3.12 + 1.829D 9.06 + 1.525D 4.776D 0.7656 2.369 + 3.548D 0.7986 2.852 + 1.529D 0.868 + 4.150D 0.7514 12.08 + 1.32D 2.65 + 1.975D 3.689 + 1.838D

Source Ek, 1974 Ek, 1974 Ek, 1974 Ek, 1974 Krajicek, 1967 Ek, 1974 Ek, 1974 Ek, 1974 Vezina, 1963 Roberts and Ross, 1965 Ek, 1974 Krajicek et al., 1961 Krajicek, 1967 Ek, 1974 Ek, 1974 Rogers, 1983 Ek, 1974 Smith and Gibbs, 1970 Krajicek, 1967 Ek, 1974

width in feet given tree dbh (D) in inches; corresponds to CWmax in text. Assuming tree crowns are circular in cross-section, maximum crown area in square feet is equal to (CWmax)2•π/4.


where summations (∑) are over all trees per acre, Di is the dbh of tree i and N is number of trees per acre. Note that ∑Di2 is equal to the stand basal area in square feet per unit area divided by π/576. A CCF of 100 (or equivalently, ∑TARmax = 100%) therefore is usually interpreted as the approximate lowest density at which a stand fully utilizes above-ground growing space. Stands with CCFs below 100 are certain to have canopy gaps. CCF values near 200 have been observed for undisturbed oak–hickory stands (Krajicek et al., 1961). Just as trees have a maximum area they can occupy, they also have a minimum tree area that is necessary for good physiological function and survival. However, minimum tree area is derived quite differently from its maximum tree-area counterpart. Unlike maximum tree area, which can be estimated from open-grown trees, minimum tree area is difficult to observe directly for individual trees. Minimum tree area requirements nevertheless can be estimated from data obtained from undisturbed, normally stocked, even-aged stands. Estimation is based on deriving minimum tree–area ratio (TARmin) equations

(e.g. an acre) that would be utilized by trees when all tree crowns are fully extended. When TARmax is less than 100%, the trees present would not utilize the available growing space even when their crowns are fully extended. Consequently, reducing TARmax below 100% by thinning will, at least temporarily, result in unutilized growing space. Calculating TARmax can be simplified by dividing equations for open-grown crown areas (e.g. Equation 6.10, or Table 6.1) by 435.6, the area comprising 1% of an acre. For Equation 6.10, TARmax is given by: TARmax = 0.0175 + 0.0205D + 0.00603D2 [6.11] where TARmax is the maximum percentage of an acre that a tree of a given dbh (D) can occupy. The sum of TARmax for all the trees on an acre is sometimes referred to as crown competition factor (CCF) (Krajicek et al., 1961) and is calculated by summing TARmax for individual trees as follows: CCF = ∑(0.0175 + 0.0205Di + 0.00603Di2) [6.12] = 0.0175N + 0.0205∑Di + 0.00603∑Di2 [6.13] 60

Oak and hickory Shortleaf pine Sugar maple Black walnut

Maximum crown width (ft)

50

40

30

20

10

0 0

4

8

239

12

16

20

24

28

Dbh (inches) Fig. 6.7. Estimated open-grown crown widths of oaks and hickories and three commonly associated species in relation to bole diameter (dbh). See also Table 6.1. (Adapted from Ek, 1974 (sugar maple), Krajicek, 1967 (black walnut), Krajicek et al., 1961 (oak–hickory), Rogers, 1983b (shortleaf pine).)

240

Chapter 6

that express tree growing space requirements for normally stocked stands (Chisman and Schumacher, 1940). Like TARmax, TARmin expresses tree area in per cent of an acre. Just as maximum tree area is a linear function of diameter and diameter squared (Equations 6.9 and 6.10), a tree’s minimum tree area can be similarly expressed by: TARmin = c0 + c1D + c2D2

[6.14]

where TARmin is the estimated minimum per cent of an acre required by a tree of a given dbh (D) in a normally stocked forest. Unlike the maximum tree–area coefficients, the coefficients for the minimum tree–area equation are not derived from measurements of crown diameters. Instead, they are estimated by regression by assuming the sum of the tree areas for all trees on an acre of undisturbed, normally stocked forest is equal to 43,650 ft2, or 100% of an acre (Chisman and Schumacher, 1940; Gingrich, 1967). Minimum tree area then can be expressed directly as a percentage of an acre. This expression has been termed stocking per cent (S%) (Gingrich, 1967) and is given by: S% = ∑ (b0 + b1Di + b2D22)

[6.15]

= b0N + b1∑Di + b2∑Di2

[6.16]

where summations (∑) are over all trees per acre, S% is the percentage of an acre filled by the minimum tree areas of all trees on that acre, Di is the dbh of tree i, N is the number of trees per acre, and b0, b1 and b2 are coefficients (usually estimated by regression). The stocking percentage represented by a single tree can be derived by solving Equation 6.16 for N = 1. When stocking percentage equations are expressed on a per acre basis (e.g. Equation 6.15 or 6.16), equations for minimum tree area in square feet can be derived by multiplying each term in the equation by 435.6 (the number of square feet in 1% of an acre). Although stocking percentage is usually the relative density measure of choice, rescaling to square feet facilitates comparing the estimated tree areas representing

maximum and minimum growing space (Fig. 6.8). Other factors being equal, the closer a tree’s crown is to its maximum size for its dbh, the faster the tree’s diameter and gross volume growth. The minimum tree–area ratio is reportedly independent of stand age and site quality (Chisman and Schumacher, 1940; Gingrich, 1967), and can be applied to mixed as well as to pure stands. The methodology for deriving minimum tree–area equations is described in more detail by others (Chisman and Schumacher, 1940; Gingrich, 1967; Roach, 1977; Ernst and Knapp, 1985; Stout and Nyland, 1986). Oak and associated forest types are seldom comprised of a single species. In applying stocking equations, it is therefore important to recognize differences in tree–area ratios among species. In developing stocking equations, coefficients for individual species can be derived by incorporating species-specific terms into stocking equations so that the Equation 6.15 expands to the more general form: no. spp.

no. trees

no. trees

S% = ∑ (b0j Nj b1j ∑ Dij b2j ∑ Dij2) j=1 i=1 i=1 [6.17] where the outer summation (∑) is over all species; the inner summations are over all i trees of species j; b0j, b1j and b2j are coefficients specific to species j; Nj is the number of trees of species j; and Dij is the diameter of tree i of species j (Roach, 1977). In forests such as the oak–hickory type of the Central Hardwood Region, the minimum tree–area ratios of the predominant species do not differ significantly (Krajicek et al., 1961; Gingrich, 1967). A single set of coefficients therefore can be used to represent the major species of the forest type. In other forest types, tree–area ratios differ appreciably among species (Roach, 1977; Stout and Nyland, 1986; Zhang et al., 1995). When such differences occur, separate coefficients for individual species or species groups can improve the accuracy of relative density equations. This is the case in the Allegheny hardwood forests of Pennsylvania, which are comprised of mixed stands of black cherry, yellow-


241

3000

Crown area (ft2)

2500

2000 Maximum crown area 1500

1000 Minimum crown area

500

0 2

6

10


26

30

Fig. 6.8. Estimated maximum and minimum tree areas in relation to bole diameter (dbh) for upland oaks and hickories in the Central Hardwood Region. The area between the two lines represents the approximate biological range of crown areas for individual trees. (From Gingrich, 1967.)

poplar, red maple, white ash, sugar maple, black birch, yellow birch, American beech, oaks and other species. Analysis of species-specific tree–area ratios identified three species groups with significantly different tree–area ratios (Stout and Nyland, 1986; Zhang et al., 1995). Stocking equations based on tree–area ratios also have been derived for northern red oak and various forest types of the eastern United States that often include oaks (Table 6.2). The silvicultural value of tree–area ratios and stocking percentage for defining relative density is reinforced by their demonstrated independence of stand age and site quality (Chisman and Schumacher, 1940; Gingrich, 1967). They also have been shown to be little influenced by variation in stand structure (Gingrich, 1967). Minimum tree–area ratio equations and equivalent stocking percentage equations can be used to calculate an approximate average upper limit of stand density corresponding to that for normally stocked stands. This upper limit, which can be graphically expressed as a line on a stand density diagram, has been termed average maximum density or average max-

imum competition (Ernst and Knapp, 1985). This line is generally interpreted as the level of density that stands tend to return to in the absence of disturbance (Gingrich, 1967; Ernst and Knapp, 1985). Nevertheless, stand density is likely to fluctuate around this line due to variation in weather, minor outbreaks of insects and disease, and other factors associated with ‘regular’ tree mortality. A potential deficiency of all measures of maximum relative density is the necessarily subjective selection of stands or plots used to derive tree–area ratios.

Stand density diagrams Stand density diagrams are graphical representations of the equations and variables that define relative stand density. In practical application, measures of relative density are often more convenient when displayed as diagrams than as equations. The most widely used type of stand density diagram in North American oak and hardwood silviculture is the one developed by Gingrich (1967).

242

Chapter 6

Table 6.2. Equations for estimating stocking per cent (minimum tree–area ratio as a per cent of an acre) by species group. Species (location)

Stocking per cent per acre a

Source

Upland oaks and hickories (Ohio, Kentucky, Missouri, Iowa) Northern red oak (Wisconsin) Sugar maple American beech (Allegheny Plateau) Black cherry Yellow-poplar (Allegheny Plateau) Red maple American basswood White ash (Allegheny Plateau) Black walnut (Unspecified) Shortleaf pine (Missouri)

0.00507N + 0.01698ΣD + 0.00317ΣD2

Gingrich, 1967

0.02476N + 0.004182ΣD + 0.00267ΣD2

McGill et al., 1991, 1999

0.003082N + 0.006272ΣD + 0.00469ΣD2

Stout and Nyland, 1986

0.02794N + 0.01545ΣD + 0.000871ΣD2


0.01798N + 0.02143ΣD + 0.001711ΣD2


0.01646N + 0.01347ΣD + 0.002757ΣD2

Schlesingerb

0.008798N + 0.009435ΣD + 0.00253ΣD2

Rogers, 1983

is number of trees per acre, ΣD is sum of diameters (inches/acre), and ΣD2 is sum of squared diameters (square inches/acre). Note that ΣD2 is equal to basal area (ft2/acre) divided by 0.005454. Minimum tree area in ft2 (TARmin) can be obtained by multiplying the calculated stocking per cent by 435.6, the number of square feet in 1% of an acre. bUnpublished equation by Richard C. Schlesinger, USDA Forest Service, Columbia, Missouri. aN

The Gingrich diagram Gingrich’s (1967) stand density diagram is based on tree–area ratio equations. The equations, in turn, are based on data from white, black, scarlet and chestnut oak stands in the Central Hardwood Region. Gingrich’s diagram incorporates measures of absolute density, relative density and stocking percentage into one graph. The measures of absolute density used are number of trees per acre (horizontal axis) and basal area per acre (vertical axis). Quadratic mean stand diameter (i.e. the diameter of the tree of average basal area) is also shown in relation to basal area and trees per acre. For any given stand, observed values of basal area and trees per acre can be used to determine stocking per cent directly from the diagram (Fig. 6.9). Gingrich’s diagram graphically defines the line representing average maximum stand density in relation to basal area, trees

per acre and mean stand diameter. The line is based on the minimum tree–area equation for upland oaks and hickories (Table 6.2). On the diagram, the line is labelled ‘100% stocking’ or ‘A-level stocking’ (Fig. 6.9). Similarly, the maximum tree–area ratio equation (Equation 6.11) defines the fewest number of trees of a given diameter sufficient to completely occupy all the growing space on an acre. On the diagram, this reference line is labelled ‘B-level stocking’. It could also be called the ‘line of imminent competitioninduced mortality’ (Drew and Flewelling, 1977). Over a wide range of stand conditions, B-level occurs between 57% and 59% of A-level stocking. Another line, labelled C-level, defines the relative density at which a stand on an average site requires 10 years to attain B-level stocking. The steep slope of the C-level line reflects the relatively rapid rate of increase in


243

7

110

6

Av e

A

100

et

ree

Ov

90

dia

me

ter

4

ers

toc

80

(in

.)

ke

d

Fu lly s

toc

ke

d

7 in 0 g pe 80 rc en t

B

60

90

70

10 0

11 0

3

50 Un d

C

ers

toc

ke

d

40

30

6 St 0 oc k

40

50

Basal area per acre (ft2)

rag

5

20 1

3

5

7

9

11

13

15

Hundred trees per acre 15

140 130

A

13 Avera ge 12 tre 11 ed iam 10 ete 9 r (i Ov n.) ers 8 toc ked

11 0 Fu

90 80

80

lly sto cke d

50

kin rst

ock ed

St oc

de

60

Un

50

C

gp

70

er

B 70

nt

90

10

100

60

7

0

110

ce


120

40 50

100 150 200 250 300 350 400 Trees per acre

Fig. 6.9. Relation of basal area, number of trees and average tree diameter to stocking per cent in upland oak stands in the Central Hardwood Region. Values are for trees ≥ 2 inches dbh. Lines of average tree diameter correspond to the dbh of the tree of average basal area or quadratic mean diameter. Upper panel is for stands with quadratic mean tree diameters from 3 to 7 inches dbh; lower panel is for stands with quadratic mean diameters from 7 to 15 inches dbh. The area between curves A and B represents the range of stocking where trees fully utilize growing space. (From Gingrich, 1967.)

244

Chapter 6

stocking (and basal area) in stands with small mean diameters. Qualitative levels of stocking (overstocked, fully stocked and understocked) are also identified on the stocking diagram. Stands are overstocked when they fall above A-level, fully stocked when they are between A- and B-levels, and understocked when they are below B-level. Stands within the fully stocked range therefore are considered to completely utilize growing space. Although stocking per cent can be calculated directly from the associated tree–area ratio equation (Table 6.2), the diagram facilitates rapid estimation when basal area and number of trees per acre are known. Stocking per cent is also related to canopy cover immediately after thinning and subsequent crown closure (Fig. 6.10). Unless residual stand densities are reduced to levels below 50% stocking, the canopy gaps created by thinning quickly close through crown expansion and correlated dbh growth. The stocking diagram illustrates the inadequacy of basal area alone as an expression of relative stand density. If basal area and stocking per cent were equivalent measures of density, the stocking per cent lines would parallel the horizontal basal area lines – thereby signifying isometry between tree crown area and basal area. Instead, the slope of the stocking per cent lines shows that trees can endure more crowding as they increase in diameter. This relation also implies that trees of large diameter require less space to support a unit of basal area than trees of small diameter. The density diagram is designed for rapid estimation of stocking per cent from two easily obtained measures of stand density: basal area and trees per acre. Note, however, that the stocking per cent equation (Equation 6.16 and Table 6.2) includes three terms: number of trees per acre (N), sum of diameters per acre (∑D) and sum of diameters squared per acre (∑D2) the latter value also is equivalent to 0.005454•basal area per acre. The density diagram incorporates only number of trees and basal area with no direct measure of sum of diameters. Nevertheless, it is important to consider both ∑D and ∑D2 when estimating

stocking per cent because together they account for variation in stand structure. This can be seen from the relation between the variance in tree diameters (2) and the difference or ‘discrepancy’ between arithmetic mean stand diameter (AMD) and quadratic mean stand diameter (QMD): 2 = QMD2 AMD2

[6.18]

where AMD is the arithmetic mean stand diameter (i.e. ∑D/N) and QMD is the tree of average basal area (i.e. [(∑D2/N)0.5]). The greater the variance in tree diameters, the greater the difference between arithmetic mean stand diameter and quadratic mean stand diameter. Moreover, if either AMD or QMD is given, the other can be algebraically derived from knowledge of 2. A valuable property of the stocking per cent equation (Equation 6.16 and Table 6.2) therefore is the indirect incorporation of information about the variance of tree diameters via inclusion of ∑D and ∑D2. The loss of accuracy in calculating stocking per cent that results from not considering ∑D would not be significant if all the trees were of equal or nearly equal diameter. But this is seldom if ever the case. When Gingrich (1967) derived the stocking per cent lines shown on his diagram, he incorporated estimates of the variance of tree diameters. He based these estimates on the observed negative linear relation between 2 and QMD. The stocking per cent lines shown on the density diagram therefore reflect average or ‘representative’ stand diameter distributions that change as the arithmetic mean stand diameter changes. Stocking per cent derived directly from the stocking per cent equation may be as much as 4% greater than indicated by the density diagram when the range of tree diameters is small. Where there is a wide range of diameters (e.g. in uneven-aged stands), stocking per cent calculated from the equation may be as much as 6% lower than that indicated by the stand density diagram. Differences of this magnitude are usually of little practical importance. However, not all density diagrams based on tree–area ratios incorporate estimates of the variance of diameters. Developers of dia-


245

A

B

C

Fig. 6.10. Representative canopy cover of oak stands in the Ozark Highlands for three levels of stocking: (A) 100%; (B) 80%; and (C) 60%. Photos for 60% and 80% stocking were taken immediately after thinning. (USDA Forest Service, North Central Research Station photograph.)

grams may wittingly or unwittingly assume that variance in diameters, and thus stand structure, is inconsequential in estimating tree–area ratios. When information on individual tree diameters is available, the tree–area equation or stocking per cent equation will produce estimates that are

more accurate than estimates read from the stocking diagram. Gingrich’s stand density diagram was developed from observations in forests comprised largely of white, black, scarlet and chestnut oaks. Other oaks may have different growing space requirements and

246

Chapter 6

thus different A-level and B-level stocking lines. For example, northern red oak attains 100% stocking at substantially higher levels of absolute density than those represented by Gingrich’s equation. Alternative stocking equations and stand density diagrams therefore have been developed for northern red oak (Sampson, 1983; Sampson et al., 1983; McGill et al., 1991, 1999; Stout, 1991) (Fig. 6.11). Based on the equation of McGill and others (Table 6.2), a stocking per cent of 100 for northern red oak equates to stocking percentages ranging from 120 to 140 on the Gingrich diagram (Fig. 6.9). Assuming the associated tree–area equations are accurate, white, black, scarlet and chestnut oaks and hickories in the Central Hardwood Region require 20–40% more growing space per tree than that required for northern red oak in Wisconsin. Two other stand density diagrams based on tree–area ratios have relevance to oak forests in the eastern United States. One is the stand density diagram for Allegheny hardwoods developed by Roach (1977). Those forests often include northern red oak as an associated species. The related stand density diagram is unique in that it accommodates adjustments in A-level stocking associated with changes in species composition. Because of the relatively narrow crowns of black cherry, white ash and yellow-poplar, the reference line for Alevel stocking increases as the percentage of those species increases. A stand density diagram based on tree–area ratios also has been developed for southern bottomland hardwood forests that include oaks as component species (Goelz, 1995). The diagram is based on the stand tables of Putnam and others (1960) for mixed bottomland stands in the lower Mississippi Valley, lower Piedmont and southern Coastal Plain. In those forests, cherrybark, laurel, Nuttall, overcup, pin, Shumard, water, willow and swamp chestnut oaks are commonly associated with other bottomland species. Despite their widespread use, some of the properties and assumptions incorporated into stand density diagrams based on tree–area ratios are not always well under-

stood or appreciated. Some characteristics worth remembering are presented below. ● Crown closure is assumed to occur when the sum of the corresponding crown areas of open-grown trees equals 43,560 ft2, i.e. ∑TARmax = 43,560 ft and CCF = 100. This is assumed to be the point of crown closure where inter-tree competition begins. However, it is difficult to perceive of a situation where this assumption would be literally true. Irregularities in tree spacing virtually ensure that tree crowns will form overlaps and gaps. The assumption nevertheless provides a quantifiable and demonstrably useful reference line that estimates the minimum number of trees necessary to occupy the available growing space. ● On a good site, trees of a given diameter may be younger than trees of the same diameter on a poor site. Trees of the same size nevertheless require the same amount of growing space, regardless of their ages (Krajicek et al., 1961; Gingrich, 1967). ● Stand density can exceed 100% stocking. This gives rise to the term ‘overstocked’. However, in the absence of disturbance, stands are expected to move toward the line of A-level stocking regardless of where they initially fall on the density diagram. ● According to current silvicultural theory, oak stands maintained in the ‘fully stocked’ zone will produce approximately the same total gross volume or biomass growth regardless of where they are maintained within that zone. However, other factors being equal, individual trees in oak stands maintained near B-level stocking will grow faster and usually produce greater total net merchantable volume than trees in stands maintained near A-level. Exceptions include oak stands grown exclusively for wood fibre or stands containing certain other hardwoods. The net merchantable growth per acre of stands containing a large proportion of black cherry (Nowak, 1996) or yellow-


7

160

6

Av e

A 140

5 O


247

120 Fu

100

lly

st

oc

ke

ra

ve

ge

rs

to

tre

ck

e

di

am

4

et

ed

er

(in

.) 3

d

80

00

0 11

1 90

60

B Un

40

80 ent rc pe g in

70 de

rs

to

ck

60 ock St

50 ed

40 30

20 100

400

700

1000

1300

1600

Trees per acre

15

180

13

A

toc

ked

90

ly s

10

Ful

140

0

11 0

160

Ave rag e tr 12 ee 11 dia me 10 ter (in. 9 Ove ) rsto 8 cke d 7

120

7 pe 0 rc en 80 t


200

50

B rsto

cke

30

d

40

Un de

60

St

80

6 oc 0 kin g

100

40 0

100

200 300 400 Trees per acre

500

600

Fig. 6.11. Relation of basal area, number of trees and average tree diameter to stocking per cent in northern red oak stands in Wisconsin. Values are for trees ≥ 2 inches dbh. Lines of average tree diameter correspond to the dbh of the tree of average basal area or quadratic mean diameter. Upper panel is for stands with quadratic mean tree diameters from 3 to 7 inches dbh; lower panel is for stands with quadratic mean diameters from 7 to 15 inches dbh. The area between curves A and B represents the range of stocking where trees fully utilize growing space. (From McGill et al., 1991, used with permission.)

248

Chapter 6

poplar (Trimble, 1968; Beck, 1986) may decrease with decreasing residual stand density. For those species, post-thinning rates of crown expansion and crown densification may be insufficient to compensate for loss of growth related to the removal of trees (Nowak, 1996). The general rule of equal growth across the ‘fully stocked’ zone therefore may not hold for some species mixes that commonly include oaks. ● There is evidence, based on long-term forest growth studies, that gross volume growth increases with increasing stand density up to a stand’s maximum biological limits of density (Zeide, 2001). Accordingly, there would be no such thing as an excessive number of trees – as implied by the labelling of the ‘overstocked’ zone of the Gingrich and similar stocking charts. In a general biological context, the concept of ‘overstocking’ therefore may be misleading. Nevertheless, the use of thinning to increase the rate of residual stand growth to more quickly obtain merchantable timber products other than wood fibre generally holds. • Tree–area ratio equations assume that interactions among species do not affect tree area. However, it would seem plausible that co-occurring species that appreciably differ in how they utilize site resources would more efficiently ‘share’ a unit space than co-occurring species with very similar resource requirements (McGill et al., 1999). If this were true, we might expect TARmin for two species with very different resource requirements to be smaller (and therefore their joint absolute maximum densities to be larger) when the two species are interacting than when they are growing apart. For example, northern red oak stands with a subcanopy of sugar maple have higher average maximum absolute densities (in basal area and trees per acre) than pure red oak stands (McGill et al., 1999). This suggests that minimum tree–area ratios for a given species may change as species composition changes.

● Stands at identical stocking per cent may develop differently, depending on their structure and history. For example, the growth of a previously undisturbed stand reduced to 70% stocking by windthrow of mostly large trees is likely to be different from a similar stand that was thinned to the same stocking by removing mostly small trees. The general direction of the response should be similar for both stands: increased stocking and basal area, and decreased number of trees. But rates of stand growth and increases in stocking would be expected to differ substantially because of differences in stand structure and composition. Stand density, therefore, does not define a unique stand condition even within the same forest type and site class. Use of stocking per cent or other measures of stand density to make accurate growth predictions thus requires specifying additional stand characteristics.

Other diagrams Stand density diagrams other than those based on tree–area ratio have been developed, and some of these have been applied to oak forests. One example is represented by Oswald’s (1982) diagram for English (pedunculate) oak and sessile oak in France. This diagram is a variant of Reineke’s (1933) stand density index. However, mean tree diameter as shown on the horizontal axis in the Reineke model is replaced by the mean height of the 100 tallest trees per hectare (Fig. 6.12). Another example is the stand density diagram for upland hardwoods developed by Kershaw and Fischer (1991), which is based on a format developed for Douglas fir (Drew and Flewelling, 1977). It incorporates the same principles as the Reineke and Gingrich diagrams, but shows trees per acre in relation to mean board foot volume per tree. It differs from other oak stand density diagrams by directly incorporating information on merchantable products.


249

Trees per hectare

2000 1600 1400 1200 1000 800

M

ax im

um

bi

ol

600 500 400

og

ic

al

de

ns

ity

300 250 200 170 140 120 100 85 70

Density index 140 120 100 85 70 16 18 20 22 24 26 28 30 32 34 Mean height of dominant trees (m)

Fig. 6.12. A stand density diagram for sessile and English oaks (non-coppice stands) in France. Stand density is expressed as the relation between trees per hectare (upper storey trees ≥7 cm in diameter at 1.4 m) and the mean height of the 100 largest-diameter trees per hectare. The uppermost line represents maximum biological stand density. Densities below the lowermost line result in annual ring widths exceeding 2 mm, which European foresters regard as undesirable for oak veneer logs. In this diagram, stand densities are indexed to the number of trees per hectare in stands with dominant trees averaging 35 m tall. Density indices between 70 and 140 identify the operational range for residual stand densities; the usual after-thinning range for oaks is between 70 and 100. French foresters refer to these indices as normes. (Redrawn from Oswald, 1982, used with permission.)

Density diagrams and stand growth In addition to their conventional use in managing stand density, density diagrams can be used to show how different stand attributes simultaneously change through time. This application can be illustrated by plotting actual or projected changes in stand basal area and trees per acre on the density diagram. Some growth and yield simulation models provide this as a graphic display option. Stand growth can be displayed in a variety of formats and units, depending on silvicultural objectives (Fig. 6.13). In the absence of disturbance and at densities below 100% stocking (i.e. below the A-level stocking line), we would expect stands to increase in basal area and decrease in number of trees per acre. Similarly, stands above 100% stocking would be expected to gradually decrease in number of trees without substantially increasing in basal area. In either case, the

expectation is for stands to move through time toward the 100% relative density line. But because the reference line of maximum density is only an approximation of an average state that fluctuates, the growth of any given stand can be expected to vary about the line after reaching it. A relation between relative density and stand growth is implicit in oak stocking diagrams. Gross volume growth is assumed to be maximum and nearly constant when stands are fully stocked, i.e. maintained between A- and B-level reference lines. Except as noted above, merchantable volume growth is often maximized by maintaining stand density near B-level. Stocking charts also can be used in conjunction with growth and yield models to compare stand growth responses to various combinations of initial basal area and trees per acre (Leary and Stanfield, 1986; Goelz, 1991). Although the relative density lines shown on stand density diagrams are independent of site

Chapter 6

7 st

oc

ke d

100

80

100 120 140

275 Trees per acre

Fu lly

110

60

B

110

120

300

t

130

Av A er 13 ag e 12 tre e 11 di O am A ver 10 et st er oc 9 (in ke .) d 8

1 per 00 cen


140

Stand density index

15

250

225

Sto 90 ckin g

250

90 100

200

300

400

6

Trees per acre

7

8

9

10

11 12

Average dbh (inches)

22 20

16

120 110 100

me

18

olu

dv

an

St ex

ind

Average tree volume (ft3)

C

90 80 14

12 200

70

220 240 260 Trees per acre

280

300

Fig. 6.13. Seventeen-year growth of an upland oak stand in Missouri (shown by arrows) displayed on three types of stand density diagrams. To better illustrate stand growth, only an enlarged part of each diagram is shown. (A) Gingrich’s (1967) diagram for oak–hickory forests of the Central Hardwood Region. (B) A Reineke diagram of stand density index for upland oaks of the eastern United States. An index of 100 represents average maximum density (the dotted self-thinning in Fig. 6.1). (Adapted from Schnur, 1937.) (C) A diagram based on the relation between average tree volume and number of trees per acre for upland oak stands of the eastern United States. A stand volume index of 100 represents average maximum density. This diagram follows the format in Fig. 6.4. (Adapted from Schnur, 1937.)

quality and stand age, rates of stand growth are not. Consequently, when growth is plotted on a density diagram, the results pertain to specific site and stand conditions.

Analysis of estimated oak growth rates based on Dale’s (1972) growth and yield model applied to a wide range of relative density classes showed that the model pre-


dicted maximum net cubic-foot volume yield at basal areas as much as 20 ft2 acre1 below B-level on the Gingrich stand density diagram (Leak, 1981). Nevertheless, stands maintained below Blevel are likely to incur defects in bole quality associated with the resulting increase in epicormic branching (Dale, 1972; Sonderman, 1985). Stand density diagrams and equations therefore should be considered guides rather than rigid rules applicable to all situations. An understanding of the principles of self-thinning and stand density are prerequisite to knowledgeable application of silvi-

251

cultural systems. Whereas the preceding chapters have focused largely on ‘natural’ ecological events, including forest disturbance, the following chapters focus on silviculture. The silviculture of North American oaks is, and has been, largely centred on the manipulation of the structure, composition and regeneration of natural stands. From an ecological perspective, a silvicultural system therefore represents a planned disturbance or series of disturbances designed to achieve specific goals. In subsequent chapters the principles discussed in this chapter are applied to silvicultural methods for oak forests.

References Beck, D.E. (1986) Thinning Appalachian pole and small sawtimber stands. Society of American Foresters Publication 86-02, pp. 85–98. Bickford, C.A., Baker, F.S. and Wilson, F.G. (1957) Stocking, normality, and measurement of stand density. Journal of Forestry 55, 99–104. Canadell, J. and Rodà, F. (1991) Root biomass of (Quercus ilex) in a montane Mediterranean forest. Canadian Journal of Forest Research 21, 1771–1778. Chisman, H.H. and Schumacher, F.X. (1940) On the tree–area ratio and certain of its applications. Journal of Forestry 38, 311–317. Curtis, R.O. (1970) Stand density measures: an interpretation. Forest Science 16, 403–414. Curtis, R.O. (1982) A simple index of stand density for Douglas-fir. Forest Science 28, 92–94. Dale, M.E. (1972) Growth and yield predictions for upland oak stands 10 years after initial thinning. USDA Forest Service Research Paper NE NE-241. Drew, J.T. and Flewelling, J.W. (1977) Some recent Japanese theories of yield–density relationships and their application to Monterey pine plantations. Forest Science 23, 517–534. Ek, A.R. (1974) Dimensional relationships of forest and open grown trees in Wisconsin. University of Wisconsin Forestry Research Note July. Ernst, R.L. and Knapp, W.H. (1985) Forest stand density and stocking: concepts, terms, and the use of stocking guides. USDA Forest Service General Technical Report WO WO-44. Frothingham, E.H. (1912) Second-growth hardwoods in Connecticut. USDA Forest Service Bulletin 96. Gevorkiantz, S.R. and Scholz, H.F. (1948) Timber yields and possible returns from the mixed-oak farmwoods of southwestern Wisconsin. USDA Forest Service Lake States Forest Experiment Station Publication 521. Gingrich, S.F. (1964) Criteria for measuring stocking in forest stands. Proceedings of 1964 Society American Foresters and National Convention, pp. 198–201. Gingrich, S.F. (1967) Measuring and evaluating stocking and stand density in upland hardwood forests in the Central States. Forest Science 13, 38–53. Gingrich, S.F. (1971) Management of young and intermediate stands of upland hardwoods. USDA Forest Service Research Paper NE NE-195. Givnish, T.J. (1986) Biomechanical constraints on self-thinning in plant populations. Journal of Theoretical Biology 119, 139–146. Goelz, J.C.G. (1991) Generation of a new type of stocking guide that reflects stand growth. USDA Forest Service General Technical Report SE SE-70, Vol. 1, pp. 240–247. Goelz, J.C.G. (1995) A stocking guide for southern bottomland hardwoods. Southern Journal of Applied Forestry 19, 103–104.

252

Chapter 6

Harper, J.L. (1977) Population Biology of Plants. Academic Press, London. Husch, B., Miller, C.I. and Beers, T.W. (1982) Forest Mensuration. Wiley, New York. Hutchings, M. (1983) Ecology’s law in search of a theory. New Scientist 16, 765–767. Kershaw, J.A., Jr and Fischer, B.C. (1991) A stand density management diagram for sawtimber-sized mixed upland central hardwoods. USDA Forest Service General Technical Report NE NE-148, pp. 414–428. Krajicek, J.E. (1967) Maximum use of minimum acres. Proceedings 9th Southern Forest Tree Improvement Conference, pp. 35–37. Krajicek, J.E., Brinkman, K.A. and Gingrich, S.F. (1961) Crown competition – a measure of density. Forest Science 7, 35–42. Leak, W.B. (1981) Do stocking guides in the Eastern United States relate to stand growth? Journal of Forestry 79, 661–664. Leary, R.A. and Stanfield, D. (1986) Stocking guides made dynamic. Northern Journal of Applied Forestry 3, 139–142. Lonsdale, W.M. (1990) The self-thinning rule: dead or alive? Ecology 71, 1373–1388. McGill, D., Martin, J., Rogers, R. and Johnson, P.S. (1991) New stocking charts for northern red oak. University of Wisconsin Forestry Research Notes 277. McGill, D.W., Rogers, R., Martin, A.J. and Johnson, P.S. (1999) Measuring stocking in northern red oak stands in Wisconsin. Northern Journal of Applied Forestry 16, 144–150. McMahon, T. (1973) Size and shape in biology. Science 179, 1201–1204. McMahon, T.A. and Bonner, J.T. (1983) On Size and Life. Scientific American Books, New York. Miyanishi, K., Hoy, A.R. and Cavers, P.B. (1979) A generalized law of self-thinning in plant populations. Journal of Theoretical Biology 78, 439–442. Norberg, R.A. (1988) Theory of growth geometry of plants and self-thinning of plant populations: geometric similarity, elastic similarity, and different growth modes of plant parts. American Naturalist 131, 220–256. Nowak, C.A. (1996) Wood volume increment in thinned, 50- to 55-year-old, mixed-species Allegheny hardwoods. Canadian Journal of Forest Research 26, 819–835. Oswald, H. (1982) Silviculture of oak and beech high forests in France. Proceedings of Broadleaves in Britain, Future Management & Research Symposium, pp. 31–39. Putnam, J.A., Furnival, G.M. and McKnight, J.S. (1960) Management and inventory of southern hardwoods. USDA Forest Service Agriculture Handbook 181. Reineke, L.H. (1933) Perfecting a stand-density index for even-aged forests. Journal of Agricultural Research 46, 627–638. Roach, B.A. (1977) A stocking guide for Allegheny hardwoods and its use in controlling intermediate cuttings. USDA Forest Service Research Paper NE NE-373. Roberts, E.G. and Ross, R.D. (1965) Crown area of free-growing loblolly pine and its apparent independence of age and site. Journal of Forestry 63, 462–463. Rogers, R. (1983) Guides for thinning shortleaf pine. USDA Forest Service General Technical Report SE SE-24, pp. 217–225. Sampson, T.L. (1983) A stocking guide for northern red oak in New England. MS thesis, University of New Hampshire, Durham. Sampson, T.L., Barrett, J.P. and Leak, W.B. (1983) A stocking chart for northern red oak in New England. University of New Hampshire Agricultural Experiment Station Research Report 100. Schnur, G.L. (1937). Yield, stand, and volume tables for even-aged upland oak forests. USDA Technical Bulletin 560. Smith, H.C. and Gibbs, C.B. (1970) A guide to sugarbush stocking. USDA Forest Service Research Paper NE NE-171. Sonderman, D.L. (1985) Stand density – a factor affecting stem quality of young hardwoods. USDA Forest Service Research Paper NE NE-561. Sorrensen-Cothern, K.A., Ford, E.D. and Sprugel, D.G. (1993) A model of competition incorporating plasticity through modular foliage and crown development. Ecological Monographs 63, 277–304. Sprugel, D.G. (1984) Density, biomass, productivity, and nutrient-cycling changes during stand development in wave-regenerated balsam fir forests. Ecological Monographs 54, 165–186. Stout, S.L. (1991) Stand density, stand structure, and species composition in transition oak stands of northwestern Pennsylvania. USDA Forest Service General Technical Report NE NE-148, pp. 194–206.


253

Stout, S.L. and Larson, B.C. (1988) Relative stand density: why do we need to know? USDA Forest Service General Technical Report INT INT-243, pp. 73–79. Stout, S.L. and Nyland, R.D. (1986) Role of species composition in relative density measurement in Allegheny hardwoods. Canadian Journal of Forest Research 16, 574–579. Trimble, G.R. Jr (1968) Multiple stems and single stems of red oak give same site index. Journal of Forestry 66, 198. Vezina, P.E. (1963) More about the crown competition factor. Forestry Chronicle 39, 313–317. Walker, N. (1956) Growing stock volumes in unmanaged and managed forests. Journal of Forestry 54, 378–383. Weller, D.E. (1987a) A reevaluation of the 3/2 power rule of plant self-thinning. Ecological Monographs 57, 23–43. Weller, D.E. (1987b) Self-thinning exponent correlated with allometric measures of plant geometry. Ecology 68, 813–821. Weller, D.E. (1989) The interspecific size–density relationship among crowded plant stands and its implications for the 3/2 power rule of self-thinning. American Naturalist 133, 20–41. Westoby, M. (1984) The self-thinning rule. Advances in Ecological Research 14, 167–225. White, J. (1981) The allometric interpretation of the self-thinning rule. Journal of Theoretical Biology 89, 475–500. White, J. (1985) The thinning rule and its application to mixtures of plant populations. In: Studies on Plant Demography. Academic Press, New York, pp. 291–309. White, J. and Harper, J.L. (1970) Correlated changes in plant size and number in plant populations. Journal of Ecology 58, 467–485. Wilson, F.G. (1946) Numerical stocking in terms of height. Journal of Forestry 44, 758–761. Yoda, K., Kira, T., Ogawa, H. and Hozumi, K. (1963) Self-thinning in overcrowded pure stands under cultivated and natural conditions. Journal of Biology 14, 107–129. Zeide, B. (1985) Production of unmanaged bottomland hardwoods in Arkansas. Proceedings of the Central Hardwood Forest Conference V. University of Illinois, Urbana-Champaign, pp. 118–124. Zeide, B. (1987) Analysis of the 3/2 power law of self-thinning. Forest Science 33, 517–537. Zeide, B. (2001) Thinning and growth: a full turnaround. Journal of Forestry 99(1), 20–24. Zhang, L., Oswald, B.P., Green, T.H. and Stout, S.L. (1995) Relative density measurement and species composition in the mixed upland hardwood forests of North Alabama. USDA Forest Service General Technical Report SRS SRS-1, pp. 467–472.

7 Even-aged Silvicultural Methods

Introduction Two silvicultural systems of managing forests are generally recognized: even-aged and uneven-aged (Smith, 1986). This chapter focuses on the methods used in the even-aged silviculture of oak forests. The complete implementation of either system can lead to a regulated forest theoretically capable of sustaining to perpetuity an even flow of timber products and other values. The traditional objective of even-aged management is to regulate a forest by managing the stands within it as a mosaic of different age classes. The trees in each stand are allowed to grow to a specific age called the rotation age. On reaching rotation age, a stand is renewed or regenerated by a final harvest that requires the application of one of several even-aged regeneration methods. Even-aged forest management is said to be based on area control because it relies on regulating forest yield by creating and maintaining stands of various age classes, with each class occupying an approximately equal area of the forest. A regulated even-aged forest thus consists of a balanced distribution of stand age classes that is maintained through time. This arrangement also sustains an even and continuous flow of wood products and a constant proportion of stands in each age class. Diameter frequency distributions of mature even-aged oak stands are often bell-shaped, i.e. normally distributed (Fig. 5.4). There also is growing interest in maintaining forests in 254

other specified conditions for social and ecological reasons. Such considerations may require a perspective different from that considered by silviculturists in the past. Although there are only two types of management systems, three types of tree age distributions have been recognized: (i) even-aged; (ii) two-aged; and (iii) unevenaged (Smith, 1986). Even-aged stands are defined as those where the difference in age between the oldest and youngest trees does not exceed 20% of the rotation. Uneven-aged stands contain at least three age classes intermingled on the same area. Two-aged stands are comprised of two age classes of trees. In silvicultural practice, a new stand resulting from the final harvest of an evenaged stand is usually considered even-aged regardless of the actual distribution of tree ages in the new stand. This convention requires a flexible definition of the evenaged state. One definition employs the term cohort to refer to all the trees, arising anew or from advance (pre-established) reproduction, that originate from a silvicultural or natural event that produces a canopy gap or large opening in the forest (Oliver and Larson, 1996). Members of a cohort are considered even-aged regardless of their actual biological ages. This definition differs from that used in plant population biology, where cohort usually denotes membership in a group of plants originating from a single seed crop (Harper, 1977). The broader silvicultural definition facilitates reference to single- and multiple-

Even-aged Silvicultural Methods

cohort stands and thus many commonly encountered tree populations resulting from natural and silvicultural regeneration events (Fig. 5.13). Spatial scale also can be considered in determining whether a stand is even-aged or uneven-aged. For example, several cohorts of trees may occur within spatial scales smaller than that normally defined as a stand. The definition of the even-aged state therefore may vary depending on the spatial scale that is relevant to management objectives. Defining the uneven-aged state similarly depends on considerations of spatial scale (Chapter 8).

Natural Regeneration Methods Stands under even-aged management are regenerated at or near the end of the rotation by one of three silvicultural methods: (i) clearcutting; (ii) shelterwood; or (iii) seed tree methods. In their usual application, the parent stand is completely removed in one or more steps that eventually leave the site free of overstorey shade. Overstorey removal allows new reproduction to become established and the advance reproduction, if present, to develop in full light. All three methods can be used for regenerating a wide range of species from shade tolerant to shade intolerant. However, their use is often necessary for regenerating intolerant and mid-tolerant species such as the oaks. There also are variations within each method that provide additional flexibility for attaining regeneration and other objectives. Although evenaged regeneration methods are applied at the end of the rotation (and thus the end of the life of the parent stand), their application is intended to initiate the new stand.

The clearcutting method Clearcutting probably is the easiest to apply and most economically efficient of the regeneration methods. In its simplest application to oak forests, the method requires only the removal of the overstorey

255

to release the oak reproduction beneath it. The success of the method depends on preexisting (‘advance’) reproduction to replace the parent stand. New reproduction of nonoaks from seed stored in the forest floor, root sprouts and newly arriving wind-disseminated seeds also can become a part of the new stand. The method’s simplicity derives from the one-step removal of the overstorey. Its economic efficiency obtains from the minimization of logging costs per unit of tree volume harvested. However, for environmental and social reasons, the method also is the most controversial of the regeneration methods. Consequently, before the method is applied, several factors should be weighed, including: (i) the ecosystem-specific suitability of clearcutting for meeting oak regeneration requirements; (ii) predictability of the regeneration outcome; and (iii) economic, environmental and social considerations.

Suitability to oak regeneration requirements Clearcutting has been used successfully to regenerate oaks in many of the drier oak forests of the Central Hardwood Region and elsewhere where oak advance reproduction intrinsically accumulates (Roach and Gingrich, 1968; Sander, 1977; Johnson, 1993). In mature oaks stands (Fig. 7.1A), such accumulation may be sufficient for replacing the parent stand after clearcutting. Regeneration success depends on a reproduction establishment period of a decade or more before the final harvest is made (Sander, 1971, 1977). Where regeneration guidelines are available, stand regeneration potential can be objectively evaluated from an inventory of the overstorey (from which stump sprouts originate) and advance reproduction (Fig. 7.1B). Oak advance reproduction of the requisite size and spatial distribution must be present at the time of final harvest if oaks are to become a major part of the next stand (Sander et al., 1984). If stand regeneration potential is deemed adequate, all trees ≥2 inches dbh should be cut if the management objective is exclusively timber production. However, clearcutting a stand without carefully considering its

256

Chapter 7

A

B

Fig. 7.1. (A) A mature even-aged mixed-oak stand in central Missouri (Central Hardwood Region). The stand is fully stocked and dominated by northern red, white and black oaks. Site index for red oak ranges from 65 to 70 ft at an index age of 50 years. Successfully regenerating such stands to oak by clearcutting depends on the oak regeneration potential. (B) An inventory of the oak advance reproduction coupled with the application of a regional regeneration guide (e.g. Dey et al., 1996b) provides an objective basis for predicting future stand composition and the utilization of growing space by oaks and other species. (USDA Forest Service, North Central Research Station photographs.)

regeneration potential may result in a new stand with few oaks. Where the oak regeneration potential is low, clearcutting will abruptly shift species composition from oak to non-oak. A common outcome is a mixture of less desirable species and poorly distributed stocking of oak stump sprouts (Fig. 7.2A). Cutting only trees with commercial value usually leaves stands of poor quality and undesirable species (Fig. 7.2B). The

residual trees are likely to capture much of the growing space at the expense of more desirable but smaller reproduction. Where management requires retaining overstorey trees to meet non-timber objectives, retained trees should be selected by silvicultural design rather than by the logger (Smith et al., 1989). When properly applied, the method is called clearcutting with reserves and creates, at least temporarily, a two-aged stand (Helms, 1998). A few


257

A

B

Fig. 7.2. (A) Twenty years after clearcutting, this stand in southwestern Wisconsin (Central Hardwood Region) is largely comprised of northern red oak stump sprouts; site index for northern red oak is 65 ft. Stocking of oak stump sprouts is poorly distributed and the remaining stocking is poor quality black cherry and American elm. (B) A commercial clearcut 5 years after harvesting an oak stand in southeastern Ohio (Central Hardwood Region). Although the residual trees are of poor quality, their crowns and the growing space they occupy will expand and hinder the development of the established reproduction. Some residual trees can be retained in clearcuts for non-timber objectives, but it should be by design rather than ‘loggers choice’. (USDA Forest Service, North Central Research Station photographs.)

trees per acre can be retained for aesthetic purposes, den trees and snags for wildlife, and acorn production for wildlife. However, oaks retained as seed trees for regeneration purposes are usually ineffective after clearcutting.

Immediately after clearcutting, the site may temporarily appear to be inadequate in tree reproduction and vegetative cover (Fig. 7.3A). However, within a year, a dense new growth of herbaceous and woody species quickly occupy the site. Some of

258

Chapter 7

A

B

C

Fig. 7.3. (A) A clearcut oak stand in the Central Hardwood Region during the first year after harvest. New growth develops rapidly from advance reproduction of trees and shrubs, stump sprouts, seed stored in the forest floor and seed disseminated from surrounding vegetation. (B) Three years after clearcutting, this stand in the Ozark Highlands of Missouri (Central Hardwood Region) was well stocked with oak seedlings and seedling sprouts. (C) Twenty years after clearcutting, this stand in the Ozark Highlands was at 100% stocking and dominated by white, black and scarlet oaks. (USDA Forest Service, North Central Research Station photographs.)


this emergent vegetation originates from plants and root systems present before harvest. Other plants originate from the seed bank in the forest floor and from windblown and animal-dispersed seed from adjacent sites. The sudden flush of growth holds soil and nutrients in place even in steep terrain. Soil erosion and site deterioration after clearcutting are usually the result of improper logging road and skid trail design, which are largely avoidable (Chapter 4). Usually within 2 or 3 years, trees begin to dominate the site (Fig. 7.3B); by stand age 20, trees fully occupy the growing space (Fig. 7.3C). In the Ozark Highlands, 400–600 stems of oak reproduction per acre from 3 to 5 ft tall may provide adequate future stocking, depending on site factors (Sander et al., 1984). However, oak stump sprouts can compensate for deficiencies in stocking from advance reproduction. Stump sprouts may be especially important in previously unthinned stands where numerous smalldiameter white oaks often comprise a subordinate canopy layer. Previously thinned stands are therefore likely to produce few stump sprouts if thinning has been properly applied by concentrating removals in the overtopped and intermediate crown classes and rotation ages are 80 years or longer. Clearcutting is not an effective regeneration method in all xeric oak forests. For example, black oak and white oak stands on droughty outwash sands in northern Lower Michigan often fail to regenerate after clearcutting even when there is abundant oak advance reproduction (Johnson, 1992a). In that region, regeneration failures occur even where oak advance reproduction is five to ten times greater per unit area than under similar oak stands in the Ozark Highlands. The site index, composition, structure, stocking and yield of oak stands in both regions are similar. The failure of clearcutting to regenerate the Michigan forests appears to be related to: (i) the small size of the oak reproduction (mostly less than 1 foot tall); and (ii) its relatedly slow growth and high mortality after overstorey removal. The latter is associated with the post-clearcutting development of a dense

259

mat of sedges (Carex spp.), which in turn kills most of the established oak reproduction (Johnson, 1992a; Congdon, 1993). These failures suggest the need for regeneration prescriptions that increase light to levels sufficient for the development of oak advance reproduction without stimulating the development of sedges. Regeneration failures related to clearcutting these northern oak forests also may be related to the removal of the frost-protecting overstorey canopy. In contrast, populations of oak advance reproduction in the Missouri forests, although of lower density than those in Michigan, include hundreds of large stems per acre that, after clearcutting, can rapidly capture growing space. Moreover, woody plants are the primary competitors of oak reproduction in the Ozark forests and late spring frosts occur there less frequently than in the northern forest. The contrasting regeneration ecology of these superficially similar oak forests emphasizes the sensitivity of oak regeneration to factors that may not measurably affect growth and yield. Such differences in ecosystem reactions to clearcutting, and disturbances in general, point out the importance of distinguishing among ecologically defined classes of oak forests (Chapter 1). In mesic oak forests, clearcutting usually fails to restore oaks to their preharvest level of importance (McGee and Hooper, 1970; Johnson, 1976; Beck and Hooper, 1986; Loftis, 1988; Stanturf et al., 1997; Jenkins and Parker, 1998). There, clearcutting often accelerates succession towards shade tolerant hardwoods such as sugar maple, red maple and American beech, or fast growing intolerant species such as yellow-poplar, white ash and black cherry (Abrams and Nowacki, 1992; Jenkins and Parker, 1998). Even when oak advance reproduction is abundant, it typically is suppressed after overstorey removal by the growth of non-oak stump sprouts and other competition (McGee and Hooper, 1970; Beck and Hooper, 1986). For example, 20 years after clearcutting an oak/yellow-poplar stand on a very productive site (yellowpoplar site index 100+ ft) in the southern Appalachians, the stand was dominated by

260

Chapter 7

A survey of 29 clearcut oak stands in the Ohio Valley 5–26 years after harvest showed that the amount of oak reproduction varied with time since clearcutting and site quality. The oak volume of the parent stands, a mixture of upland oaks (chestnut, white, black and scarlet oaks), comprised at least 60% of the total volume. On poor sites (oak site index 50–59 ft), dominant and codominant oaks became increasingly more

yellow-poplar and other non-oaks (Beck and Hooper, 1986). This conversion occurred even though there were more than 5000 stems per acre of oak advance reproduction (Fig. 7.4). Nevertheless, oaks sometimes succeed after clearcutting mesic forests if a large proportion of the overstorey oaks sprout (P.S. Johnson, 1975; Wendel, 1975), or where competition is not severe or is controlled (Johnson et al., 1989; Jacobs and Wray, 1992). 10,000

Trees per acre

A

Sweet birch

1,000

Other Yellow-poplar Oaks

100

Red maple

10 0

5

10

15

20

500 B

Trees per acre

400 300 200 Other Yellow-poplar

100

Red maple

0

Oaks Sweet birch

5

10

15

20

Stand age (years) Fig. 7.4. Twenty-year change in stand composition after clearcutting a cove hardwood stand in the southern Appalachian Mountains of North Carolina (Central Hardwood Region). (A) All reproduction ≥4.5 ft tall. (B) Free-to-grow (not overtopped) reproduction. ‘Other’ species include black locust, eastern hemlock, white ash, black cherry, hickories, American basswood, blackgum, sassafras, sourwood and flowering dogwood. The oak group includes northern red, black, chestnut and white oaks. The parent stand was 53% oak and 33% yellow-poplar by volume. (From Beck and Hooper, 1986.)


abundant with stand age and accounted for 64% of all stems in stands 15 years old and older (Hilt, 1985a). On medium sites (oak site index 60–69 ft), oaks attained moderate importance after clearcutting and changed little with increasing stand age. On good sites (oak site index 70–80 ft), dominant and co-dominant oaks decreased with time and accounted for only 11% of stems in stands 15 years old and older, by which time yellow-poplar, black cherry and ash dominated the sites (Fig. 7.5). In this region, the compositional outcome is largely determined by the interaction of time and site quality: oaks ultimately emerge as dominants within two decades on poor sites (Hilt, 1985b). Yellow-poplar and other site-demanding hardwoods ultimately emerge as dominants on the more productive sites. On sites of intermediate quality, oaks often regenerate to intermediate levels of importance. The compositional and structural changes occurring after clearcutting are expressions of secondary succession. Each

silvicultural ‘outcome’ represents a point along one of several possible trajectories or successional pathways. Poor (xeric) sites typically possess few pathways (i.e. future compositional and structural possibilities) because many species, especially those that are sensitive to soil moisture stress, are excluded from those ecosystems by their failure to initially colonize and/or by high mortality rates. In mesic ecosystems, successional pathways are more numerous because virtually all species from those adapted to dry sites (xerophytes) to those that are more moisture-demanding (mesophytes) are physiologically capable of surviving and growing there. It is thus the competition environment that largely excludes oaks from the mesic ecosystems. Conversely, the moisture-demanding nonoaks usually are excluded from xeric ecosystems (Wuenscher and Kozlowski, 1971; Abrams, 1992). However, the common presence of oaks in mesic ecosystems suggests that certain kinds of disturbances can favour the oaks. Based on the outcome

70

Oaks as a percentage of all trees

261

Site index 50–59

60 50 40 30

60–69

20 70–80

10 0 5–9

10–14 Stand age-class (years)

15+

Fig. 7.5. Change in the relative proportion of oaks in relation to site index in mixed oak stands in southeastern Ohio (Central Hardwood Region). The oaks include white, black, scarlet and chestnut oaks. Other species present include red and sugar maples, black cherry, white ash, hickories, bigtooth aspen, flowering dogwood, sassafras and eastern hophornbeam. The maples and yellow-poplar dominated stands on the two better site classes by stand age 15. (Adapted from Hilt, 1985a.)

262

Chapter 7

of numerous reported trials, it is nevertheless apparent that clearcutting, by itself, often fails to maintain oaks at preharvest levels of stocking in mesic ecosystems. Despite its importance, the presumed requirement for the presence of oak advance reproduction may not be an irrevocable rule. This may be especially true in some mesic and hydric ecosystems. Silvicultural alternatives to relying on advance reproduction have been described for northern red oak in the Driftless Area of southwestern Wisconsin and adjacent states (Johnson et al., 1989; Bundy et al., 1991; Jacobs and Wray, 1992). There, the site index of red oak in the more mesic ecosystems ranges from about 65 to 70 ft (Gevorkiantz and Scholz, 1948; Gevorkiantz, 1957). On those sites, thousands of red oak seedlings per acre may become established in a single year following a good acorn crop (Scholz, 1955; Johnson, 1974). Numbers of seedlings also can be increased by mechanical soil scarification timed to coincide with an acorn crop. Scarification effectively reduces competition (Bundy et al., 1991). It also facilitates direct contact between germinating acorns and a mineral seedbed, which can increase initial seedling establishment if done before leaf fall. Scarification in combination with direct-seeded northern red oak acorns produced similar increases in seedling establishment under a shelterwood in Pennsylvania (Zaczek et al., 1997). After overstorey removal, the growth of red oak reproduction on productive sites is potentially rapid and less dependent on the development of a large root system and a long establishment period than is the case for the more xerophytic oaks. Thus, large numbers of red oak seedlings and a favourable postharvest environment can combine to effect successful regeneration. However, unlike many other mesic ecosystems, the density of competing trees in Driftless Area stands is relatively low and often comprised of species such as black cherry, paper birch, elms and hickories (Johnson, 1976; Martin et al., 1992). After clearcutting, the rapid decline of non-oaks from the main canopy is similar to that

which occurs in young red oak stands in New England (Oliver, 1978; Hibbs and Bentley, 1984). Soil scarification has been used in the Driftless Area to effectively reduce competition and simultaneously prepare a patchy seedbed of mineral soil (Bundy et al., 1991; Jacobs and Wray, 1992). One technique uses a tractor-mounted blade set 6 inches above ground. Careful application of the treatment during a good acorn crop results in mechanical breakage of low vegetation, including tree reproduction and shrubs, while simultaneously effecting only moderate soil disturbance. In one trial, the method increased the height and leaf area of red oak reproduction two growing seasons after treatment (1 year after clearcutting) (Bundy et al., 1991). A similar treatment consisting of mechanical uprooting of low vegetation during logging, or in a separate operation before or after logging, also has been proposed (Jacobs and Wray, 1992). Experience has shown that oak advance reproduction survives the treatment because of its deep taproot and capacity for sprouting after top injury. The recommended time to apply the treatment is in the autumn during a good seed year before acorns drop. Timing the treatment with acorn fall can increase numbers of seedlings as well as reduce competition. Applying herbicides to the understorey before clearcutting and concurrently with a good acorn crop also can facilitate oak regeneration. Even though the herbicide kills some of the oak advance reproduction, one trial resulted in about 1000 red oak stems per acre 0.5 inch dbh and larger 11 years after clearcutting; about 27% were co-dominant or larger (Johnson et al., 1989). Strategies for regenerating northern red oak in the Driftless Area are discussed in more detail by Jacobs and Wray (1992). In New England and the Lake States, northern red oak sometimes regenerates under pine stands if there is a nearby red oak seed source (Cline and Lockard, 1925; Crow and Isebrands, 1986). When the pine is harvested, the oak advance reproduction then may capture the site (Sampson et al., 1983). Such pine-to-oak successions may


be facilitated by the dispersal of acorns by animals. Small mammals disperse red oak acorns up to about 60 ft (Sork, 1984) while blue jays disperse them up to 2.5 miles (Fig. 2.13). Although most of the dispersed acorns are consumed by animals, even the small surviving proportion of blue jay-dispersed acorns can produce significant numbers of oak seedlings because of the large numbers of acorns that are dispersed and the favourable germination and growth environments they are carried to (Johnson and Webb, 1989). In bottomland oak forests, success in regenerating oaks by clearcutting has been mixed, ranging from successful to unsuccessful (e.g. Bowling and Kellison, 1983; Gresham, 1985a,b). There, prediction of stand replacement after clearcutting may be confounded by flooding, which may affect oak and non-oak advance reproduction differently. Physiological tolerance of bottomland oaks to inundation varies among species and tolerance is dependent on length of the flood period, depth of inundation, and other factors (McKevlin, 1992). Several bottomland non-oaks such as water hickory, water tupelo, swamp tupelo, green ash and red maple are more physiologically tolerant of inundation than other oaks (Hosner and Minckler, 1960, 1963; Hosner and Boyce, 1962; Broadfoot and Williston, 1973; McKnight et al., 1981; McKevlin, 1992). However, floods of short to moderate duration limited to the dormant season may favour oak reproduction. For example, water oak seedlings in an East Texas floodplain survived winter and early spring flooding because, unlike their shallow-rooted competitors, some of the oak seedlings germinated after flooding, survived physical damage, resisted uprooting during flooding, and resprouted when damaged (Streng et al., 1989). Consequently, oaks may comprise a relatively large proportion of the older, and thus larger, pool of advance reproduction in bottomlands because of their deep roots and relatively high survival and sprouting rates (Streng et al., 1989) (Fig. 3.1). Like the upland oaks, it is the larger advance reproduction that is

263

most likely to capture the growing space after clearcutting. The outcome largely depends on the species composition and size distribution of the advance reproduction. Different reactions to flooding among species and the mix of species present at the time of flooding thus may largely determine the regeneration outcome in bottomland oak forests. In some cases, bottomland oaks have successfully regenerated from seedlings established after clearcutting (Golden and Loewenstein, 1991; Golden, 1993; Nix and Lafaye, 1993). In two bottomland sites in South Carolina, cherrybark, shumard and water oak seedlings established after clearcutting resulted in over 500 oaks per acre that attained canopy dominance 5 years after clearcutting (Nix and Lafaye, 1993). Success was attributed to a moderately good acorn crop followed by early winter logging in wet weather, which in turn scarified the soil and buried many acorns. Despite such successes, it would be silviculturally more prudent to rely on oak advance reproduction of the requisite size and number. One guideline for regenerating bottomland hardwoods deems clearcutting a viable option only if 200–500 seedlings per acre of desirable species are present as advance reproduction (McKevlin, 1992). The composition and structure of bottomland clearcuts at the end of the first decade after cutting may not reflect the potential of oaks to ultimately dominate stands. For example, water and swamp chestnut oaks collectively increased from 32 to 44% of total stand basal area from stand age 14–22 years in a bottomland forest in Mississippi. This gain was made at the expense of sweetgum, ironwood, pine, blackgum and other species, whose collective basal areas remained relatively unchanged, but whose survival rates were lower than the oaks during that period (Fig. 7.6). As a result, the oaks emerged as the major dominant species by stand age 22. The emergence of bottomland oaks from inferior to dominant crown classes is related to spacing between oaks and competitors. When cherrybark oak com-

264

Chapter 7 60 Basal area (ft2 acre–1)

A 50

Oaks

40 Sweetgum

30 20

Ironwood Pine

10

Other Blackgum

0 14

19

22

1.00 B

Survival probability

0.95 Oaks (k = 0.977)

0.90 0.85

Other hardwoods (k = 0.959)

0.80 0.75 0.70 14

16

18

20

22

Stand age (years) Fig. 7.6. Stand development after clearcutting a bottomland oak/mixed hardwood stand in east-central Mississippi (Southern Pine–Hardwood Region). (A) Change in basal area with stand age. The oaks are predominantly water oak but include swamp chestnut oak; pine includes loblolly and spruce pines; other includes magnolia, elms, hickory and red maple. (B) Survival of oaks compared to the average of all other hardwoods. The curves are based on the negative exponential rates (k) calculated from observed 22nd-year survival (authors’ analysis). Actual survival between stand ages 14 and 22 was not observed. (Adapted from Bowling and Kellison, 1983, by permission of Society of American Foresters, Bethesda, Maryland. Not for further reproduction.)

petes with sweetgum in Mississippi lowlands, height growth and thus the outcome of competition between the two species depends on the average spacing between dominant/co-dominant trees (Clatterbuck and Hodges, 1988). When spacing between is less than 18 ft, sweetgum initially grows faster than oak. However, by stand age 32, the oaks are significantly taller than the sweetgum under these competitive conditions. In contrast, where the spacing between dominant and co-dominant competitors

is greater than about 18 ft, the height growth of oak is similar to that of competitors (Fig. 5.10B). The faster growth of oak under the more competitive (‘restricted’) condition also resulted in greater average clear bole length of oak at stand age 40 (40 ft) than in the unrestricted mode (23 ft). The ultimate height growth advantage of oak in the restricted mode occurred even though the oaks were 3–6 ft shorter than sweetgum during the brushy stand initiation stage of stand development.


Because site quality and competition vary greatly among different kinds of oak forests, ecological classification also is a potential silvicultural tool for assessing which stands are unlikely to regenerate. Defined ecological classes of forests can be used to distinguish among oak stands that appear to be similar (e.g. similar cover types) but that behave differently in their regeneration characteristics. But even within a defined ecological class, it is desirable to have a more specific basis for assessing an individual stand’s regeneration potential before clearcutting is applied because of the temporal and spatial variation in reproduction density and size. So how can the required reproduction characteristics be identified given that a stand’s regeneration potential is not realized until after the overstorey is harvested? Although experienced silviculturists may be able to assess the regeneration outcome from visual examination of the advance reproduction and overstorey, more objective methods are often desirable.

Regeneration models One method for objectively assessing regeneration potential involves the use of predictive regeneration models. Among the models applicable to clearcutting oak forests are those developed for the Ozark Highlands (Sander et al., 1976, 1984; Johnson and Sander, 1987; Dey, 1991; Dey et al., 1996b). The region is transitional to the Great Plains and includes the southern half of Missouri and extends into northern Arkansas and northeastern Oklahoma (McNab and Avers, 1994). The upland forests there are typically dominated by various combinations of black, white, scarlet, northern red, southern red, post, and blackjack oaks, and other hardwoods that are sometimes mixed with shortleaf pine and eastern redcedar (Braun, 1972). Site index for black oak and scarlet oak ranges from about 40 to 80 ft at an index age of 50 years (McQuilkin, 1974). The predominant associated hardwoods include hickories, sassafras, blackgum and flowering dogwood. Although shortleaf pine and eastern

265

redcedar are frequently absent from individual stands, the associated hardwoods are usually present. In these relatively dry forests, oak reproduction typically accumulates beneath the canopy of the parent stand for several decades (Liming and Johnston, 1944). There, the shoots of oak reproduction repeatedly die back and the roots of survivors slowly attain large size. Typical of ecosystems where oak reproduction accumulates, the oak advance reproduction remains largely in a suppressed state until disturbance substantially reduces overstorey density. The reproduction present at the time of disturbance is an important component of the initial state of the new stand because of its potential for capturing growing space after disturbance. The oaks of the Ozark Highlands are ‘persistent’ (sensu Veblen, 1992) because their presence is sustained over successive generations. This persistence contrasts with the more mesic oak forests to the north and east and with the bottomland oak forests to the south where dominance by oaks often may last only one generation because of successional replacement by long-lived, faster growing or more shade tolerant hardwoods (R.L. Johnson, 1975; Loftis, 1990b; Nowacki et al., 1990). In contrast, the dominance of the hardwoods associated with oaks in the Ozark Highlands usually lasts for only two decades after overstorey removal. The high mortality rates and limited development of the non-oaks in this ecosystem relegates them to saplings and reproduction in older stands (Braun, 1972; Dey, 1991). Even though these forests are often called oak–hickory forests, the hickories usually comprise a minor part of the overstorey. During the first decade after overstorey removal, the composition of the reproduction largely depends on the mix of seed and plants present when the disturbance occurs. Although this initial mix itself is relatively unpredictable, the non-oaks are quickly relegated to the sub-canopy as competition intensifies and crowns close. Within two decades, the oaks have reemerged as the dominant species. This pattern of stand development conforms to the

266

Chapter 7

‘competitive sorting’ model described by Margalef (1963, 1968) and Peet (1992). However, in the Ozark Highlands the outcome is influenced by topography. The ascendence of oaks to dominance by competitive sorting is rapid and predictable on hot southwest-facing slopes and on neutral southeast and northwest slopes. It is less certain on cool northeast slopes where the pre-disturbance accumulation of oak reproduction is less pronounced and shade tolerant competitors are sometimes capable of replacing the oaks or slowing their re-emergence to dominance (Sander et al., 1984). The outcome is probabilistically specifiable for each species and depends on their size at the time of clearcutting (Fig. 7.7). Even though the non-oaks of the Ozark Highlands are usually unimportant as dominant components of older stands, they nevertheless persist from one generation to the next. Their abundance and early rapid growth after clearcutting also interfere with the re-emergence of the oaks. The oaks and associated hardwoods nevertheless re-establish a ‘compositional equilibrium’ (sensu Veblen, 1992) within two decades of disturbance. The consistency of this pattern of stand redevelopment simplifies the prediction of changes in species composition and therefore modelling the regeneration process. In the Ozark Highlands, predicting changes in species composition after clearcutting thus is less a question of losing the oaks through successional displacement than it is of how the proportions of the various oak species will change and how quickly and completely the oaks will capture growing space in the regenerated stand. ADVREGEN is a probabilistic individual-tree regeneration model applicable to the Ozark Highlands. It was developed to assess the adequacy of the oak, hickory and blackgum regeneration potential in stands considered for harvesting by clearcutting (Sander et al., 1984; Johnson and Sander, 1987). However, the model is applicable to any method of regeneration requiring complete overstorey removal on areas of about one-third of an acre or larger. The model provides a simple ‘yes’ or ‘no’ answer to

the question of whether a given stand will be adequately stocked 20 years after final harvest. The criterion for defining minimum adequacy of stocking assumes that by stand age 20 there must be at least 221 dominant and co-dominant oaks per acre averaging 4.5 inches dbh (the average expected dbh of dominant and co-dominant trees in 20-year-old stands). That level of stocking equates to C-level stocking on Gingrich’s (1967) stocking chart (Fig. 6.9). The trees in stands at C-level stocking will fully utilize growing space within 10 years (i.e. by about stand age 30) assuming that trees are well distributed (Gingrich, 1967; see also Chapter 6). Application of the model is relatively simple and is facilitated by the ADVREGEN computer program or by referring to related tables (Sander et al., 1984). To use the model, the height and basal diameter of the largest stem of advance reproduction must be measured on 1/735-acre (4.3-ft radius) field plots and the slope position and aspect of each plot must be specified. A 1/735-acre plot equals the minimum growing space required for a tree 4.5 inches dbh as defined by Gingrich (1967). Co-dominant and dominant trees averaging 4.5 inches dbh occur in stands averaging about 3 inches dbh (based on trees 1.6 inches dbh and larger). Because 3 inches is the smallest mean stand diameter shown on Gingrich’s (1967) stocking chart, it represents the earliest point in stand development (approximately stand age 20 on average sites) that the utilization of growing space by trees (stocking) can be practically determined. ADVREGEN also facilitates estimating the probable contribution of stump sprouts to future stocking. To do that, the diameters of overstorey oaks must be sampled and stand site index must be determined. The regeneration model is based on estimates of the probability that a seedling or seedling sprout of a given initial size (basal diameter and height) will survive and grow to dominant or co-dominant crown class after clearcutting. For a given species, large initial diameters and heights are associated with high probabilities. The


267

A

Fig. 7.7. (A) Large oak advance reproduction such as this 7-ft tall seedling sprout in a Central Hardwood stand has a high dominance probability, i.e. a high chance of becoming a dominant or co-dominant tree after complete overstorey removal. (USDA Forest Service, North Central Research Station photograph.) (B) Dominance probabilities for advance reproduction (seedling or seedling sprout) of different species in the Ozark Highlands of Missouri (Central Hardwood Region). In these examples, dominance probability is the probability that a tree will attain an intermediate-orlarger crown class 21 years after clearcutting. In each case, probabilities are for 6-ft-tall advance reproduction with 1-inch basal diameters growing on neutral aspects (southeast- or northwestfacing) mid-slopes. The five species groups shown are the predominant hardwoods within this ecoregion. Probabilities were generated by the regeneration model ACORn (Dey et al., 1996b). 0.6

B

Dominance probability

0.5

0.4

0.3

0.2

0.1

0.0 Hickory

Sassafras

Blackgum

resulting estimates, called dominance probabilities, are generated by a predictive equation (Fig. 7.8). Dominance probabilities are extended to stand age 20 by assuming an annual negative exponential ‘retainment’ rate of 0.99 (Johnson and Sander, 1987). A ‘stocking value’ for each

Dogwood

Oaks

plot is then calculated from the reciprocal of the 20-year dominance probability and the binomial probability distribution (Sander et al., 1984; Johnson and Sander, 1987). If a stand’s mean stocking value equals or exceeds 30%, the stand is deemed adequately stocked.

268

Chapter 7

0.8 a

0.6 0.5 0.4 0.3 1.75 1.50

0.2 b

0.1

1.25

Ba sa ld iam et er (in ch es )

Dominance probability

0.7

1.00

0.0 5

0.75

4

3

2

Heigh

t (ft)

0.50 1

0 0.25

Fig. 7.8. Estimated dominance probabilities for oak advance reproduction in relation to initial (preharvest) size of reproduction in the Ozark Highlands of Missouri (Central Hardwood Region). Dominance probability is here defined as the probability that seedling or seedling sprout of a given height and basal diameter immediately before clearcutting will be dominant or co-dominant 5 years later. (a) Probabilities for mid-slopes on neutral aspects (southeast- and northwest-facing slopes); (b) Probabilities for lower slopes on cool aspects (northeast-facing slopes). Probabilities for other aspect/slope combinations lie between the two response surfaces shown, which represent averages for black, white, scarlet, post and blackjack oaks. (Derived from a logistic regression model from Sander et al., 1984.)

The ADVREGEN model is simple in concept and application because the definition of a dominance probability integrates growth and survival into a single value. The specified size of the inventory plot, which considers the future growing space requirements of trees, also simplifies prediction and field application. Although the model does not explicitly consider competition effects, such effects are implicit in the effects of topographic factors (advance reproduction) or site index (stump sprouts) on dominance probabilities. The limitations of ADVREGEN include the lack of specification of the composition and structure of the future stand, the use of a single stocking criterion, restriction to stands where there will be complete overstorey removal (i.e. where all trees 2 inches dbh and larger will be cut), regeneration stock-

ing estimates that are limited to oaks, hickories and blackgum, and a lack of generality of application to conditions not considered by the data from which the model was built. Model development and application are discussed in more detail elsewhere (Sander et al., 1984; Johnson and Sander, 1987). A similar regeneration model applicable to northern red oak reproduction in the southern Appalachians predicts 20th-year dominance probabilities from preharvest basal (ground line) diameter and site index. Probabilities increase with increasing basal diameter and decreasing site index across the site index range of 70–90 ft (Table 7.1). The smaller dominance probabilities associated with the higher site indices are attributable to the increasing competition from yellow-poplar and other vegetation


269

Table 7.1. Twentieth-year dominance probabilitiesa for northern red oak advance reproduction in the southern Appalachians. Basal diameterb of advance reproduction (inches) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1–1.5 1.6–2.0

Oak site indexc (probability) 70 0.01 0.02 0.04 0.06 0.09 0.13 0.17 0.21 0.25 0.29 0.38 0.46

80 0 0.01 0.02 0.03 0.04 0.06 0.09 0.12 0.15 0.18 0.29 0.41

90 0 0 0.01 0.01 0.02 0.03 0.04 0.06 0.08 0.11 0.19 0.34

aThe

probability that a stem of red oak advance reproduction will be dominant or co-dominant 20 years after overstorey removal. (From Loftis, 1990a.) bMeasured at ground line. cHeight in feet at an index age of 50 years based on the curves of Olson (1959).

associated with increasing site quality. In application, the expected future numbers of dominant and co-dominant red oaks per unit area can be calculated by multiplying the observed number of seedlings in each of several basal diameter classes by the dominance probability for each class and summing those products across all diameter classes. Because northern red oak is usually a minor to moderately important component of the mixed mesophytic forests of the region, the model itself makes no assumptions about the adequacy of future red oak stocking. A model called ACORn (A Comprehensive Ozark Regenerator), like ADVREGEN, is applicable to the Ozark Highlands. It can be used to predict the composition and structure of stands 21 years after complete overstorey removal (Dey et al., 1996a,b). This probabilistic individual-tree, distance-independent model simulates the development of oak stands including all the oaks native to the Ozarks and the four major associated hardwoods (hickories, blackgum, flowering dogwood

and sassafras). ACORn makes no assumptions about the stocking adequacy of the future stand. Instead, the model predicts the future distribution of tree diameters by species (Fig. 7.9). The output from ACORn can be used as input for existing growth and yield models. This facilitates projecting stand growth and change through the next rotation. Like ADVREGEN, the application of ACORn requires an inventory of the advance reproduction and overstorey together with information on site quality as expressed by slope position, aspect and site index. Because of its complexity, using the model requires a computer. The model can be applied to ‘real’ stands or used to compare various hypothetical situations. The ACORn model assumes that the height growth of reproduction, including advance reproduction and stump sprouts, is related to the apparent size of the root system rather than reproduction origin, per se. Accordingly, all forms of reproduction including new seedlings, seedling sprouts and stump sprouts represent a continuum of growth potential determined by root

Chapter 7

24

A

Data input: overstorey

20

Trees per acre

White oak 16 Black oak 12 Other species 8 4 0

Advance reproduction (genets/acre)

2.5

4.5

6.5

8.5 10.5 12.5 14.5 Dbh class (inches)

16.5

18.5

B

Data input: advance reproduction

500 400 300 200 100 0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1

1.1 1.2

Basal diameter class (inches)

24

C

400

Model output: regenerated stand at age 21

2* Trees per acre

270

300

8

*Stocking % in dbh class

200 15 100

22 15 4

3

1

6.5

7.5

8.5

0 0.5

1.5

2.5

3.5 4.5 5.5 Dbh class (inches)


size, which in turn is correlated with basal tree diameter. Rate of height growth at first increases rapidly with increasing initial basal diameter, reaches a maximum, and then decreases (Fig. 2.27). For oak advance reproduction, initial height also is considered by the model. However, for various physiological and ecological reasons, initial size relations explain only a small to moderate proportion of the variance in the postharvest growth of reproduction. To produce realistic simulations of the size distributions of trees, ACORn jointly employs the regression estimates of future tree heights and the observed variation around those estimates (Dey, 1991). The prediction errors are used to estimate the probability that a seedling, seedling sprout or a stump sprout will attain a specified 21st-year diameter based on their initial size and site-related factors. These probabilities then are reduced by multiplying them by estimated survival probabilities. The resulting individual-tree probabilities facilitate predicting future diameter distributions of trees that survive and grow to stand age 21 based on a preharvest inventory of the advance reproduction and the overstorey. ACORn is limited to predicting stand composition and structure at stand age 21, and like ADVREGEN is limited to predicting the outcome after complete overstorey

271

removal. Also, contributions to future stocking from reproduction originating from seed cannot be accounted for. The latter eliminates prediction of two regionally important species, eastern redcedar and shortleaf pine. FORCAT is a regeneration model designed to simulate the development of clearcut stands on the Cumberland Plateau (Waldrop et al., 1986). The Cumberland Plateau is centred in eastern Tennessee and extends northwards into eastern Kentucky and southwards into northern Alabama (Smalley, 1979, 1982, 1984, 1986). This is a floristically diverse region that includes forests dominated by oaks, pines and numerous other hardwoods. Site index for upland oaks ranges from about 50 to 80 ft at an index age of 50 years. Variation in the region’s undulating to steep topography, together with soil depth and soil texture, largely controls site quality and therefore species composition, succession and regeneration dynamics. The drier sites, which are characterized by southerly-facing slopes, dry ridges and thin soils, are usually dominated by oaks and hickories sometimes mixed with shortleaf pine, Virginia pine and eastern redcedar. Common associated hardwoods include blackgum, red maple, sourwood, flowering dogwood and sassafras (Waldrop et al., 1986). On the drier sites, oak reproduction

Fig. 7.9. (Opposite) Graphical representation of input and output for the regeneration model ACORn (A Comprehensive Ozark Regenerator) (Dey et al., 1996b). The model predicts, from a preharvest stand inventory (tree list), the 21st-year composition, structure and stocking of oak stands that are proposed for complete overstorey removal in the Ozark Highlands of Missouri (Central Hardwood Region). (A) Diameter distribution of overstorey trees. ACORn requires an inventory of the diameters of overstorey trees (listed by species) and an estimate of stand site index. This facilitates predicting contributions to the new stand from stump sprouts originating from cut overstorey trees. The graph represents a mature black oak/white oak stand at 84% stocking (based on Gingrich, 1967) growing on a southeast-facing slope (black oak site index 63). The stand is dominated by black oak. (B) Distribution of basal diameters of advance (preharvest) reproduction. The model requires an inventory of the advance reproduction (a list of basal diameters and heights by species) to predict its contribution to the future stand. Information on the slope position and aspect of each reproduction sample plot or the stand as a whole also is required. In this example, white oak reproduction predominates. (C) Model output: the regenerated stand. Based on input from A and B, the model predicts the new stand will be dominated by white oak, which contributes 72% stocking of the total projected 94% at stand age 21. Other species include hickories, blackgum, sassafras and flowering dogwood. The computer model produces graphical output similar to that shown in (C) as well as tables. Output can be obtained by crown classes, source of reproduction (stump sprout or advance reproduction) and user-specified species or species-groups. (Adapted from Dey et al., 1996b.)

272

Chapter 7

often naturally accumulates. Stand structure, composition and succession therefore are similar to the Ozark Highlands and other xeric oak forests of the Central Hardwood Region. On the richer sites on north- and eastfacing slopes, valley bottoms and coves, stands are often dominated by diverse mixtures of tolerant and intolerant mesophytes such as yellow-poplar, sugar maple, American beech, black cherry and white ash mixed with minor to moderate proportions of northern red, white and other oaks (Braun, 1972). Lush subcanopies of flowering dogwood, sourwood, bigleaf magnolia, umbrella magnolia, American hornbeam and other shade tolerant species are common (Carpenter, 1976). Because oak advance reproduction usually does not accumulate under those conditions, regenerating oaks is problematic on these sites. FORCAT predicts future stand composition and structure from measurements of the initial state, i.e. dbh, height and species composition of the parent stand overstorey. The model can consider up to 33 tree species and can simulate the composition and structure of stands of any age up to 100 years. The model user also can begin the simulation at any stand age beginning with the preharvest state or from subsequent years. Unlike ACORn, FORCAT does not directly consider the presence of advance reproduction. Therefore, in application, FORCAT does not require measurements or counts of reproduction. Regeneration events are simulated by various model subroutines that probabilistically account for sprouting from harvested overstorey trees, seed production and germination. Regional variation in temperature and soil moisture effects are considered by constraining the maximum growth potential of each species based on growing-season degree days and soil moisture days above specified critical thresholds. Site-specific effects are considered by providing separate values of a diameter growth constant for each species in each of 20 different ‘land types’ in the MidCumberland Plateau as defined by Smalley

(1982). Tree survival is predicted from tree age and expected diameter growth rate. The model also can simulate the effects of prescribed burning. In tests of FORCAT’s precision, predictions became increasingly more accurate with increasing stand age (up to 100 years). Its questionable ability to accurately predict the composition of young stands (e.g. at age 20) may be silviculturally problematic. FORCAT’s generalized structure makes it adaptable to a wide range of species, site and ecological conditions like those characteristic of the Cumberland Plateau. This flexibility is derived from the model’s lineage to FORET, a gap model designed to predict long-term forest succession in Tennessee (Shugart and West, 1977). Another regeneration model covers the Allegheny Hardwood Region from the Allegheny Plateau of northwestern Pennsylvania (ecoregion province 211, Fig. 1.2) northwards into New York where it is transitional to the Northern Hardwood Region (ecoregion province 221a, Fig. 1.2) (Marquis et al., 1992). To the south, the applicable area extends into West Virginia and Maryland (ecoregion provinces 221a and M221, Fig. 1.2). Complete silvicultural guidelines, including the regeneration model are available for three commonly occurring hardwood forest types within the region: cherry–maple, beech–birch– maple and oak–hickory (Marquis et al., 1992). The latter occur primarily on the drier sites. However, one or more species of oak (including northern red, white, black, scarlet and chestnut oaks) are potential components of all three types. Important non-oaks include black cherry, red maple, sugar maple, American beech, yellow birch, sweet birch, hickories, white ash, yellow-poplar, pin cherry, striped maple eastern hophornbeam and other species (Marquis et al., 1975). Serious regeneration problems often occur throughout the region because of heavy deer browsing and competition from ferns, grasses and shade tolerant sub-canopy tree species (Marquis, 1974, 1981; Horsley, 1982; Horsley and Marquis, 1983).


Comprehensive guidelines are available for prescribing silvicultural treatments for stands within the region (Marquis et al., 1992). Application of the guidelines is facilitated by a computer program, SILVAH, which incorporates criteria for evaluating stand regeneration potential. SILVAH also produces printed output containing data summaries and stand diagnostics. The regeneration model included in SILVAH is applicable to stands under even-aged management. Using the model requires preharvest counts of reproduction by species groups within sample plots of 6 ft radius. The counts are weighted by height and vigour classes and adjusted for interfering factors including intensity of deer browsing, soil characteristics, and competition from non-commercial tree species and other plants. The resulting rating then is used to determine whether a reproduction plot is stocked (Marquis et al., 1992). A stand’s regeneration potential is deemed adequate if at least 70% of plots are rated as stocked. Regeneration from sapling-size trees (0.5–6 inches dbh) also can be considered. Three commercially valuable species groups are recognized: black cherry, oaks and other ‘desirable’ species. Other tree species are categorized as ‘undesirable’ and are treated by the model as an ‘interfering’ factor in the regeneration of the commercially valuable species. There also is a provision for evaluating regeneration after final harvest. Regeneration adequacy can be projected by using the SILVAH program or by hand cal-

273

culation using the tables and decision rules in Marquis et al. (1992). The regeneration model was developed from the relation between the observed density of advance reproduction before final harvest and stand stocking after harvest (Grisez and Peace, 1973; Marquis and Bjorkbom, 1982). The model emphasizes black cherry because of its regional importance and considers other species, e.g. oaks and other ‘desirable’ species, as rather broad species groups. Stocking criteria for oaks were largely adapted from those developed in other regions and modified by general observations on oak regeneration within the Allegheny region (Marquis et al., 1992). Because the model is in part statistically derived, it possesses some of the characteristics of other statistically based models as discussed above. However, the model differs from the models previously discussed because of its partly statistical, partly expert opinion design. In southern bottomland oak stands, regeneration potential can be assessed by assigning a score of 1–3 to trees (both advance reproduction and overstorey) within 1/100-acre plots (Johnson and Deen, 1993). Small advance reproduction is assigned a value of 1 whereas larger reproduction is assigned a value of 3; conversely, small overstorey trees are assigned a value of 3 and larger trees a value of 1 or 0, which reflects their stump sprouting potential (Table 7.2). Because the scoring system is relatively simple, an inventory of advance reproduction and the overstorey

Table 7.2. Scoring system for assessing the adequacy of the regeneration potential of 1/100-acre plots in southern bottomland hardwood stands.a Advance reproduction height (ft) 3.0 Score aTo

1

2

3

Overstorey tree dbh (inches)b six seedlings per square foot) will result in small seedlings with limited growth capacity. Obtaining a desirable density therefore requires adjusting sowing rates based on the percentage of acorns expected to germinate. After sowing, the seedbed should be mulched to prevent acorn desiccation and to protect acorns from freezing. In addition, it is usually necessary to protect seedbeds from bird and mammal predation until germination is complete and leaves are about onehalf fully expanded. If acorns are inadvertently oversown, seedbed density can be reduced by removing (‘roguing’) seedlings (preferably the smaller ones) until the desired bed density is obtained. Roguing should be done early in the first growing season.

279

Seedlings can be lifted any time after they are fully dormant in the autumn. Cold storage at 33–38°F (0.5–3°C) and proper packaging of seedlings from time of lifting to time of planting is essential for maintaining physiological quality of seedlings (Webb and von Althen, 1980). Seedlings can be planted in the autumn, winter or spring provided they are dormant. Seedling size and root morphology at the time of planting are important determinants of the physiological quality of seedlings and thus planted tree performance (Johnson, 1988; Kormanik, 1989; Schultz and Thompson, 1989, 1990, 1991; Kaczmarek and Pope, 1993; Teclaw and Isebrands, 1993a,b; Thompson and Schultz, 1995; Dey and Parker, 1997b; Kormanik et al., 1998). Based on studies in the Eastern Hardwood Region, oak seedlings with relatively large basal diameters (e.g. ≥3/8 inch 1 inch above the root collar) perform better than smaller seedlings. Planting seedlings that have been grown for 1 year in the nursery then lifted and transplanted back to the nursery (called 1+1 transplants) have been shown to outperform 1- or 2-year-old seedlings that are not transplanted in the nursery (Johnson, 1984). Similar but more cost-effective advantages can be obtained by undercutting seedlings in the nursery the first or second year (Schultz and Thompson, 1990; Buchschacher et al., 1991; Weigel and Johnson, 1998a,b, 1999). Seedlings can be undercut by drawing a tractor-mounted blade or other root-cutting device at a prescribed soil depth – usually 6–8 inches below the soil surface (Fig. 7.11A). This practice severs the tap root and stimulates the development of lateral roots in the nursery (Fig. 7.11B) and thus the number of new roots that develop after field planting (Johnson, 1988). The combined effect of undercutting and low seedbed density can greatly increase the number of large lateral roots that are essential for post-planting root regeneration and thus successful field performance (Table 7.3). However, to obtain the full advantage of undercutting, the blade used to lift seedlings should be set below the undercutting depth. Undercutting should not be confused with root pruning,

280

Chapter 7

A

B

Fig. 7.11. (A) Undercutting northern red oak seedlings in the nursery with a tractor-drawn blade in early summer of their second year. (B) Undercutting stimulates lateral root development (0 = not undercut, 1 = once undercut, 2 = twice undercut). These seedlings are also root pruned to a common length (10 inches) to facilitate field planting. Because undercutting retards overall seedling growth, a given nursery bed will yield fewer undercut seedlings of a given diameter than not-undercut seedlings (Schultz and Thompson, 1990). A favourable time to undercut is when seedlings are in the resting (lag) phase between first and second flushes of shoot growth. The date of occurrence of this stage may occur any time from late spring to mid-summer, depending on nursery latitude, weather, nursery management practices and other factors. (USDA Forest Service, North Central Research Station photographs.)

which is done after lifting to reduce roots to a common length (Fig. 7.11B) to accommodate proper placement in a planting hole of the same depth.

Removing seedling tops 6–8 inches above the root collar just before planting further improves planted tree performance when trees are planted under shelterwoods.


281

Table 7.3. Response of northern red oak seedlings to variation in nursery bed density and undercutting treatments.a

Bed density/ (undercutting treatment)b 3/ft2 (undercut) (not undercut) 6/ft2 (undercut) (not undercut) 12/ft2 (undercut) (not undercut) Statistical significancee

Mean seedling Mean seedling heightc diameterc (inches) (inches) 15.3 20.9 14.9 19.1 15.2 18.1 D, U, DU

0.32 0.39 0.29 0.33 0.27 0.30 D, U, DU

Mean number of lateral roots > 0.04 inch (1 mm) diameter Normal Wound 1st order rootsd Total 13.4 10.8 10.6 8.8 8.7 6.6 D, U

6.1 — 5.8 — 6.2 —

19.5 10.8 16.4 8.8 14.9 6.6

aFrom

Schultz and Thompson (1990); based on data from the Illinois State Forest Nursery. beds were reduced to the prescribed seedling densities in the spring before the second growing season. Seedlings were undercut during the second growing season at a depth of 6 inches when taproots at that depth were 0.25 to 0.5 inches diameter. cMeasured after two growing seasons in the nursery; diameters were measured 0.5 inch above the root collar. dWound roots are roots > 0.04 inch (1 mm) in diameter that develop at or near the undercutting wound. eVariables statistically significant at = 0.01: D = bed density; U = undercutting treatments; DU = interaction between D and U. bNursery

Although shoot clipping may not benefit seedlings planted in the open (including new clearcuts), there is no evidence it reduces seedling performance (Johnson, 1984, 1988, 1989). For a given type of oak seedling, outplanting performance varies with site quality, competition from established tree reproduction and other vegetation (Johnson and Rogers, 1985).

Oak plantation establishment Oaks can be established as pure or mixed plantations on old fields or other open areas largely devoid of forest vegetation. On these sites, mechanical planters and site preparation equipment can often be used. Subsequent mechanical weeding and row thinning methods such as those commonly applied to pine plantations also can be used. In the United States, probably more is known about oak plantation establishment in southern bottomlands than elsewhere. This has partly resulted from the availability of bottomland agricultural fields that were abandoned because of frequent flood-

ing. These areas were originally covered by oaks and other lowland hardwoods and rank among the most productive forests within the temperate region. Oak site index on southern bottomlands ranges from 80 to 100 ft or more at an index age of 50 years. The productivity of bottomland forests therefore may sometimes justify relatively high investments in site preparation, planting and weeding. Species that have been successfully planted include Nuttall, water, cherrybark, Shumard, willow, white and swamp chestnut oaks. Recommendations are to plant seedlings of 3/8-inch caliper or larger with roots pruned to a length of about 8 inches. Bottomland oaks grow best on moist, welldrained soils of medium texture and high fertility. Because each bottomland oak species has specific site requirements, successful oak planting requires careful matching of species to site (Fig. 1.17). Site preparation is often necessary to obtain acceptable growth and to facilitate the use of mechanical planters or seeders. Sites that have been under cultivation for a long time may require breaking plough

282

Chapter 7

pans or other compacted soil layers with specialized equipment. Discing, mowing and herbicides can be used to speed the growth of planted oaks during the first few years. However, even without weed control, trees growing on site index 80–90 ft can be expected to reach 15–20 ft in height in 10–15 years under normal conditions (Kennedy, 1993). Cleaning may be required after that to release planted trees from overtopping competitors. Direct seeding also has been successfully used in southern bottomlands, especially in the lower Mississippi River Valley. More than ten oak species have been successfully direct-seeded (Johnson and Krinard, 1987). As in planting, species growth requirements should be matched to site. An array of experiences in planting and direct seeding of bottomland oaks are reported in more detail elsewhere (e.g. Johnson, 1981; Krinard and Francis, 1983; Kennedy and Krinard, 1985; Johnson and Krinard, 1987; Krinard and Johnson, 1988; Wittwer, 1991; Bullard et al., 1992; Schweitzer et al., 1997). Oak plantations are sometimes established using ‘tree shelters’. Tree shelters are transparent or translucent plastic tubes that protect individual trees from animal damage and also create a greenhouse effect around each tree (Potter, 1991). They are commercially available in a range of materials, durability, sizes (diameter and height) and colours (Windell, 1991). Sizes typically range from 3 to 6 inches in diameter and 4 to 5 ft in height. Five-foot shelters are recommended where deer browsing is severe or snow is deep (Kays, 1996; Schuler and Miller, 1996). Commonly available colours include white, green, tan and brown. Some are circular in cross section whereas others are square or rectangular. Trees in shelters usually grow faster than non-sheltered trees. Accelerated growth may be due to increased air temperature, carbon dioxide, and reduced wind inside the shelter (Potter, 1988, 1991; Mayhead and Jones, 1991; Windell, 1991; Minter et al., 1992). Although ambient air temperatures inside shelters may exceed 38°C (101°F), leaf temperatures on actively

transpiring trees may be 7–11 degrees lower (Potter, 1988). However, inside and outside temperatures of shelters with perforated walls may not differ (Minter et al., 1992). Several studies have demonstrated the effectiveness of shelters in increasing the survival and height growth of oak seedlings. The period of accelerated growth is most dramatic while seedlings are still in the tree shelter, and slows after they emerge. Emergence may take 2 years or longer depending on tree shelter height and other factors. In southern Britain, sheltered sessile oak seedlings grew five times faster in height than unsheltered seedlings over 4- and 5-year periods (Potter, 1988, 1991). In the United States, reported height growth of northern red oak in tree shelters ranges from about 30 to 230% greater than that of non-sheltered seedlings (cf. Lantagne, 1991; Teclaw and Isebrands, 1991; Zastrow and Marty, 1991; Minter et al., 1992; Bardon and Countryman, 1993; Smith, 1993; Walters, 1993; Gillespie et al., 1996; Ponder, 1997). Variation in growth responses among studies may be related to differences in competition, site quality, climate and weather, shelter design and colour, overhead shade, seedling quality, length of period observed herbivory and other factors. However, tree shelters have not always accelerated the growth of planted oaks (Lantagne, 1996; Teclaw and Zasada, 1996; Lantagne and Miller, 1997). Tree shelters also may increase oak seedling survival (Potter, 1988, 1991; Lantagne, 1991, 1996; Zastrow and Marty, 1991; Bardon and Countryman, 1993; Smith, 1993). Because of the relatively high cost of the tree shelters and their installation and maintenance, it may be difficult to justify the use of shelters unless animal or other types of damage are severe, future tree value is high, or other factors merit the investment. Shelters nevertheless can reduce the cost of applying herbicides because the shelter itself protects the seedling from herbicide damage. Shelters also protect against damage from mechanical equipment such as mowers and string trimmers, and facilitate


later finding planted trees for inspection and cultural treatments. Tree shelters can be used in a variety of situations ranging from conventional plantation establishment in old fields and other treeless areas, to enrichment plantings (see the following section) in clearcuts and under shelterwoods. When tree shelters are used under shelterwoods, shelters that block little light are recommended (Potter, 1991; Schuler and Miller, 1996). Other factors being equal, shelterwood density itself affects the growth of seedlings in tree shelters. In a Wisconsin study, the height growth of northern red oak seedlings in tree shelters decreased with increasing shelterwood density (Teclaw and Isebrands, 1991). But even under high-density shelterwoods (100% crown cover), seedlings in tree shelters grew faster than unsheltered seedlings. Where late spring frosts are frequent and severe (such as in the Northern Hardwood Region), tree shelters used in combination with shelterwoods may be necessary to obtain any growth advantage from the tree shelter. Tree shelters did not benefit northern red oak seedlings planted in clearcuts in northern Wisconsin because of dieback caused by late spring frosts, but did benefit seedlings under shelterwoods that provided 50% or 75% crown cover (Teclaw and Zasada, 1996). However, tree shelters had a detrimental effect on the height growth of northern red oak planted under shelterwoods in a northern Michigan study (Lantagne and Miller, 1997). Tree shelters also have potential application in sheltering natural reproduction (Potter, 1991; Kittredge et al., 1992), direct seeding with acorns (Smith, 1993; Walters 1993; Bailey et al., 1996; Schuler and Miller, 1996), rehabilitating partially failed plantings (Potter, 1991; Gillespie et al., 1996), and revegetating strip-mined lands (Farley et al., 1996). Seedlings grown in tree shelters develop a nearly columnar stem with practically no taper (Potter, 1988). Although rapid height growth helps keep the sheltered tree above competitors and out of reach of browsing animals, stems are usually so weak that the

283

tree shelter is needed for physical support. While this acceleration in stem growth occurs at the expense of root growth, the delay in root growth is more than compensated for by the third year after planting (Ponder, 1996). It nevertheless may take 5 years or longer for the tree to develop sufficient diameter to stand by itself (Potter, 1988, 1991). Tree shelters therefore must be sufficiently durable to last until trees can emerge from the shelter. Many of the commercially available tree shelters are made of polypropylene, which is not durable in sunlight unless an ultraviolet inhibitor is added (Potter, 1988; Windell and Haywood, 1996). After trees emerge from the top of the tree shelter, height growth slows and more growth is allocated to stem diameter (Schuler and Miller, 1996; Strobl and Wagner, 1996; Windell and Haywood, 1996). The top rim of a shelter also must be smooth to prevent abrasion and damage to the thin stems. Although promising in principle, much remains to be learned about tree shelters and their application to establishing oaks. Guidelines for using tree shelters with northern red oak, including direct seeding and shelterwood plantings, are presented by Schuler and Miller (1996).

Enrichment planting Enrichment planting is defined as planting to improve the proportion of desirable species or to increase biodiversity by establishing young trees among existing forest growth (Helms, 1998). The method potentially can be combined with any natural regeneration method. In oak forests, experience in enrichment planting largely has been limited to clearcutting and shelterwood methods.

Planting in clearcuts Oaks can be successfully planted in clearcuts despite the numerous reported failures of such plantings (e.g. Hilt, 1977; Loftis, 1979; McGee and Loftis, 1986). Opportunities for planting oaks in clearcuts vary among ecosystems. In

284

Chapter 7

upland oak stands in the eastern United States, probably the best opportunities for planting the major oak species occur within the oak site index range of 60–75 ft. There, site quality is sufficient to meet the growth requirements of most oak species, but competition is not too severe. Outside this site index range, it may be difficult to justify planting oaks for ecological and economic reasons. Enrichment plantings are most effectively used in conjunction with existing natural regeneration guides. Guidelines for evaluating the natural regeneration potential of oak stands are available for several regions (see ‘Regeneration models’, this chapter). Planting costs accordingly can be minimized by planting only the number of trees needed to attain a given future stocking goal (Johnson and Rogers, 1985; Johnson et al., 1986). However, this strategy requires information on the expected performance of the planted oaks. The outcome will depend on site quality, overstorey and understorey composition and structure of the present stand, planting stock size, animal browsing and other factors. Other factors being equal, planting oaks in clearcuts has been less effective than planting oaks under shelterwoods. One reason for planting failures in clearcuts is the delayed onset of root growth in bare-root seedlings (Johnson et al., 1984). Although container-grown seedlings do somewhat better in clearcuts (Johnson et al., 1986), the use of tree shelters (as discussed above) in combination with competition control may provide better results provided that the additional costs can be justified. Because of the problems in planting oaks in clearcuts, methods of oak enrichment planting are discussed in more detail in the following section on planting under shelterwoods. Mixtures of oak and pine can be created by combining clearcutting with enrichment planting. Where oak regeneration dynamics favours the persistence of oaks, one method for establishing such mixtures is to plant pine. The method has been shown to produce oak–pine mixtures with aesthetic and wildlife benefits as well as cost-effi-

cient timber production benefits. The method has been employed in the Southeast as a low-cost alternative to the high site preparation costs required in establishing pure pine stands. It is especially applicable to cut-over pine and mixed oak–pine stands that have not regenerated to pine, but where the natural oak regeneration potential is high. The site index for white oak on these sites typically ranges from 60 to 70 ft. One method of planting pine in oak clearcuts, developed in the Georgia Piedmont, is called the ‘fell and burn technique’ (McMinn, 1986). The method requires commercial clearcutting followed by the felling of all residual trees greater than 5 ft tall during mid-April to early June. At that time, most of the residual hardwoods are near or at full leaf expansion. The timing of this treatment is critical because: (i) the felled vegetation is needed to carry a summer burn; and (ii) cutting at the peak of leafout reduces the carbohydrate reserves of the hardwood sprouts, which in turn slows their regrowth. By cutting trees after they have leafed out, they also dry more thoroughly through ‘transpirational drying’. The felled trees are then allowed to dry until early July or later, when turkey and quail nesting is completed. The stand is then burned, which produces a high-intensity fire over a moist fuelbed. The result is additional top-kill and carbohydrate reserve depletion of the oaks and other hardwoods. Although burning does not reduce the number of hardwood sprouts, it does reduce their height growth (McMinn, 1986). Burning removes about 80% of the forest floor, but leaves about two-thirds of the root mat intact to prevent erosion. In late winter or early spring after burning, about 450 genetically improved pine seedlings per acre are hand planted. In the Piedmont, both shortleaf pine and loblolly pine are planted, but in the Coastal Plain, loblolly pine is preferred. Total costs in applying the method are about half those of conventional pine regeneration methods. The method results in high survival and rapid growth of the planted pines because


competition from hardwoods is effectively reduced, but not eliminated. The method thus produces a mixture of pine, oaks and other hardwoods. Wildlife benefits include browse for deer and cover for many other species. As stands mature, the oaks also provide mast. Other benefits include improved insect and disease resistance compared to pure pine stands. Although the prescription calls for removing all residual stems before burning, some residual trees probably could be retained to reinforce structural diversity without detracting from other objectives.

Planting under shelterwoods Oak planting can be combined with the shelterwood method to supplement the natural regeneration potential of a stand. Like natural oak reproduction, planted oaks also can benefit from a shelterwood. The period under a shelterwood allows seedlings time to become established before they compete with the surge of competition that develops after complete overstorey removal. For bare-root nursery stock, this recovery period is critical because the physiological disruptions to seedlings from lifting, handling and planting delay the initiation of root and shoot growth the first year after planting (Johnson et al., 1984; Johnson, 1988; Struve and Joly, 1992). Shelterwoods of appropriate density thus provide planted oaks with sufficient light while allowing time for them to re-establish and expand their root systems before final shelterwood removal (Dey and Parker, 1996, 1997a). In turn, this increases the proportion of planted trees that can successfully compete with other established trees after shelterwood removal. Although the use of tree shelters is also feasible under shelterwoods (as discussed above), in many if not most situations the additional expense of tree shelters is not necessary if appropriate silvicultural methods are used (as discussed below). Exceptions may include areas where deer browsing or other animal damage is severe. A practical problem in integrating planting with the shelterwood method is quanti-

285

tatively expressing and accurately predicting the expected outcome. Unlike plantings designed to create a monotype (i.e. a single-species stand), planting oaks under shelterwoods can take advantage of a stand’s natural regeneration potential. The number of planted trees (and thus planting costs) required to obtain a given future stocking thereby can be minimized. However, planting trees among naturally established trees and vegetation complicates predicting the overall regeneration outcome. Uncertainties are largely the consequence of unknowns related to the competitive struggle between planted trees and competitors in a relatively heterogeneous physical and biotic environment. The two major determinants of planting success, planted tree survival and growth, have traditionally been treated as separate and independent responses. In establishing monotypes by planting, survival by itself may provide a sufficient measure of planting success – especially when intensive preplanting site preparation eliminates competitors. In a shelterwood setting, it is more useful to define a planted tree’s competitive capacity, i.e. its expected capacity to survive and grow at a rate sufficient to attain and maintain dominance among its competitors over a specified period. Accordingly, it makes little difference whether a planted tree dies or grows so slowly that it becomes overtopped by competitors. In either case, the seedling has failed silviculturally. A suitable quantitative expression of competitive capacity accordingly would consider survival and growth simultaneously. Ideally, this quantification also would account for the effects of nursery stock quality and planting environment. The latter includes site quality, anticipated changes in competition and planned modifications of the environment before and after planting including weeding, control of shelterwood density and number of years planted trees remain under the shelterwood. Dominance probabilities (see ‘Regeneration models’, this chapter) provide a useful and convenient quantitative expression of competitive capacity. In a

286

Chapter 7

shelterwood context, dominance probability is the probability that an individual planted tree attains dominance or co-dominance a given number of years after planting or after shelterwood removal. These probabilities are likely to change through time as a result of differences in the survival and growth of planted trees and their competitors. In principle, dominance probabilities can increase, decrease, remain constant or reverse their direction through time. They therefore provide a potentially flexible quantitative expression of a planted tree’s competitive capacity. Dominance probabilities can be calculated as observed averages for a specific type of seedling in a given planting environment and number of years after planting. Alternatively, they can be estimated by various statistical methods (e.g. Weigel and Johnson, 1998a,b, 2000). The silvicultural value of dominance probabilities (P) lies not so much in the probabilities themselves, but in their reciprocals (i.e. 1/P). The reciprocals define how many seedlings are needed to obtain one competitively successful tree at a future point in time. These reciprocals are sometimes called planting factors (Johnson, 1984; Johnson et al., 1986; Weigel and Johnson, 1998a,b). The shelterwood method provides flexibility in controlling stand density (and therefore overhead light), timing the removal of the shelterwood and controlling understorey competition. Four important questions in planting oaks under shelterwoods are: • What level of overhead shading (or shelterwood density) is appropriate? • Is control of understorey vegetation necessary? • How long should the shelterwood be retained? • After final overstorey removal, is further silvicultural intervention necessary to ensure satisfactory survival and growth of planted trees? Reduction in overstorey density is usually necessary for satisfactory growth of trees planted under shelterwoods. Residual

stand densities near 50–60% stocking or 70–80% crown cover often provide the requisite conditions (Gottschalk and Marquis, 1982; Pubanz and Lorimer, 1992; Dey and Parker, 1996, 1997a). Thinning below those levels often stimulates the growth of understorey competitors at the expense of planted trees. Controlling low shade from shrubs, small trees, stump sprouts and herbaceous vegetation may be as important as controlling high shade (Horsley, 1991; Pubanz and Lorimer, 1992; Teclaw and Isebrands, 1993a,b). Control of understorey vegetation provides the dual advantages of increasing light during the shelterwood period as well as reducing competition after shelterwood removal. The length of time planted trees should remain under a shelterwood therefore partly depends on how quickly the understorey competition develops and whether additional competition control is economically justifiable. The effect of shelterwood density, understorey density and various seedling factors is demonstrated by the performance of northern red oaks planted in the Boston Mountains of northern Arkansas (Central Hardwood Region). Based on dominance probabilities 11 years after planting and 8 years after complete shelterwood removal, northern red oaks planted under low to moderate shelterwood densities (40% and 60% stocking) outperformed trees planted under higher shelterwood densities (80% stocking) (Fig. 7.12). For a given initial seedling size and type, dominance probabilities increased with time and with increasing intensity of weed control. Probabilities decreased with increasing site quality (as expressed by site index). The latter effect is related to the greater abundance and rapid growth of competitors on the better sites. There, the dominant competitors are blackgum, red maple, black cherry, flowering dogwood and other shade tolerant species. For a given planting environment, seedling characteristics that influenced planting success included initial seedling size (expressed as basal stem diameter measured 1 inch above the root collar), and


whether or not the tops of planted seedlings were cut off (clipped) 8 inches above the root collar before planting. Dominance probabilities increased with increasing initial basal stem diameter (i.e. caliper) (Fig. 7.12), and other factors being equal, probabilities were greater for clipped than for unclipped seedlings The effect of the timing of overstorey removal is demonstrated by a 19-year study on planting northern red oak in the Ozark Highlands of Missouri. In this region, opportunities for planting red oak under shelterwoods often occur on north- and east-facing slopes, where site index for black oak typically ranges from 60 to 75 ft. Although site quality there is favourable for northern red oak, the natural oak regeneration potential is often low (Fig. 3.26). The number of years planted trees remain under a shelterwood significantly affects dominance probabilities. In the Missouri study, highest probabilities occurred when shelterwoods were retained for 10 years. Shelterwoods retained for 3 or 6 years were second best, while those retained 0 years (i.e. trees planted in clearcuts) were least successful (Fig. 7.13). Dominance probabilities also increased with increasing initial shoot caliper. For a given initial caliper, 2-year-old transplants outperformed 2-yearold seedlings and clipped trees performed better than unclipped trees (Fig. 7.14A). Where oaks occur with yellow-poplar, planting oaks under shelterwoods represents a special problem. Yellow-poplar outgrows most co-occurring oaks, is long-lived and regenerates aggressively from seed and sprouts after complete or moderately heavy overstorey removal (Beck, 1991; Beck and Della-Bianca, 1981). The capacity of oaks to successfully regenerate naturally in competition with yellow-poplar is therefore severely limited. As a consequence, oaks are declining in abundance where the two species co-occur. Attempts to regenerate mixed oak and yellow-poplar stands by planting have produced a long history of planting failures, especially where site quality is high (e.g. red oak site index ≥75 ft) (Olson and Hooper, 1972; Russell, 1973; McGee and Loftis, 1986). In mixed oak and yellow-poplar stands

287

in West Virginia (oak site index 60–70 ft) that were planted to northern red oak after clearcutting, 30–50% of planted seedlings were considered competitively successful after 5 years (Wendel, 1980). Northern red oaks planted under mixed oak/yellowpoplar shelterwoods in southern Indiana (black oak site index 75 ft) responded similarly (Weigel and Johnson, 2000). In this experiment, the shelterwood was thinned to 60% stocking and retained for 3 years. Five years after shelterwood removal (8 years after planting), dominance probabilities of planted trees ranged from 0.55, depending on initial seedling caliper, shoot clipping and undercutting treatments (Fig. 7.15). However, these probabilities declined rapidly during the next 5 years as a result of suppression from overtopping yellow-poplar. Ten years after shelterwood removal, dominance probabilities for all classes and initial sizes of planted seedlings declined to

The Ecology and Silviculture of Oaks

The Southern Forest: Geography, Ecology, and Silviculture

The Oaks of Lindores

Silviculture in the Tropics

The Red List of Oaks

War for the Oaks

The Ecology of Plants

The Ecology of Seeds

The Ecology of Agroecosystems

The Ecology of Plants

The Ecology of Seeds

The ecology of Sulawesi

Ecology of the Acanthocephala

The Ecology of Freedom

The ecology of plants

Dumbarton Oaks Papers 57

The Ecology of Arabic

Dumbarton Oaks Papers 56

Dumbarton Oaks Papers 53

The Ecology of Phytoplankton (Ecology, Biodiversity and Conservation)

Dumbarton Oaks Papers, 61

Dumbarton Oaks Papers 54

Earthworms for Ecology and Profit: Earthworms and the Ecology

The geobiology and ecology of Metasequoia

Evolution and ecology : the pace of life

The Geobiology and Ecology of Metasequoia

Limnoecology: The Ecology of Lakes and Streams

Ecology and Evolution of Flowers

Basics Of Environment And Ecology

Ecology and Evolution of Flowers

The Ecology and Silviculture of Oaks

The Southern Forest: Geography, Ecology, and Silviculture

The Oaks of Lindores

Silviculture in the Tropics

The Red List of Oaks

War for the Oaks

The Ecology of Plants

The Ecology of Seeds

The Ecology of Agroecosystems

The Ecology of Plants

The Ecology of Seeds

The ecology of Sulawesi

Ecology of the Acanthocephala

The Ecology of Freedom

The ecology of plants

Dumbarton Oaks Papers 57

The Ecology of Arabic

Dumbarton Oaks Papers 56

Dumbarton Oaks Papers 53

The Ecology of Phytoplankton (Ecology, Biodiversity and Conservation)

Dumbarton Oaks Papers, 61

Dumbarton Oaks Papers 54

Earthworms for Ecology and Profit: Earthworms and the Ecology

The geobiology and ecology of Metasequoia

Evolution and ecology : the pace of life

The Geobiology and Ecology of Metasequoia

Limnoecology: The Ecology of Lakes and Streams

Ecology and Evolution of Flowers

Basics Of Environment And Ecology

Ecology and Evolution of Flowers

Recommend Documents