t*;i,-J-«'--<5ir -. ...... -..ii..^,^, - , ^.^ .; . p., — ^. -,'.-S--.%W.l- :;:'^;-:-r-:-.>-:7^i^ US/IBP SYNTHESIS SERIES | 12 AN ARCTIC ECOSYSTEM The Coastal Tundra at Barrow, Alaska Jerry Brown Philip C. Miller Larry L. Tieszen Fred L. Bunnell Dpwden, Hutchincon C& Ross, Inc. ^ North American tundra. 77?^ map also shows the position of the summer Arctic Front and sites referred to in this volume: 1-4) U.S. Tundra Biome research areas, 5) Banks Island, 6) Devon Island, 7) Axel Heiberg, 8) Mackenzie River, 9) Aleutian Islands, 10) Yukon River. Eagle Summit (3) is located in the Yukon-Tanana Uplands. The shaded portion of the map depicts areas of alpine tundra only, while the Alaskan map opposite depicts all tundra or treeless regions. (After Komdrkovd and Webber 1978.) ARCTIC OCEAN BarrowG Beaufort Sea Glaciers Generalized distribution of tundra, forests and glaciers in Alaska. The physiographic provinces of the Arctic Slope are shown. (After Viereck and Little 1975.) AN ARCTIC ECOSYSTEM US/IBP SYNTHESIS SERIES This volume is a contribution to the International Biological Program. The United States' effort was sponsored by the National Academy of Sciences through the National Committee for the IBP. The lead federal agency in providing support for IBP has been the National Science Foun- dation. Views expressed in this volume do not necessarily represent those of the National Academy of Sciences or of the National Science Foun- dation. Volume 1 MAN IN THE ANDES: A Muhidisciplinary Study of High- Altitude Quechua/Paw/ T. Baker and Michael A. Little 2 CHILE-CALIFORNIA MEDITERRANEAN SCRUB ATLAS: A Comparative JKna\ys\s/ Norman J. W. Thrower and David E. Bradbury 3 CONVERGENT EVOLUTION IN WARM DESERTS: An Examination of Strategies and Patterns in Deserts of Argentina and the United States/ Gorc^ow H. Orians and Otto T. Solbrig 4 MESQUITE: Its Biology in Two Desert Scrub Ecosystems/ B.B. Simpson 5 CONVERGENT EVOLUTION IN CHILE AND CALIFORNIA: Mediterranean Climate Ecosystems///aro/<^/4. Mooney 6 CREOSOTE BUSH: Biology and Chemistry of Larrea in New World Deserts/ r.y. Mabry, J.H. Hunziker, and D.R. DiFeo, Jr. 1 BIG BIOLOGY: The US/IBP/ PF. Frank Blair 8 ESKIMOS OF NORTHWESTERN ALASKA: A Biological Perspective/Pflw/ L. Jamison, Stephen L. Zegura, and Frederick A . Milan 9 NITROGEN IN DESERT ECOSYSTEMS/N.E. West and John Skujins 10 AEROBIOLOGY: The Ecological Systems Approach//?oZ7er/ L. Edmonds 1 1 WATER IN DESERT ECOSYSTEMS/DaAz/e/ D. Evans and John L. Thames 12 AN ARCTIC ECOSYSTEM: The Coastal Tundra at Barrow, Alaska/Zero' Brown, Philip C. Miller, Larry L. Tieszen, and Fred L. Bunnell 13 LIMNOLOGY OF TUNDRA PONDS, BARROW, ALASKA/yo/2A7 E. Hobble US/IBP SYNTHESIS SERIES 1 12 AN ARCTIC ECOSYSTEM The Coastal Tundra at Barrow, Alaska Edited by Jerry Brown U.S. Army Cold Regions Research and Engineering Laboratory Philip C, Miller San Diego State University Larry L. Tieszen August ana College Fred L. Bunnell University of British Columbia Dowden, Hutchincon CS? Ross, Inc. Stroudsburg Pennsylvania Copyright © 1 980 by The Institute of Ecology Library of Congress Catalog Card Number: 79-22901 ISBN: 0-87933-370-7 All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means — graphic, electronic, or mechanical, including photocopying, recording, taping, or information and retrieval systems — without written permission of the publisher. 82 81 80 12 3 4 5 Manufactured in the United States of America. Library of Congress Cataloging in Publication Data Main entry under title: An arctic ecosystem. (US/IBP synthesis series ; 12) Includes bibliographical references and index. 1. Tundra ecology — Alaska — Barrow. 2. Coastal ecology — Alaska — Barrow. 3. Ecology — Arctic regions. I. Brown, Jerry. II. Miller, Philip C. HI. Tieszen, Larry L. IV. Bunnell, Fred L. V. Series: US/IBP synthesis series ; 12. QHI05.A4A77 574.5'26 79-22901 ISBN 0-87933-370-7 Distributed world-wide by Academic Press, a subsidiary of Harcourt Brace Jovanovich, Publishers. FOREWORD This book is one of a series of volumes reporting results of research by U.S. scientists participating in the International Biological Program (IBP). As one of the 58 nations taking part in the IBP during the period July 1967 to June 1974, the United States organized a number of large, multidisciplinary studies pertinent to the central IBP theme of "the bio- logical basis of productivity and human welfare." These multidisciplinary studies (Integrated Research Programs), di- rected toward an understanding of the structure and function of major ecological or human systems, have been a distinctive feature of the U.S. participation in the IBP. Many of the detailed investigations that repre- sent individual contributions to the overall objectives of each Integrated Research Program have been published in the journal literature. The main purpose of this series of books is to accomplish a synthesis of the many contributions for each principal program and thus answer the larger questions pertinent to the structure and function of the major systems that have been studied. Publications Committee: US/ IBP Gabriel Lasker Robert B. Piatt Frederick E. Smith W. Frank Blair, Chairman PREFACE Tundra covers about 5.5"7o of the land surface of the earth (Rodin et al. 1975), but justification for studying it goes far beyond its areal extent. Of the world's major ecosystems, tundra has the lowest temperatures and the shortest growing seasons. Thus we may expect to find there the limits of biological accommodation and adaptation to low temperature. Largely because of the climate, which is so inhospitable for humans and so unsuitable for traditional forms of agriculture, tundra areas have never really been developed. However, increasing demands have been placed upon tundra to provide energy, minerals, food, and recreation. Often, alternative uses of tundra resources are not compatible. There are conflicting demands for wilderness, recreation areas, development of natural resources, and retention of the traditional life-styles of the indi- genous people. The research within the U.S. Tundra Biome was developed as part of both the U.S. IBP Analysis of Ecosystems program (National Acad- emy of Sciences 1974, Blair 1977), consisting of five Biomes (Grassland, Desert, Tundra, Western Coniferous Forest and Eastern Deciduous For- est), and the International Tundra Biome program, comprising some 14 other national study sites (Rosswall and Heal 1975, Wielgolaski 1975a, b, Bliss 1977, Heal and Perkins 1978, Sonesson 1980, BHss et al. 1981). It had become apparent by early 1970 that a field program centered on the coastal tundra at Barrow, Alaska, would be required to develop fully the U.S. IBP ecosystem approach. The area around Barrow had a long heritage of ecological research (Reed and Ronhovde 1971, Britton 1973, Gunn 1973), and this research contributed significantly to the plan- ning and initiation of the U.S. Tundra Biome program (Brown et al. 1970). Because of a combination of circumstances relating to the rapidly expanding oil developments in arctic Alaska and the new wave of envi- ronmental consciousness, a modest program of basic and applied re- search in tundra was initiated in 1970 (Brown 1970). The following year, a full-fledged Biome program was officially recognized, with Barrow chosen for intensive ecosystem research. Prudhoe Bay, the site of major arctic oil development, became an area for comparative coastal tundra research (Brown 1975). Two alpine sites. Eagle Summit in central Alaska and Niwot Ridge in the Colorado Front Range, provided comparative data from high- and mid-latitude alpine tundras (see map inside front cover). Vll viii Preface The 1970 field program concentrated on initiating the field design and establishing a series of field experiments and control plots. During the summers of 1971, 1972 and 1973 a vast array of field data were gath- ered from the Biome research area at Barrow. Summer 1974 was devoted to initial synthesis in a summer-long workshop that formed the basis of this volume, a companion aquatic volume (Hobbie 1980), and a volume on primary producers (Tieszen 1978a). Several broad objectives guided the research design of the U.S. Tundra Biome program from its inception: 1) to develop a predictive un- derstanding of how the tundra system operates, particularly as exempli- fied by the wet coastal tundra of northern Alaska; 2) to obtain the neces- sary data base from a variety of cold-dominated ecosystems represented in the United States so that their behavior could be modeled and simu- lated and the results compared with similar studies underway in other cir- cumpolar countries; and 3) to bring basic environmental knowledge to bear on problems of degradation, maintenance, and restoration of the temperature-sensitive and cold-dominated tundra and taiga ecosystems. The ecosystem approach and the use of ecological models were inte- grating and research tools of the U.S. IBP Biome studies. Miller et al. (1975) summarized the development of "box and arrow" representations of the tundra ecosystem. Modeling in the U.S. Tundra Biome program emphasized processes rather than the total ecosystem. This was done to maximize the interactions among field observation, hypothesis formula- tion, experimentation, and incorporation of results into working models. Such models are regarded as necessary steps leading to the eventual de- velopment of meaningful whole-ecosystem simulations. The ecological model is a research tool, not an objective. Because of this, modeling cannot be separated from field experiments, and discus- sions of the two are intertwined throughout much of the volume. Bunnell (1973) emphasized the heuristic value of models that fail to predict accu- rately or to mimic adequately the behavior of the real world. Such failure indicates that either the model structure, certain parameter values, or the basic hypotheses are incorrect, and thus contributes directly to our un- derstanding. Many hypotheses and model structures have been tried and modified or replaced as our understanding has developed. In this book the models that incorporate our understanding as of the mid-1970's are discussed and used to explore the behavior of organisms and processes under a variety of conditions. In some cases, the predictions of models have been subjected to testing, and results are presented. In other cases, the evaluation of model behavior remains a topic for future research. Our intent has been to produce an integrated discussion of a tundra ecosystem rather than a collection of independent papers on its compo- nent parts. We hope that the reader will be motivated to read the book as such. Books suffer from the constraint that they are necessarily unidi- Preface ix mensional. Although the reader may begin and end at any point, he must proceed Hnearly between those points. We view ecosystems as mukidi- mensional, with complex lines of interaction and influence running throughout. In resolving this complexity into the linear structure of this volume, we have fallen back on the relatively familiar divisions of abiotic setting, primary production processes, soil and decomposition processes, and herbivory. Within each of these subdivisions, the reader will find the common theme of the limitation of rates of biological processes by low temperature and related conditions of short growing season and the pres- ence of permafrost. The Editors ACKNOWLEDGMENTS Approximately $5.5 million was expended on this program for re- search and logistical support. Direct financial support of the Biome-wide program was derived from three major sources: the National Science Foundation, the State of Alaska and the petroleum industry through the University of Alaska. The NSF funding was under the joint sponsorship of the U.S. Arctic Research Program (Division of Polar Programs) and the U.S. International Biological Program (Ecosystem Analysis). The Army Research Office and the Department of Energy (previously AEC and ERDA) both contributed funded projects to the Program. Industry support was provided through unrestricted grants from: Atlantic Rich- field Company, Alyeska Pipeline Service Company, BP Alaska, Inc., Cities Service Company, Exxon Company, USA (Humble Oil and Refin- ing Company), Gulf Oil Corporation, Marathon Oil Company, Mobil Oil Company, Prudhoe Bay Environmental Subcommittee of the Alaska Oil and Gas Association, Shell Oil Company, Standard Oil Company of California, Standard Oil (Indiana) Foundation, Inc., and Sun Oil Com- pany. In addition to the directly funded support, parent institutions, almost without exception, provided their staff members with a variety of on-campus and other support, which is gratefully acknowledged on behalf of the entire program. In a program such as this, involving hundreds of scientists and doz- ens of institutions and support agencies, it is difficult to present a com- plete list of acknowledgments. Foremost, the Office of Naval Research through its Naval Arctic Research Laboratory (NARL) at Barrow, Alaska, provided the field and laboratory support without which the U.S. Tundra Biome program would have been impossible. Two former directors of NARL, Dr. Max C. Brewer and John Schindler, deserve par- ticular credit for facilitating the logistic support of the Tundra Biome program. Dr. Larry L. Tieszen ably served as the intensive site director at Barrow during the several summers of major activity. Alpine field sup- port at Eagle Summit and Niwot Ridge was enhanced through the Uni- versity of Alaska's Institute of Arctic Biology and the University of Col- orado's Institute of Arctic and Alpine Research (INSTAAR) at its Mountain Research Station, respectively. Administratively, the Tundra Biome Center at the University of Alaska provided vital contractual and support services. Dr. George C. West, Director of the Tundra Biome Center, Dr. Keith Van Cleve, and Mr. David Witt deserve particular credit. XI xii Acknowledgments The Cold Regions Research and Engineering Laboratory (CRREL) of the U.S. Army Corps of Engineers provided overall management, spe- cialized logistic and equipment support, and editorial services through an interagency agreement with NSF and in the spirit of Public Law 91-438, which authorized support of the U.S. IBP. Stephen L. Bowen, Technical Editor, and the CRREL editorial and photographic staffs were respon- sible for preparation of all internal Biome reports. This volume was edited and prepared through the final camera-ready copy stage by Mr. Bowen, assisted by Donna R. Murphy, Editorial Assistant, and Harold Larsen, Scientific Illustrator. Their patience, endurance and invaluable assistance over the years are sincerely acknowledged by the editors and the many Biome participants. Kathleen Salzberg, INSTAAR, provided editorial assistance on the references and prepared the subject index to this book. Sylvia Barkley verified many of the computations. A consultant committee conducted regular reviews of the program for the NSF, and its constructive advice over the life of the program is appreciated. Members included Dr. William Mayer, Chairman, Dr. Donald Wohlschlag, Dr. Hugh Raup, Dr. William Cooper and Dr. Fred- erick Sparrow. In 1970 and prior to the formation of this committee, a special field review committee headed by Dr. Philip L. Johnson made recommendations for the direction of the ensuing Biome program; its foresight has been greatly appreciated. The constructive guidance of the NSF staff is also gratefully acknowledged. NSF staff members of the Ecosystem Analysis Program included, over the five-year span. Dr. Philip L. Johnson, Dr. Charles C. Cooper, Dr. John Neuhold, Dr. Wil- liam Hazen, and Dr. Jerry Franklin. Dr. George Llano provided overall NSF program management on behalf of the Division of Polar Programs, and his valuable polar experience contributed significantly to the con- tinuity of the tundra program. This volume or its individual chapters were reviewed at several stages by the following reviewers, whose guidance and advice are greatly appreciated: Dr. Francis E. Clark, Dr. Robert S. Hoffmann, Dr. Rich- ard T. Holmes, Dr. Albert W. Johnson, Dr. Philip L. Johnson, Dr. Rob- ert S. Loomis, Dr. Gary A. Maykut, Dr. Samuel J. McNaughton, Dr. Atsuma Ohmura, Dr. William A. Reiners, Dr. Boyd R. Strain and Dr. Martin Witkamp. The Biome management and the editors of this book extend a spe- cial note of appreciation to all participants in the program for their un- derstanding and indulgence as the program evolved and during the vari- ous stages in the preparation of this volume. We especially thank those who are listed as contributors for their informal input to individual chap- ters. Dr. Stephen F. MacLean, Jr. provided considerable assistance in the planning and implementation of the many consumer-based projects, in organizing portions of this volume, and in numerous discussions con- Acknowledgments xiii cerning tundra structure and function. Along with the many authors and contributors to this volume, many other investigators, undergraduates, graduate students, technicians, and support staff participated in the Tundra Biome program. Appendix 1 contains a complete list of all Biome participants and the affiliation of the principal investigators. In addition, a considerable number of inter- national Tundra Biome members participated in various U.S. activities; these individuals are also listed in Appendix 1. As this book goes to press, we are saddened to report that the Naval Arctic Research Laboratory has closed its doors to the support of scien- tific research. This results from the lack of funds to operate the Labora- tory. Those of us who have literally spent hundreds of field seasons con- ducting field and laboratory research in the Barrow environs hope that the closure of this scientific facility on the shores of the Arctic Ocean is a temporary action and that future generations of scientists will enjoy the logistic opportunities we have been privileged to experience over the past several decades. JERRY BROWN Former Director, U.S. Tundra Biome USA CRREL Hanover, New Hampshire CONTENTS Foreword v Preface vii Acknowledgments xi List of Authors xix List of Contributors xxiii The Coastal Tundra at Barrow J. Brown, K. R. Everett, P. J. Webber, S. F. Maclean, Jr., and D. F. Murray The Arctic Coastal Plain: A Geographic Perspective, 1 Barrow Research Area, 10 Terrain Subdivisions and Formation, 16 Soils: Description and Distribution, 21 Vegetation, 25 2: Climate, Snow Cover, Microclimate, and Hydrology 30 5. L. Dingman, R. G. Barry, G. Weller, C. Benson, E. F. LeDrew, and C. W. Goodwin Introduction, 30 Climate, 30 Snow Cover, 37 Microclimate, 44 Hydrology, 51 Summary, 64 Biophysical Processes and Primary Production 66 P. C. Miller, P. J. Webber, W. C. Oechel, and L. L. Tieszen Introduction, 66 Primary Production in the Barrow Wet Meadow Tundra, 67 Vertical Distribution of Biomass and Canopy Structure, 74 Influence of the Canopy on the Physical Environment, 78 Control of Evapotranspiration by Plants and Plant- Water Rela- tions, 91 Simulation of Plant-Water Relations, 97 Summary, 100 XV xvi Contents 4: Photosynthesis 102 L. L. Tieszen, P. C. Miller, and W. C. Oechel Introduction, 102 Intrinsic Factors Affecting Carbon Dioxide Exchange, 103 Extrinsic Factors and the Rate of Photosynthesis, 110 Simulation Analysis of Photo- synthesis and Vascular Canopy Interactions, 120 Summary, 137 5: Control of Tundra Plant Allocation Patterns and Growth 140 F. S. Chapin III, L. L. Tieszen, M. C. Lewis, P. C. Miller, and B. H. McCown Introduction, 140 Growth Patterns of Tundra Gramin- oids, 141 Allocation of Carbon Compounds, 148 Carbon Cost of Plant Growth, 158 Nutrient Absorp- tion, 165 Nutrient Allocation, 169 Reproductive Allocation and Population Structure, 175 Effect of Grazing on Allocation and Population Structure, 177 Variability in Growth and Allocation Patterns, 181 Summary, 184 6: The Vegetation: Pattern and Succession 186 P. J. Webber, P. C. Miller, F. S. Chapin III, and B. H. McCown Introduction, 186 Topographic Variation and Vegeta- tion Patterns in the Coastal Tundra, 187 Growth Forms and Environmental Control, 194 Functional Relationships of Growth Forms and Environmental Gra- dients, 200 Plant Succession and Response to Perturba- tions, 206 Summary, 217 7: The Soils and Their Nutrients 219 P. L. Gersper, V. Alexander, S. A. Bark ley, R. J. Barsdate, and P. S. Flint Introduction, 219 Soil Physical Properties and Nutri- ents, 220 Inputs and Outflows of Nitrogen and Phos- phorus, 234 Transformation and Transport of Nitrogen and Phosphorus Within the Soil, 245 Summary, 253 Contents xvii 8: The Microflora: Composition, Biomass, and Environ- mental Relations 255 F. L. Bunnell, O. K. Miller, P. W. Flanagan, and R. E. Benoit Introduction, 255 Taxonomic Structure: Its General Character and Heterogeneity, 257 Microflora Biomass: Its Distribution in the Environment and Changes Through Time, 268 Environmental Controls on Microfloral Biomass, 281 Summary, 288 9: Microflora Activities and Decomposition 291 P. W. Flanagan and F. L. Bunnell Introduction, 291 Potential of the Microflora to Utilize Substrates, 294 Abiotic Variables and Microflora Ac- tivities, 298 Decomposer Activities and Decomposition, 309 Summary, 333 10: The Herbivore-Based Trophic System 335 G. O. Batzli, R. G. White, S. F. Maclean, Jr., F. A. Pitelka, and B. D. Collier Introduction, 335 Herbivory at Barrow — Lemmings, 337 Predation on Lemmings, 360 Factors Influencing Lem- ming Populations, 371 Herbivory at Prudhoe Bay — Caribou, 384 Comparison of Grazing Systems, 399 Summary, 409 11: The Detritus-Based Trophic System 411 5. F. Mac Lean, Jr. Introduction, 411 Abundance and Biomass of Soil In- vertebrates, 412 Life Cycles of Tundra Soil Inverte- brates, 424 Energetics of Soil Invertebrates, 432 Energy Structure of the Detritus-Based Trophic System, 436 The Role of Soil Invertebrates in Nutrient Cycles, 441 Abundance, Productivity, and Energetics of Avian Insectivores, 444 Summary, 456 xviii Contents 12: Carbon and Nutrient Budgets and Their Control in Coastal Tundra 458 F. S. Chapin III, P. C. Miller, W. D. Billings, and P. I. Coyne Introduction, 458 Standing Crops, 458 Average An- nual Fluxes, 465 Long-Term Changes in Coastal Tun- dra, 478 Summary and Conclusions, 481 References 483 Appendix 1: U.S. IBP Tundra Biome Projects, Personnel, Site Locations, 1970-1974 545 Appendix 2: Location of Principal Biome Plots 559 Subject Index 561 LIST OF AUTHORS Vera Alexander Institute of Marine Science, University of Alaska Fairbanks, Alaska 99701 Sylvia A. Bark ley t Biology Department, San Diego State University San Diego, California 92182 Roger G. Barry Institute of Arctic and Alpine Research, University of Colorado Boulder, Colorado 80309 Robert J. Barsdate Institute of Marine Sciences, University of Alaska Fairbanks, Alaska 99701 George O. BatzH Department of Ecology, Ethology, and Evolution University of Illinois, Urbana, Illinois 61801 Robert E. Benoit Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 Carl Benson Geophysical Institute, University of Alaska Fairbanks, Alaska 99701 W. D wight Billings Department of Botany, Duke University Durham, North Carolina 27706 Jerry Brown U.S. Army Cold Regions Research and Engineering Laboratory Hanover, New Hampshire 03755 t Affiliation at time of participation. XIX XX List of Authors Fred L. Bunnell Faculty of Forestry, University of British Columbia Vancouver, British Columbia, Canada V6T 1W5 F. Stuart Chapin III Institute of Arctic Biology, University of Alaska Fairbanks, Alaska 99701 Boyd D. Collier Biology Department, San Diego State University San Diego, California 92182 Patrick I. Coyne U.S. Department of Agriculture, Southern Plains Range Research Station, Woodward, Oklahoma 78801 S. Lawrence Ding man Institute of Natural and Environmental Research University of New Hampshire, Durham, New Hampshire 03824 K. R. Everett Institute of Polar Studies, Ohio State University Columbus, Ohio 43210 Patrick W. Flanagan Institute of Arctic Biology, University of Alaska Fairbanks, Alaska 99701 Philip S. Flinty Department of Soils and Plant Nutrition, University of Cahfornia Berkeley, California 94720 Paul L. Gersper Department of Soils and Plant Nutrition, University of California Berkeley, California 94720 Cecil W. Goodwin Department of Geography, Pennsylvania State University State College, Pennsylvania 16802 Ellsworth F. LeDrew Department of Geography, University of Waterloo Waterloo, Ontario, Canada N2L 3G1 Martin C. Lewis Department of Biology, York University Downsview, Ontario, Canada M3J 1P3 List of Authors xxi Stephen F. MacLean, Jr. Institute of Arctic Biology, University of Alaska Fairbanks, Alaska 99701 Brent H. McCown Department of Horticulture, University of Wisconsin Madison, Wisconsin 53706 Orson K. Miller Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 Philip C. Miller Biology Department, San Diego State University San Diego, California 92182 David F. Murray Institute of Arctic Biology and the Museum University of Alaska, Fairbanks, Alaska 99701 Walter C. Oechel Biology Department, San Diego State University San Diego, California 92182 Frank A. Pitelka Museum of Vertebrate Zoology, University of California Berkeley, California 94720 Larry L. Tieszen Department of Biology, Augustana College Sioux Falls, South Dakota 57102 Patrick J. Webber Institute of Arctic and Alpine Research, University of Colorado Boulder, Colorado 80309 Gunter Weller Geophysical Institute, University of Alaska Fairbanks, Alaska 99701 Robert G. White Institute of Arctic Biology, University of Alaska Fairbanks, Alaska 99701 LIST OF CONTRIBUTORS Rodney J. Ark ley Department of Soils and Plant Nutrition, University of California Berkeley, California 94720 Mary A I less io Leek Department of Biology, Rider College Lawrenceville, New Jersey 08648 Edwin M. Banks Department of Ecology, Ethology and Evolution University of Illinois, Urbana, Illinois 61801 Dirk Bare! Research Institute for Nature Management Arheim, The Netherlands Pille Bunnell Institute of Animal Resource Ecology, University of British Columbia, Vancouver, British Columbia, Canada V6T 1W5 Martyn Caldwell Department of Range Science, Utah State University Logan, Utah 84321 Roy E. Cameron Argonne National Laboratory, Environmental Statement Project Division, Argonne, Illinois 60439 G. Keith Douce Department of Entomology, University of Georgia Athens, Georgia 30603 Douglas A. Johnson U.S. Department of Agriculture, Crops Research Laboratory Utah State University, Logan, Utah 84322 Vera Kom6rkov6 Institute of Arctic and Alpine Research, University of Colorado Boulder, Colorado 80309 xxni xxiv List of Contributors Gary A. Laursen Office of Naval Research, Arlington, Virginia 22217 Jay D. McKendrick Agricultural Experiment Station, University of Alaska Palmer, Alaska 99645 Herbert Melchior Alaska Department of Fish and Game, Fairbanks, Alaska 99701 Barbara M. Murray Institute of Arctic Biology, University of Alaska Fairbanks, Alaska 99701 David Norton Outer Continental Shelf Program, Geophysical Institute University of Alaska, Fairbanks, Alaska 99701 Ronald G. Osborn t Department of Biology, San Diego State University San Diego, California 92182 Samuel Outcalt Department of Geography, University of Michigan Ann Arbor, Michigan 48104 James R. Rastorfer Department of Biological Sciences, Chicago State University Chicago, Illinois 60628 Emanuel D. Rudolph Department of Botany and Institute of Polar Studies Ohio State University, Columbus, Ohio 43210 Uriel Safrielt Department of Zoology, Hebrew University, Jerusalem, Israel William C. Steere New York Botanical Garden, New York, New York 10460 Wayne St oner '\ Biology Department, San Diego State University San Diego, Cahfornia 92182 t Affiliation at time of participation. List of Contributors xxv Albert Ulrich Department of Soils and Plant Nutrition, University of California Berkeley, California 94720 Donald A . Walker Institute of Arctic and Alpine Research, University of Colorado Boulder, Colorado 80309 AN ARCTIC ECOSYSTEM The Coastal Tundra at Barrow J. Brown, K. R. Everett, P. J. Webber, S. F. Maclean, Jr., and D. F. Murray THE ARCTIC COASTAL PLAIN: A GEOGRAPHIC PERSPECTIVE The word tundra is broadly used to refer to the landscapes that are found above the altitudinal or latitudinal treeline. The classification of tundra has been reviewed by Barry and Ives (1974) and Murray (1978). In Alaska, lowland tundra covers large portions of the Aleutian Islands and the delta of the Yukon and Kuskokwim Rivers in the southwest, the Sew- ard Peninsula in the west, and the Arctic Slope (see map inside front cover). Alpine tundra is found in all the mountain ranges and at eleva- tions above about 1000 meters in upland terrain such as the Yukon- Tanana Uplands. The Arctic Slope, that part of northern Alaska that drains to the Arctic Ocean, covers 200,000 km^ an area the size of the state of Nebras- ka. It consists of three major physiographic provinces: the Brooks Range, the Arctic Foothills, and the Arctic Coastal Plain (Figure 1-1). These provinces differ in topography, geology, climate and history, and consequently in fauna and flora. Permafrost underlies all land surfaces at depths up to approximately 600 m. The Brooks Range, with peaks as high as 2700 m, has cirque glaciers in its central and eastern sections. Variations in slope and topography lead to large differences in microclimate and in soil properties, and thus there are diverse habitats for plants and animals. Only the floodplains of the larger river drainages have extensive stands of a single vegetation community, usually shrub thicket. The valley bottoms contain sedge meadows and well-developed shrub tundra that is dominated by willow {Salix lanata, S. pulchra, S. glauca and S. alaxensis) and dwarf birch {Betula exilis). The slopes have dry meadow or heath communities dominated by Dryas octopetala. Above about 1800 m vascular plants are limited to protected sites, and lichen cover is discontinuous. Such areas are analogous to the polar deserts of high latitudes, and might be called alpine deserts. I ?3 s: Qj ^ij 5 ^ ■^ 5 ^ ^ ^ ■o 0 -s: On cu ^ [^ c3 *^ ^^*— e 00 ^ o * cs o ■^^ O) t\ <4j 53 0. % E ^ (V) O — tf)C\J CO S o V. '■**» s -**- •^« ^ ?3 •^ !<. o X O. o~ 1 -•^ 5j ^ ■^- o s; 0 ^ ?3 U) ^ <4; flo-i \j s: s: ^ 3 5^ S § f*- -s: • o 1 (=0 -5: to S ^ » 53 s: ^ < CO o^ CQ < T3 C 03 O O J= "iJ 3 O < ^ " 'S D. H < on o o 03 00 C ■•-' UD O = 2 c CO W ° a, u O ^ >. T3 W 3 CQ O CQ ON ^ ■rj- o 1 ^1 00 'T' o "^ r^ O >/-| OS O 5 T o ON 1 O •^ -I ^ 00 r- ON ^^ u k. (/5 u lU 0 iA CO (55 0 O. T3 00 C r~ ns On m «N O 1 ^" ■^ -I 1 2 S§ 2§ ON _J C o c c 3 o 00 ■ + o c c 00 O 2 i^ OO— 1 1 d SO OS _l •^ -1 Tt"^ lo^ vo°° (N*^ r^*^ (N"^ —OO j^vo ;;^^ 'i^'* Jo Id id id d ^J oj oj r-J o '^ = J 1 - 0^ m ■ OO . c U ^ U ^ ^Ui>^ c CO ^. w — : -J X) 3 D. C 3 CO ioS u — < X) 3 D. C 3 _co o? o O x> £ E H = U- O ^ U d-§ ^ ON c — CO 5 CO D. C D. O E o X u c o 1^ ■a o X D. 1) XJ 3 C _c t/1 OS u '** '5 C (U O ^ u. ^ i CO . o k- k- JO c •o c G CO u^ u— u ^ O Q. O E c CO t/5 o ir. ■o C c o C/5 8 J. Brown et al. Prudhoe Bay — have only a limited subset of its total flora. This is a re- flection of the particularly severe summer climatic conditions that exist near the coast, but moderate inland across the plain to the south. This zone of littoral tundra (Figure 1-1) is characterized by low species diver- sity, dominance of grasses and sedges (Cantlon 1961), rarity of tussock tundra, and the absence of erect shrubs. The limit of littoral tundra fol- lows roughly the 7°C July mean temperature isotherm. The floras of the coastal tundras at Barrow and at Prudhoe Bay are strikingly dissimilar, with Sc^renson Coefficients of Similarity (Sdrenson 1948) that are much lower than those describing the comparisons be- tween provinces (Table 1-2). The Barrow peninsula is dominated by wet, acid soils. The Prudhoe Bay region, in contrast, is influenced by two large rivers, the Sagavanirktok and Kuparuk, that originate in the moun- tains and contribute carbonates to the soils. However, even in this region, acid soils occur beyond the influence of the rivers. Differences in parent materials and topography have a marked ef- fect on the vegetation of the two regions. The lichen genera Cladina and Cladonia, which are associated with acid soils, are represented by 19 taxa in the Barrow region, but only 4 at Prudhoe Bay. The species Collema tunaeforme, Evernia perfragilis and Fulgensia bracteata, which are char- acteristic of calcareous substrates, are present at Prudhoe Bay but not at Barrow. The absence of sphagnum mosses and the scarcity of polytrica- ceous mosses around Prudhoe Bay are examples of probable pH control of distribution in the bryoflora (Steere 1978a). The effect of pH on vascular plants is noticeable in the distribution of Carex species; the cal- ciphiles C. atrofusca, C. membranacea and C. misandra are common in the vicinity of Prudhoe Bay but are absent from the Barrow area (Mur- ray 1978, Walker and Webber 1979). A rich arctic-alpine floristic ele- ment is found on the numerous exposed and well-drained habitats that form on the coarse-textured soils of river banks, gravel bars, dunes and pingos close to Prudhoe Bay. The vegetation in this area shows a greater affinity with that of more southerly tundra than does the vegetation of the Barrow peninsula. The terrestrial vertebrate fauna of northern Alaska is composed en- tirely of homeothermic birds and mammals and lacks the heterothermic reptiles and amphibians which extend well into the Asian Arctic. Pitelka (1974) listed 189 species of birds for the Arctic Slope, 76 of which are considered to be regular breeders (Table 1-2). Because of the mobility of birds, there are many records of accidental or casual visitors to the Arctic Slope. The number of species is very similar in the three physiographic provinces: 45 species are found in both the Coastal Plain and Foothills, and 50 in the Brooks Range. However, the composition of the bird fauna changes considerably from the coast to the mountains. Thirty species of waterfowl and shorebirds and three species of passerine birds occur on The Coastal Tundra at Barrow 9 the Coastal Plain, while in the Brooks Range passerines account for 21 species and waterfowl and shorebirds for 18 species (Kessel and Cade 1958, Pitelka 1974). Only 13 bird species breed regularly in all three pro- vinces. Thirty-five species, or 46*yo of the regular breeders, are limited in their breeding distribution to the Arctic Slope. Thus the avifauna exhib- its a much higher level of endemism than does any other animal or plant group. The breeding bird faunas of the Barrow and Prudhoe Bay regions are more similar than their floras. The Arctic Slope is well known for its wildlife resources, which in- clude dall sheep (Ovis d. dalli), wolves (Canis lupus), grizzly and polar bears {Ursus arctos and U. maritimus), wolverines {Gulo gu/o), moose {Alces alces) and caribou (Rangifer tarandus), as well as golden eagles {Aquila chrysaetas) and peregrine falcons {Falco peregrinus). Musk oxen {Ovibos moschatus) were a part of the fauna until the mid- to late 1800s, when they were eliminated by man. They have been successfully reintro- duced to the Arctic Slope in the Arctic National Wildlife Range and near Cape Thompson (Figure 1-1), and can once again be considered resident mammals of the Coastal Plain and Foothills. Bands of caribou from two major herds roam the Arctic Slope, and have always been important in the economy of the Eskimo. The terrestrial mammal fauna of the Arctic Slope is limited to about 24 species. The number of species present increases from the Coastal Plain to the Brooks Range. All of the species of the Coastal Plain except polar bears live in the Foothills as well, and all but polar bears and musk oxen live in the mountains of the Brooks Range. Faunal overlap between provinces is very high, as is overlap between the fauna of the Arctic Slope and more southerly locations. Only 5 of the 24 terrestrial mammals are endemic to arctic tundra. Microtine rodents are a substantial part of the mammal fauna. Whereas in the Canadian Arctic Archipelago only the collared lemming (Dicrostonyx groenlandicus) is found, on the Coastal Plain at Barrow there are two species, the collared and the brown lemmings (Lemmus si- bericus, = trimucronatus). At Prudhoe Bay, and inland from the coast, these two species occur together with the tundra vole {Microtus oecon- omus). And these microtines are joined by the red-backed mouse {Cleth- rionomys rutilus) in the foothill tundra and by the singing vole {Microtus miurus) in rocky habitats of the southern Foothills and the Brooks Range. The adjacent Arctic Ocean is rich in marine mammals such as ringed and bearded seals (Phoca hispida and Erignathus barbatus), walrus {Odobenus rosmarus), and bowhead and beluga whales (Balaena mysti- cetus and Delphinapterus leucas). The marine mammals, probably more than any other resource, were responsible for the development of the rich culture of the people living along the arctic coast of Alaska. 10 J. Brown et al. The invertebrate fauna of the Arctic Slope is poorly known. However, the recently built road from interior Alaska to the Prudhoe Bay oilfield has provided access to collecting sites along a latitudinal transect, allowing study of the distribution of invertebrates. Soil mites (Acari) and springtails (Collembola) are among the most diverse of the invertebrates; 127 mite species and 115 Collembola species were recorded along the transect (Table 1-2). Since these data come from a limited number of samples, they do not represent the total fauna of the Arctic Slope. The number of species recorded from the three physiographic pro- vinces is similar. The most Collembola species are found in the Foothills and Coastal Plain, while the Acari are most diverse in the Brooks Range and the Foothills. Overlap between provinces is high. Recently, MacLean et al. (1978) showed that the mite and Collembola species from a coastal tundra site in northern Chukotka, in northeast Asia, overlapped consid- erably with the fauna of the Arctic Slope. Fifty-nine of ninety mite spe- cies and 50 of 79 Collembola species found in the Chukotkan tundra have also been found in Alaska. In contrast to the distribution of soil microarthropods, many other invertebrate groups increase in number of species from the Coastal Plain to the Brooks Range. For example, at Barrow there are only small numbers of a few species of herbivorous insects, including only a single species of the plant-sucking Homoptera, which is an important group worldwide. However, the number of species of herbivorous insects in- creases almost four-fold from the Coastal Plain to the Foothills along the road transect, then declines markedly from the Foothills to the Brooks Range. The distribution parallels that of the vascular plants on which the insects feed. However, the distribution of host plants is clearly only one factor in herbivore distribution, since these plant species are commonly found farther north than their characteristic herbivores. BARROW RESEARCH AREA Location The Barrow peninsula is situated at the northern extremity of the Coastal Plain (71°18 'N, 156°40 'W). It is a triangular-shaped land mass bounded by the Chukchi Sea on the west and the Beaufort Sea and Elson Lagoon on the east (Figure 1-3). The earliest known site of human habi- tation, Birnirk, was submerged by an encroaching sea some 1200 to 1500 years ago (Ford 1959). The present village of Barrow (population 2700) is the largest Eskimo settlement in the State of Alaska. Since 1920 the National Weather Service has operated a first-order weather station at Barrow. About 6 km north of the village there are The Coastal Tundra at Barrow 11 -71° 20' 156 50' I 56° 40' 156 30' FIGURE 1-3. Physiographic and historical features of the Barrow penin- sula. A) North Salt Lagoon, B) Middle Salt Lagoon, C) South Salt Lagoon, D) Imikpuk Lake, 1) Drainage channel for Middle Salt Lagoon. The shaded area around the lagoons and Imikpuk Lake is now tundra, but was previously a large embayment. The box outlines the Biome area shown in Figures 1-7, l-ll and 1-12. Point Barrow is 5.8 km northeast of Birnirk. X-Y indicates the approximate location of the cross section shown in Figure 1-6. several government-operated facilities, including the Naval Arctic Re- search Laboratory (NARL), which has supported research on tundra ecology since 1947 (Reed and Ronhovde 1971, Britton 1973, Gunn 1973). In the 1940s and early 1950s Barrow was the main supply base for the ex- ploration of Naval Petroleum Reserve No. 4, redesignated National Pe- troleum Reserve, Alaska, in 1976. It also was a supply station during the construction of distant early warning sites (the DEW line) in the 1950s. The coastal tundra at Barrow has low relief and is dominated by a pattern of ice-wedge polygons, shallow oriented lakes, drained lake basins and small ponds. Elevations range from sea level to 5 m along the northern shores of Elson Lagoon and rise to greater than 10 m south- westward across the peninsula. North of 71°15 ', approximately 65% of the surface is covered by polygonal ground (Sellmann et al. 1972), and half of this consists of high-centered or low, flat-centered polygons. The 12 J. Brown et al. FIGURE 1-4. Aerial view looking north across the U.S. Tundra Biome research area. The ice-covered Arctic Ocean is in the background. The Naval Arctic Research Laboratory camp complex is in the upper right corner. The ice-covered water body is Middle Salt Lagoon. Polygonal terrain is visible in the foreground. (CRREL photograph.) remainder of the landscape contains recently drained lake basins, gently sloping terrain and lakes. In 1970 an area 3 km from the sea near NARL was selected for the main U.S. International Biological Program Tundra Biome research ef- fort (Figure 1-4). The area contains a sequence of drained lake surfaces (Figure 1-5). A small entrenched stream, Footprint Creek, flows across the northern portion and its marshy feeders, together with polygon troughs, drain the western and southern portion. The drainage area north of the creek consists of gentle, hummocky slopes and high- centered polygons with deepened troughs along with non-patterned mea- dows and mixed high- and low-centered polygons. Three primary terrestrial sites and one aquatic site were established (Figure 1-5): Site 1, immediately north of Footprint Creek, was used for a series of experimental simulations of human-induced impacts: heated soils, oil spills, physical disruption and greenhouse effects. I56°44 The Coastal Tundra at Barrow 56°40' I56°38' 13 Limit of Mid-Wisconsi Transgression \ 1 J I X f4^ llFootprint\ \ '^^^^' I Lake V nonN. I 1 \(Drained)i A , L -TI'IS --71° 18 --1\°\1 FIGURE 1-5. U.S. Tundra Biome intensive re- search area (S-1, S-2, S-4, S-7) and associated landscape features. 1) oldest basin, 2-5) inter- mediate aged basins, 6) youngest basin. Foot- print Lake was drained in 1950 through a bull- dozed trench. Site 2, immediately south of Footprint Creek, was chosen for its relatively uniform wet meadow. Control plots and experiments to simu- late natural perturbations, including changes in grazing and nutrient re- gime, were established. Site 4, south of site 2, was selected for its microtopographic diversity and included a number of plots that contained representative polygonal terrain. These three sites constitute the "Biome research area" referred to in this book. Site 7, west of Footprint Creek, contained well-developed, low- centered polygons filled with water that were used for aquatic studies (Hobbie 1980). Locations of plots and sample areas within these and related sites are shown on the map in Appendix 2. 14 J. Brown et al. Y Biome Sites X - 10]- 14 000 Footprint 8500-11,000(18,500) u.„^.,,. n'O !.\ /Ice Wedge Creek 25,300 )^=?Ss ^^^ce wedge cre« 25,300^>?^ Sed.m.n,, 2,650-4,570 9 -10 -20 --I0 HUffMrn^nirm: FIGURE 1-6. Idealized geologic cross section across the Barrow penin- sula showing location and age of radiocarbon-dated organic materials (see Figure 1-3 for the approximate location of the section). The radio- carbon dates in years BP are from Brown (1965) except 18,500 from Everett (unpubl.) and 12,160 from Lewellen (1972). Geologic History The general course of events over the past 25,000 years has been re- constructed on the basis of local and regional stratigraphic and geo- morphic data and correlations (Lewellen 1972, Sellmann and Brown 1973). Figure 1-6, an idealized section across the Barrow peninsula that incorporates radiocarbon dates and stratigraphic information, illustrates the major geologic units and surface features. The materials shown in the upper part of the cross section were deposited or reworked by an inva- sion of the sea that extended inland to about the current 8-m elevation. This elevation is not precise, since the surface relief changed as the sedi- ment refroze, producing an increase in volume due to the formation of ground ice. During this transgression, which occurred approximately 25,000 to 35,000 years ago (mid-Wisconsinan time), the surface of the ex- isting permafrost would have been lowered, but permafrost probably did not disappear entirely. Subsea permafrost is currently found to depths of over 100 m offshore from the present coastline (Lachenbruch and Mar- shall 1977, Lewellen 1977). With the retreat of the sea, a gravelly beach ridge-shoal complex built up north and east of the Biome research area. This ridge complex currently reaches elevations of 6 to 8 m. North of the ridge, but landward of the present-day active beach, an embayment formed which persists in fragmented form as the lakes and sloughs from Elson Lagoon to Barrow Village. Until 1945 these lakes and sloughs were connected during peri- ods of high water at approximately the 2.5-m elevation. At that time. Middle Salt Lagoon was artificially drained to prevent flooding around NARL and the level of the slough was decreased to nearly sea level (Fig- ure 1-3). The Coastal Tundra at Barrow 15 As the area occupied by the Biome research sites was exposed by a late glacial regression of the sea, geomorphic processes related to extreme cold climate again developed. The sediments that had thawed beneath the shallow waters of the sea refroze to form a thick and continuous per- mafrost section. Presumably, ice wedge formation was active under the arctic climate that existed at that time, and has continued to the present. Evidence from pollen recovered from a buried ice wedge and radiocar- bon dating of organic matter immediately west of the Biome research area indicates that by at least 14,000 BP a tundra as cold as at present, but somewhat drier, existed (Colinvaux 1964, Brown 1965). Fecal pellets of microtine rodents recovered in the same ice wedge sample indicate the presence of these small mammals at that time. Since ice wedges were ac- tively growing 14,000 years ago, it is reasonable to assume that the thaw lake cycle (Britton 1957) as we know it today was also an active geo- morphic process then. The detailed historical reconstruction of the Biome research area that follows is based upon this assumption. In the Footprint Creek drainage a lacustrine peat immediately over- lying the marine sediment at 2.2 m elevation has been radiocarbon dated at 12,160 ±200 years (Figure 1-6). This date provides a maximum age for the thaw lake. And since this peat directly overlies marine sediment, it probably also dates one of the earliest thaw lakes in the area. Aerial photographic interpretation suggests that the entire area was originally covered by one large lake (Figure 1-5). The shoreHne of this lake abuts the surrounding uplands at an elevation of 5.5 m on both the east and west sides. The topographic high point on the west is known to contain large quantities of segregated ice and ice wedges (Brown 1965) and is probably a remnant of the primary land surface. The net effect of this early lake was to thaw the ice-rich permafrost to a depth of 3 m and re- work materials of the type still found to the west. In the process most of the sea salts were removed from the sediments that form the parent mate- rial of the present soils. The initial lake probably drained as headward erosion of the Middle Salt Lagoon slough cut through a shallow pass in the elevated beach ridge-shoal complex. With this draining or lowering of the lake level, the newly exposed lake sediments refroze, tundra vegetation developed, and a new cycle of ice wedge formation began. In time, smaller lakes developed in this large lake basin. Each lake drained as headward erosion of the small tundra-covered stream chan- nels intercepted the borders of the lake basins. The youngest and smallest lake apparently formed the basin within which the ponds studied in the Biome aquatic program are situated (site 7; Hobbie 1980). The present- day polygonal ground is not only the product of current ice wedge for- mation, but probably represents previous cycles of ice wedge growth, both on land and under shallow lakes or ponds. The Biome research area 16 J. Brown et al. is slightly higher than the pond-covered areas to the east and west, sug- gesting that it is older. It is unlikely that more than one major lake has occupied the area of the terrestrial research sites, but the adjacent aquatic research site has gone through at least three lake cycles. The activities of man have led to significant changes in the Biome re- search area since the mid-1940s. The tracks of off-road vehicles have produced thermal disturbance and subsequent formation of small ponds at ice wedge intersections, shifts in vegetation composition and, in some cases, severe erosion. More dramatic is the alteration in drainage caused by the lowering of Middle Salt Lagoon to sea level and the complete draining of Footprint and Dry Lakes in 1950. These actions created a new erosional base level in the basin (Figure 1-5). Prior to this lowering, the water level in Middle Salt Lagoon stood at approximately 2.5 m ele- vation. Since then there has been massive and rapid headward erosion as Footprint Creek readjusts to the new base level conditions (Lewellen 1972). After the water level was lowered, the floor of the previously flooded valley was exposed and Footprint Creek entrenched further to form the present marginal terrace. The gravel road on the east side of the Biome research area has caused some ponding and mehing of ice wedges. TERRAIN SUBDIVISIONS AND FORMATION Ahhough tundra might appear to be a featureless plain, it actually possesses a considerable variety of landforms, both on meso- and micro- scales, and is an extremely dynamic landscape (Britton 1957). The Biome research area and areas immediately adjacent to it consist of several ma- jor terrain units, each composed of characteristic landforms and associ- ated microtopographic units (Figure 1-7). Adjacent to Footprint Creek are alluvial terraces, floodplains, and steep- and gently sloping stream banks. Immediately to the south are the weakly developed polygons that compose much of site 2 (Figure 1-4). Site 4, farther to the south, consists of a highly polygonized landscape containing both high-centered and well-developed low-centered polygons (Figure 1-8). Polygons give rise to several microtopographic units — rims of low- centered polygons, tops of high-centered polygons, basins (centers) of low-centered polygons, and polygon troughs — and to a diverse range of soil types, vegetation, and habitats. These recur over short distances, commonly on the order of several meters (see Figures 1-8 and 1-9). Since the troughs are interconnected they serve as pathways for the movement of water, nutrients and plant litter, especially during snowmek. In winter they are extensively used by lemmings under the deeper snow and low density depth hoar (Chapter 2). The basins of low-centered polygons are relatively poorly drained and many are areas of continuing organic The Coastal Tundra at Barrow 17 FIGURE 1-7. Landscape map of the Tundra Biome research area. The area mapped is the same as that in Figures 1-1 1 and 1-12. The black areas are gravel roads and pads. The dashed lines indicate off-road vehicle trails present in 1971. Heavy solid lines outline the major soil complexes, associations, and units (Figure 1-11). The rectangle is commonly referred to as site 4. Details of the hatched rectangle are shown in Figure 1-8. 1) High-centered polygons, relief >1 m. 2) Mixed high-centered and low- centered polygons, relief 4:0.5 m, frost boils common. 3) High-centered polygons, broad trough, relief <0. 5 m. 4) Low-centered polygons, relief <0.5 m. 5) Large, low-centered polygons, many orthogonal, relief 4:0.5 m. 6) Large, high-centered poly- gons, narrow troughs, relief 0.1 to 0.5 m. 7) Polygonal pattern, relief <0.1 m. 8) Sloping areas (2 to 6%) mar- ginal to streams, frost boils and discontinuous polygonal cracks. 9) Vegetation-covered water- way, discontinuous polygonal cracks. A) Alluvium. d) Disturbed area. 18 J. Brown et al. □ rims and low relief, high-centered polygons ^M polygon centers m polygon troughs meadows S frost boils V*,'.*. >^ 1^ "5 s-JT^ ^^^%<^. A' --C ^ '<#! [~] Carex aqualilis-Duponlia fisheri (V) [Tj c^^^^'o'^of^yif ^ (HI) I 1 Dupontia fisheri-Calliergon (V) [~] Catliergon (VI) f~| Corex aquatilis (IV) humus-Carex fl [ j bare soil-soil lichen (II) FIGURE 1-8. Vertical photograph (center), soils (top), and vegetation (bottom) of a portion of the site 4 grid. (After Walker 1977). The complex- ity of the microlopography decreases from right to left. Mixed high- and low-centered poly- gons are on the extreme right. The center portion is mostly well-developed low-centered polygons. The dark colored centers may contain standing water following snowmelt and during wet sumtners. The rims of polygons are outlined in lighter tones. Troughs are linear elements, dark where they are very moist or contain standing water. The left third consists of polygons with little or no microrelief contrast. See Figures 1-7, 1-11, and 1-12 for the re- lationship of the grid location to the terrain, soils, and vegetation of the entire research area. See Figure 1-9 for features on transect A-B. The Coastal Tundra at Barrow 19 Vegetation Type I . Ill Land Rim Form VI VI Basin Rim Trougti Rim Basin Rim Trough Rim 0 A I deciduous shrub 15 20 Distance along Section (meters) 25 30 single ^ rosetted », erect mot , coespitose | ...,>,,. ^ bryophyte •I- dicotyledon Y dicotyledon dicotyledon " monocotyledon I monocotyledon B lichen BASIN RIM TROUGH 0 f m\ 5 - p 10 - o Depth - 20 - g ■ 25 -Alb 1 30 - ? Oi Oe/Oi Oe2 0e3/0p Oo/AI B2g Oa2 Oe/Oi B2g Pergelic Cryoquept or Pergelic Cryosoprist ^^M Oe2 Diagnostic Soil Horizons, Subdivision ond Process b = buried f =frozen g=gleyed W^ CI I I Oi Oe Go Al B2 CIg Pergelic Cryoquept or Pergelic Cryohemist Pergelic Cryoquept or Histic Pergelic Cryoquept FIGURE 1-9. Ground oblique photograph of a lov^ -centered polygon, its microtopographic profile and thaw depths, and idealized soil prof iles for each microtopographic unit (basin, rim, and trough). The profile and thaw depths are lines established by the wooden stakes shown in the pho- tograph. Progression of depth of soil thaw is for summer 1973. 20 J. Brown et al. accumulation. The rims of low-centered polygons, the tops of high- centered polygons, and particularly the outer edges of both are more ex- posed to summer and winter microchmatic extremes. Since polygonal terrain forms the basis for much of the biological and pedological varia- tion found in wet coastal tundra, a brief review of its development follows. Shallow, narrow troughs averaging 0.5 m deep by 1 to 2 m wide and underlain by ice wedges form the outlines of polygons having diameters that range from a few meters to more than 30 m (average 12 m). Ice wedges develop as a result of cracking of the ground due to contraction during periods of intense and rapid winter cooling. The narrow thermal cracks are subsequently filled by ice in the form of winter hoarfrost or from spring meltwater that freezes. The cracking and ice filling, repeated over many centuries, results in the growth of vertical wedge-shaped masses of ice which penetrate many meters deep and may expand to several tens of meters wide (Lachenbruch 1962). The increase in volume caused by the expanding ice wedges produces buckling or heaving of the surfaces on either side of the wedges and par- allel to them. The ridges or rims so produced, together with melting of the tops of the wedges, cause depressions or troughs immediately above the wedges that further define the polygonal surface pattern (Figure 1-10). As a rule, polygonal terrain becomes more deformed and elevated with the passage of time, as ice wedges expand laterally and polygons subdivide, i.e. secondary ice wedge cracks form within the polygons. Troughs and central basins may deepen as the underlying wedges melt in response to climatic or microclimatic changes. Many of the basins may remain filled with water to form permanent or seasonal ponds. Where the drainage level has been lowered, as on stream banks and the shores of drained lakes, the trough produced by thawing of ice, ther- mokarst, and thermal erosion may result in topographic reversal of low- centered polygons. That is, a polygon center that was once low becomes elevated and better-drained with respect to the deepening troughs, and the low-centered polygon is thus converted to a high-centered polygon (Figure 1-10). In extreme cases, particularly where the underlying miner- al soil is sandy, the depression of the troughs may be more than 1 m, and the raised centers then consist of dry, peaty soil. Because of their elevated position, the raised areas are vulnerable to desiccation and wind erosion as well as to slumping into the troughs. These processes appear to be ac- centuated by lemmings, which find the features ideal nest sites and riddle them with burrows, causing further desiccation and oxidation of the peat. In situations where organic materials are removed altogether, the depth of summer thaw increases and frost heaving may convert the sur- face to small, bare frost scars or boils, or into hummocks. The sequence of events just described produces the contrasting types 2 - Surface Crack t Primary Ice Wedge 10 The Coastal Tundra at Barrow 21 J 20 25m 25 m FIGURE 1-10. Idealized ev- olution of polygonal ground from initial stage (top) to low-centered polygon (mid- dle) and finally high-cen- tered polygon (bottom). (Modified from Drew and Tedrow 1962 and French 1976.) of polygon systems seen on the Biome research area: the large, low- centered, rectangular type (orthogonal) polygons, many of whose centers are permanently water-filled, and the smaller and presumably older non- orthogonal type that shows a gradation of forms, including those with highly elevated centers or tops, broad, flat centers, and low-centered basins (Outcalt 1974). The more complex nonorthogonal polygons cover over 75% of the Biome research area. The microrelief pattern superimposed on the regional topographic and moisture gradients within the Biome research area creates diverse edaphic and biotic conditions which are further illustrated in the follow- ing sections. SOILS: DESCRIPTION AND DISTRIBUTION Despite the widespread development of polygonal ground most of the soils in the Biome research area have formed on flat to very gently sloping topography under cold, moist conditions that favor the accumu- lation of organic matter. A high proportion of the soils have a tripartite morphology: a histic or organic-rich surface horizon; a horizon of silty clay to silt loam textured mineral material, commonly gleyed and with variable amounts of included or enmixed organic materials; and an un- derlying perennially frozen organic-rich horizon that is commonly coex- tensive with the horizon above. The sequence of horizons has been inter- preted as reflecting the burial of organic materials by lacustrine sedi- 22 J. Brown et al. ments as a result of the thaw lake cycle or in some cases by frost heaving (Tedrow 1977). The surface horizon and mottling within the mineral hor- izons are the product of the current soil-forming processes of organic matter accumulation and gleization. Two soil orders are represented within the Biome research area: Inceptisols, mineral soils with poorly differentiated horizons, and Histosols, soils composed primarily of organic materials (Soil Survey Staff 1975). Because of the high moisture, the Inceptisols are classified as aquepts. Histosols are differentiated taxonomically into suborders re- flecting the state of decomposition of their organic materials. Because all the soils have a cold temperature regime the prefix Cry is added to each subgroup name. The term Pergelic preceding the subgroup name indi- cates the presence of permafrost. The term Histic indicates the peaty character of the surface 25 cm. The areal distribution of the soils or soil combinations is related to landscape units of the Biome research area (Figure 1-11). Weakly leached soils (Pergelic Cryochrepts), represented by the Arctic Brown Soil (Tedrow and Cantlon 1958), are present only on coarser-textured deposits on primary land surfaces such as beach ridges (Figure 1-3) and are not represented within the Biome research area. The soils, much like the vegetation, are arranged along a topo- graphic gradient ranging from the relatively freely drained and weakly leached Pergelic Cryochrepts (Arctic Brown Soils) to soils developed under conditions of extreme wetness, such as some Pergelic and Histic Pergelic Cryaquepts (Half Bog Soils). Intermediate and somewhat better-drained elements of the topographic-moisture gradient also have Histic Pergelic Cryaquepts and Pergelic Cryaquepts (Meadow Tundra Soils) developed in association with weakly expressed low-centered poly- gons and low relief high-centered polygons. Where the low-centered polygonal pattern is strongly developed, an association of soils occurs that is closely related to the elements of that pattern (Figure 1-9). Within the depressed polygon centers (basins), mod- erately decomposed organic materials constitute Pergelic Cryohemists (organic soils with more than 40 cm of organic materials — Histosols). On the rims of low-centered polygons the soils may be Pergelic Cryosaprists made up of highly decomposed organic materials, extensions of the basin Cryohemists, elevated and oxidized as the polygon rim expanded in re- sponse to ice wedge growth, or Pergelic Cryaquepts, little differentiated, mottled mineral soil with widely ranging amounts of organic materials. The latter are more common toward the troughs and show the effects of ice wedge expansion by the disruption of their horizons and enmixing of mineral and organic materials. Within the polygon troughs the soils may be Pergelic Cryaquepts, composed of coarse, fibrous organic material overlying gray or mottled materials. Where the fibrous organic material reaches a thickness of 25 The Coastal Tundra at Barrow 23 SOI CI C2 L COMPLEX (C) Pergellc Cryosaprist Histic Pergelic Cryaquepf Pergelic Cryoquept C3 SOIL ASSOCIATION(A) Pergelic Cryohemist A Pergelic Cryosaprist Pergelic Cryoquept Pergelic Cryoquept Histic Pergelic Cryoquept OTHER Pergelic Cryosaprist Pergelic Cryoquept Ul: Aerie Pergelic Cryoquept U2 Pergelic Cryoquept FIGURE 1-11. Generalized soils map of the Biome re- search area. Soil complexes (C) include two or more soils with no consistent spatial relationship. Soil asso- ciations (A) consist of two or more soils (three at Bar- row) in a predictable spatial arrangement, e.g. associ- ated with polygon rims, centers and troughs. In other areas (U) a single soil dominates. Details of the soils for site 4 are shown in Figures 1-8 and 1-9. CO S CO s I PQ < is o — so o eg *" in c. — 'o y) >-m ^ 3 i i a It -* o •a '_» — a> ■J = '/■ w^ c o J^ ? 6 c; 1 , Qb O .':i ^ o J. O ^ JZ d — E » c. E _ * -5 < "7 - t— -O — 3n — = ^ i O eg — C- 00 ^' o -a c cot E ^ o P. g^ E ? Z Q 9 z .A Q 2 Q Z z Q Z o »/^ d d — m — (N 2 z Z 2 Z ! — > > _ > — > > > > < s ~ o — -3 'o S o ~ o Q. £1. Sd^ ^ >, o lei- -= 0-3 X < o >-. s re o r-i £ — re ■o JZ '-» -3 s 3 3 E Stl c J. y T3 c >. "C c ^ eg OS "7 •^ — r o 'o 1/ C- "3 OJ It 3 CO H re > 3 S 1 o o "re >/". 1 5 o !> >i -a i» u ■o re 1 u w 5 o Q£ 1 1 Z 1 J. 1 1 1 1 24 The Coastal Tundra at Barrow 25 cm, the soils are considered Histic Pergelic Cryaquepts. Rarely is the or- ganic material of sufficient thickness that Pergelic Cryohemists or the very fibrous Cryofibrists can be recognized. This association of soils repeats from polygon to polygon. On the centers of prominent high-centered polygons PergeHc Cryo- saprists are found. They are the products of oxidation of Pergelic Cryo- hemists of the former polygon basin. Lighter-colored Aerie Pergelic Cry- aquepts are found on the coarser-textured and better-drained mineral soil slopes marginal to Footprint Creek. The maximum thaw depths encountered across the entire Biome re- search area ranged between 24 and 60 cm (Table 1-3). Thaw variation as a function of microrelief showed less deviation from the mean early in the season than later in the summer, when site characteristics had had time to exert their full influence. However, thaw differences between rims of low-centered polygons and centers of high-centered polygons de- crease during the season as the tundra surface becomes generally drier. Measurements of mean mid-August thaw depth from 1970 to 1974 in 10 control plots on the Carex-Oncophorus meadow (site 2) showed only a 2-cm difference between years (25.2, 26.9, 26.7, 24.7 and 24.7 cm). The small difference in mean depth between summers with quite different meteorological conditions indicates that the soil that remains near satu- ration is subject to a relatively small mean fluctuation in thaw. However, for a more diverse terrain some 4 km to the northeast. Brown (1969) measured average thaw between 33 and 43 cm over the period 1962 to 1966. VEGETATION Sedge meadows cover about three-quarters of the Biome research area (Webber 1978) (Table 1-4, Figure 1-12). The meadows are domi- nated by a single species, Carex aquatilis, and commonly have only a few secondary species, such as Eriophorum angustifolium, E. scheuchzeri and Dupontiafisheri. They also have a large moss component consisting of species of Calliergon and Drepanocladus. Lichens are a minor compo- nent of these meadows (Murray 1978). Complete species nomenclature for the Biome research area is given in Murray and Murray (1978). The most striking feature of the vegetation is that it changes char- acter every few meters in response to change of microtopography and drainage (Figure 1-8). The vegetation changes are best indicated by varia- tions in the subordinate species, especially the mosses and grasses. Thus the ubiquitous Carex aquatilis dominates the meadow vegetation types which occur along a moisture gradient from dry to wet. Eight major vegetation types were recognized and mapped (Figure B3 ^^ <4; -5: s: o 51» SQ < H i_ D. 2 O W D. ^ O ^ o E C HJ E o Q c o 00 >^ "o a T3 c o op a. o - I s: o •t:^ 3 (3 t 3''C5 •a c OJ c o 00 _>. o a T3 U C o CO ^ c ,— ^ 0 "" 2 a ^ a> >. « 00 *-• j= > s 3 ■-J ■a !« -o-a c o CS N 1^ "c c o o 00 00 >> O Q. lU D. .0 5 s: 3 ^ ^2 c o 00 _>> o a T3 u c c o ^ 3 — t:;; 00 t « o S ■Is -2^ 0^^"$ I I 53 -3 Q, o -^ •i2 >; -^ 53 c: -s: §".2 0(J 00 3 tX s: 53 •is ^1 ^ ^ 00 53 c:^ =s ^ - ■=5^3 S ci S 00 c o 00 _>> o a T3 c s ^ -= 5 l>5 <3 — (3 U UQ O l> s: ^ c n o 00 _>. "o o. T3 C s 3 S| 00 ?3 3-2* '^ -5: O oj •2 -i: c ■^ >) t ■?§^ 53 I: .3 'c: 3 -3 3 (b 53 QUO ■ ? c — 00 o| (/I ^ c CQ a I 3 Co ? s: O E =: §^ 3 ■C ■- lJ ci. 00 s: ^ u !> C c/3 c Si 1) u T3 X) o c C3 io 0, 5U :^ ? 3. (VI ,53 c ?3 ^■2 S-E 2 5b -Si c « ^ b '-J ?3 1) 1^ w 0 U E X E U 1» > o o c 3 o E T3 O C T3 o u D. (/5 26 The Coastal Tundra at Barrow 27 I56'>4l Site 7- '^-7n7'45" 71° 17 30 '.uzula heoth(l) Arctophila pond margin(VII) ^ Carex-Oncophorus ^ meadow (IV) \Salix heath (II) Wyi Dupontia meadow(V)| 1 CocA/ffOA/'<7 meadow |tro/-e>r->Oo(7 meadow ■ Carex-Eriophorum I (III) ^^ meodow (VI) \G rovel road FIGURE 1-12. Vegetation map of the Biome research sites. The types have been coded according to their most abundant vegetation. The de- tails of the vegetation for site 4 are shown in Figure 1-8. (After Webber 1978.) 1-12) and an estimate of the area occupied by each type was computed (Table 1-4; Webber 1978). Luzula heath (Type I), which occupies only 3% of the Biome research area, is characteristic of high-centered poly- gons and some raised beaches. The low density of species and individuals within the Luzula heath is due to dryness, elevation above the water table, thin snow cover, and exposure to winds. The characteristic growth-forms present are caespitose graminoids and lichens. Characteris- 28 J. Brown et al. tic species include Luzula confusa. Potent ilia hyparctica, Alectoria nigri- cans and Pogonatum alpinum. Salix heath (Type II) is characteristic of the sloping creek banks and the centers of some low-centered polygons which drain readily. It occu- pies 1% of the research area. Salix rotundifolia, a prostrate deciduous shrub, Arctagrostis latifolia and Saxifraga nelsoniana characterize this type. The Salix heath has the highest number of forbs (22) and the high- est overall number of species (70) of any of the vegetation types. The sandy soil associated with Salix heath is usually well drained and has the greatest seasonal depth of thaw. Carex-Poa meadow (Type III) is the most extensive vegetation type, covering 41% of the research area. Bryophytes, primarily mosses, and lichens cover relatively larger areas than do the graminoids. The mosses Pogonatum alpinum and Dicranum elongatum and the lichens Cetraria richardsonii and Dactylina arctica are characteristic species. The Carex- Oncophorus meadow, Dupontia meadow, and Carex-Eriophorum mea- dow all have the same superficial meadow physiognomy. The Carex-Poa meadow is separated from the other meadow vegetation types by the abundance of Poa arctica, which is prominent when in flower, and by the presence of a high lichen diversity. Carex-Poa meadow is best developed in the drier parts of the polygonized sedge meadow complex where large, flat, and sometimes slightly raised low-centered polygons with barely dis- cernible troughs occur. The Carex-Poa meadow is also common on hummocky rims of low-centered polygons in wetter areas and in areas with pronounced polygon development. Carex-Oncophorus meadow (Type IV), the second most extensive vegetation type, occupying 2\% of the map area, develops on moister sites than the Carex-Poa meadow. It occurs on flat polygon centers and in drained, shallow polygon troughs. It is distinguished from Carex-Poa meadow on the basis of reduced lichen cover and the presence of mosses such as Calliergon sarmentosum, Oncophorus wahlenbergii, and Aula- comnium turgidum, and of Dupontia fisheri and Peltigera aphthosa. The Carex-Oncophorus meadow has the greatest number of graminoid species and was the vegetation type most intensively studied during the Tundra Biome program. Dupontia meadow (Type V) occupies l^o of the mapped area. It is characteristic of flat, slowly draining sites and wet polygon troughs. Eri- ophorum russeolum is also common in this type, but the low cover of Carex and the greater abundance of Dupontia and Eriophorum angusti- folium distinguish it from the other meadow types. Woody dicotyledons are essentially absent but several forbs occur. Cerastium Jenisejense, Stellaria edwardsii and S. laeta may form distinctive mats, and Saxifraga cernua is a common erect forb. The vegetation that composes the Carex-Eriophorum meadow The Coastal Tundra at Barrow 29 (Type VI) is variable. This type includes the sparse vegetation of the basins of low-centered polygons as well as the more abundant vegetation of some pond margins. It has no woody species, and only a few lichens. Saxifraga foliolosa is characteristic of the basins and CalUergon sarmen- tosum is characteristic of the pond margins. Drepanocladus brevifolius is common throughout this type. Low species diversity, thick organic mat, and highly organic soils distinguish Type VI from the other meadow types. Spring snow cover on this type is moderately deep and late to recede. The vegetation type termed Arctophila pond margin (Type VII) is also found in slow-flowing streams and is the most distinct of the princi- pal types, although it occupies only Z^/o of the area. Woody plants and h- chens are totally absent. Single-shooted graminoids, erect forbs and bry- ophytes all contribute to this vegetation type, although the emergent grass Arctophila fulva is characteristic. Ranunculus pallasii and Caltha palustris are occasionally present. The pleurocarpous mosses CalUergon giganteum and Drepanocladus brevifolius are the principal bryophytes. Late in the growing season the substrate oi Arctophila pond margin may occasionally become anaerobic and have a characteristic odor of hydro- gen sulfide. The Cochlearia meadow (Type VIII) is a rudimentary community that develops on recent alluvium in the creek bed that crosses the Biome research area, where snow commonly accumulates and remains late in the growing season. The sandy, moist alluvial soil of the Cochlearia mea- dow thaws deeply and the active layer may exceed 1 m in late summer. This vegetation type is floristically rich, but extremely variable. Cochlearia officinal's, Phippsia algida and Stellaria humifusa are the principal species present. The variations in vegetation and soil across microtopographic fea- tures are a dominant aspect of the coastal tundra at Barrow. Many of the following chapters deal with the structure and function of tundra organ- isms across these unique landscape units. Climate, Snow Cover, Microclimate, and Hydrology S. L. Dingman, R. G. Barry, G. Weller, C. Benson, E. F. LeDrew, and C. W. Goodwin INTRODUCTION The environmental conditions within a few meters above or below the ground surface constitute the microclimate of a region, and it is to these conditions that most terrestrial organisms must adapt. The micro- climate is characterized by the radiation, temperature, and moisture re- gimes of the near-surface atmospheric and soil layers. These regimes are determined largely by the regional climate, as modified by local top- ography and by the vegetation cover. Quantitatively, the microclimate is described by the energy and water balances at the surface. These balances are complexly related. The net radiation input to the land surface provides the energy that is utilized in surface physical and biological processes. The energy balance express- es how this energy is partitioned, and the water balance is determined by the energy balance. At the same time, the water balance influences the magnitude of the energy balance components through latent heat ex- changes and through the effects of snow on surface conditions. For comparison, descriptions of radiation, energy and water bal- ances at other arctic sites can be found in Ohmura (1972), Weller and Holmgren (1974a and b). Woo (1976), Ohmura and Muller (1976), Stewart and Rouse (1976), Courtin and Labine (1977), Ryden (1977), and LeDrew and Weller (1978). Dingman (1973) published an annotated bibliography covering much of the pre-1972 literature on the water bal- ance in arctic and subarctic regions. CLIMATE Climatic Setting Traditionally, areas where the average temperature of the warmest month is below 10 °C have been identified as having a tundra climate 30 Climate, Snow Cover, Microclimate, and Hydrology 31 (Kbppen 1936). More recently, temperature data were used to identify air masses, and Bryson (1966) showed that arctic tundra areas in North America have at least a 50% frequency of arctic air in July. The median location of the Arctic Front in July corresponds approximately with the northern limit of boreal forest (see map inside front cover), although in eastern North America the definition of the frontal zone and its relation- ship with vegetation boundaries is less certain (Barry 1967). The exact causal relationship between this frontal position and the biota remains uncertain, but the lack of trees in the arctic tundra must be attributable in part to the poleward decline of available surface energy. Annual net radiation at the treehne is about 670 MJ m"^ in Alaska and 750 MJ m'^ in central Canada, and decreases to 400 to 600 MJ m"^ over the tundra (Hare and Ritchie 1972). The general weather conditions in northern Alaska are a result of the patterns of atmospheric circulation. The mean sea level pressure map shows that in winter Alaska is influenced by easterly arctic airstreams as- sociated with a deep low over ihe Aleutians and a ridge of high pressure from the Mackenzie Valley across the Arctic Ocean towards eastern Siberia. A ten-year analysis of daily weather maps by Putnins (1966) il- lustrates the winter maximum of anticyclonic patterns with high pressure cells predominantly to the north and east of Alaska. Winter conditions at Barrow are similar to those over most of the Arctic Slope, but during summer its coastal location gives it a modified arctic tundra climate (Watson 1959, Searby and Hunter 1971). The temp- erature regime north of the Brooks Range is continental, with an annual range of 32 °C for monthly mean temperatures (Table 2-1). Extreme max- ima and minima have ranged between 26 °C and -49 °C at Barrow, and between 39°C and -61 °C (unofficial reading) at Umiat in the central Col- ville River Valley (Conover 1960). Mean daily temperatures are below -20 °C at Barrow from December through March. February is the coldest month, with 90% of hourly temperatures below -18 °C (Rayner 1960a, b). The severe temperatures are accompanied by moderate wind speeds, averaging 5 ms"', which often cause blowing snow. In contrast to inland localities, the wind is seldom calm at Barrow; speeds above 12 ms'' are also infrequent (Figure 2-1). Winds along the arctic coast are easterly ex- cept in the vicinity of Barter Island, where the regional topography fre- quently causes strong westerly winds in winter (Schwerdtfeger 1975). The winter temperature regime is closely related to the presence of a surface temperature inversion which has an average intensity of about 12°C over a vertical distance of 750 m (Bilello 1966). A ground-based in- version is present on 62% of winter days, and for half of these days the inversion layer is more than 1000 m thick (Holzworth 1974). The result- ing stability of the lower troposphere implies that day-to-day tempera- ture changes are largely determined by changes in cloud cover and the 3 c ■•*,* c ^ < X s: 53 Q -V 03 ^ ^ »^ Z ,0 '^ k: 0 0 1 § ^ .6c C/) i^ ~- a, |; .^ "5j ^ > C3 "-- < 0 /^ /^ r^ -^ •^ O On O 00 t-~-' On t^ O O — — 00 O ""1 (N NO NO •^ V- 1^ <-n 00 O rs I I I I I •/-I m 00 r- 00 /-l NO •^ Tt Cvl ro 00 rt 1/1 1^ NO NO NO O ON r- rn 00 00 00 f- (N 00 — cnI m (N fN cn) 1 1 1 1 1 1 CN) m rn (N r<-i 1 1 1 1 1 1^ NO 00 r<^ On •rf 00 NO lO "/-^ NO 1 r4 1 1 1 > NO 00 1^ 00 On NO On ro NO m rr, 00 r~- ^ 00 r^ cnI rvl 00 On 00 On Tt NO — 00 NO q O r^ (N NO e s >r^ t^ rn r~ NO rs| ^^ ..^ (N a. a F u at O00rnON OnnO (Nno— ••^■^no r-(N 0Or^«Nr<-i>r) oorn r^OOrnw^OO >/^On >/~. — O 00 r^ — ; Tt H — -^ r-' NO NO NO d m r<^ 00 — • — • (N ri On On o ■a 3 o - 03 o ^^ . . 3 o C/5 32 Climate, Snow Cover, Microclimate, and Hydrology 33 204- u S -!0 - E -20 - -30- -40- -50 \ January _i L. _i L. 4 8 12 Wind Speed, m s"* FIGURE 2-1. Cumulative fre- quencies of hourly air tempera- tures and wind speeds at the vil- lage of Barrow, 1945-54. The fre- quencies are expressed as the cumulative percentage frequen- cies of the total distribution. (After Rayner 1960a, b.) effect of clouds on infrared radiation, rather than by changes of air mass. Deep Pacific cyclones affect southern Alaska and the Bering Sea but seldom penetrate to the Arctic Slope. When they do, the frontal sys- tems tend to move above the 1000- to 1500-m-deep inversion layer and may not affect the weather at the surface. The persistence of anticyclonic conditions during late winter leads to clear skies and the receipt of a high percentage of the possible solar radia- tion. The mid- to late winter period has an average of 10 or more days per month with clear skies, associated with the 45 to 5097o frequency of anticyclonic patterns. Temperatures begin to rise in late winter but lag 4 to 6 weeks behind the increase in solar radiation. The mean daily temperature is above freezing from June through August, but fails to reach 5°C even in July as a result of coastal cloudiness and the effect of the Arctic Ocean. The Arc- tic Ocean maintains a cover of close pack ice in summer except for open water areas which extend some 30 to 100 km from the coast by August or September. Even in August the open water is only about 3 °C (Searby and 34 S. L. Dingman et al. Hunter 1971). This cold surface cools the lowest layers of the atmos- phere, thus providing a source of cool air and sea breeze conditions (Moritz 1977, Walsh 1977). Forty-three percent of hourly temperatures in July are at or below 0°C (Rayner 1960a, b), and on the average there are only 91 days with a mean daily temperature above 0°C each year. Cloudiness increases during May and June as moisture becomes available from open areas in the sea ice. Fog formed over the Arctic Ocean drifts inland, often as low stratus that forms following warming of the surface air. Convective activity further modifies the cloud type to stratocumulus. At the coast, heavy fog occurs on about one day in three from June through August. In May and June this fog and cloud cover re- duce the proportion of possible solar radiation actually received at the surface and depress temperatures near the coast. Simulations by Lord et al. (1972) show the importance of such cloud cover in suppressing the di- urnal range of temperature at the surface. However, until snowmelt the effect of cloud cover is in part offset by the multiple reflections between the snow surface and the clouds. During the summer temperatures increase from the coast inland. The mean daily temperature in July at Umiat for the years 1947-53 was 12°C. This inland warmth was demonstrated in July 1966 when the air temperature at the coast averaged 10.6°C when winds were southerly and 2.7 °C when winds were off the ocean (Weaver 1970, Barry et al. 1976, Myers and Pitelka 1979). According to Brown et al. (1975) the mean tem- perature gradient inland from the arctic coast is 6°C per 100 km, al- though they suggest that much of this warming occurs close to the coast where the gradient may be two to three times greater (Walsh 1977). The tundra at Barrow is typical of the very cool, moist zone in the immediate vicinity of the coast. In the Prudhoe Bay region, the area south of the bay is transitional to the warmer inland zone which extends into the Foothills, but still is significantly cooler than locations 20 km or more in- land (Walker and Webber 1979). Average thawing indices, the sum of the positive differences between daily mean temperature and 0°C, for 1970-73 are 304 at Barrow but 477 at Prudhoe Bay (Brown et al. 1975). From July through September the strong temperature gradient from the snow-free land to the cold, largely ice-covered Arctic Ocean sets up a horizontal density gradient that is referred to as the Arctic Front. An average of four weak lows per month travel eastward along this frontal zone from the Siberian arctic coast. About 35% of the annual precipita- tion falls, mainly as rain, during these three months in association with these systems (Table 2-1). The Arctic Front is generally at a height of 300 to 700 m over the coast, and reaches the ground about halfway between the coast and the Foothills (Conover 1960). When the front moves in- land, the shallow layer of arctic air that covers the whole Arctic Slope produces cold, foggy weather. This air tends to advance up the river val- Climate, Snow Cover, Microclimate, and Hydrology 35 leys so that hilltops a few hundred meters in elevation may be warmer than the valleys below. Less frequently, the front retreats poleward and thus allows the entire area to be warmed by land winds. An analysis of the Arctic Front by Streten (1974), using very high resolution radiometer (VHRR) imagery from the NOAA-2 satellite, showed considerable varia- bility in cloudiness associated with the frontal zone. It also confirmed the front's northward advance from early to midsummer, as determined by Barry (1967), and its southward retreat by August. By October, the circu- lation regime has reverted back to its winter mode as the arctic inversion redevelops over the snow-covered land. Annual precipitation on the Arctic Slope is light, although data are inadequate to describe regional trends in any detail. Clebsch and Shanks (1968) found that amounts in July and August 1956 were 50% greater at stations located 2, 15 and 55 km inland than at the first-order weather station at Barrow. Theoretically, easterly airflow along the arctic coast should tend to produce divergence and suppression of precipitation over the coastal strip (Bryson and Kuhn 1961), although this effect is less in higher latitudes than in low latitudes. At the present time, data on sur- face and upper winds and on precipitation amounts are inadequate to test this hypothesis. Total precipitation is undoubtedly higher to the south and in the Brooks Range, with an estimated 500 mm falling on the upper McCall Glacier at 2275 m (Wendler et al. 1974). Kilday (1974) indi- cated that annual totals exceed 500 mm over most of the Brooks Range and are 1000 mm in the wettest eastern section. In the coastal tundra at Barrow, rainfall intensity is low, with frequent drizzle and light snow falling during May and June from the coastal stratus clouds. Freezing rain also occurs during this season as well as during the autumn transi- tion. The maximum 6-hour total rainfall recorded is 15 mm (Miller 1963); a 24-hour snowfall of 38 cm has been recorded in October. Interannual Variability and the Representativeness of the Biome Years The interannual variability, expressed by the standard deviation of mean monthly temperature, is approximately 3 °C during the winter months, October through February, but declines to just over 1 °C in July and August. The winter variability is associated with shifts in the Arctic Frontal zone. The extreme variations from the 1924-73 mean occurred in January 1930, with a deviation of + 11.4°C, and January 1925, with a deviation of -8.1°C. During the summer months there is less departure from the long-term average; the extremes are -i-4.9°C in August 1954 and -3.6°C in August 1956. Another approach to calculating interannual variability using cumu- 36 S. L. Dingman et al. June Temp 09 ±1 Z'C; Ppt, 7.9 ±7.2 mm DryP- H HI H Cooll ■ ■ H ^^H ■ ■ worm ^^mr//i m m ^^" ^« Moist ^B ■ ■ WA im 111)11 ' ' 1 ' -^ ' ^T — ] — I — T-n — I — I — I — r — I — 1 — p August Temp: 3.2 ILe'C; Ppt. 23.6 1 14.4mm n — I — r — I — I — I — I — r — 1 — I — 1 — I — I — I — I — I — r — i — i — i f r — i I i — i — i — r — i — i — i — i — i — i — i — i — r "(" i — i — i — i — i — i — i — i — r- 1924 1930 1940 1950 I960 1970 □ 2StdDev FIGURE 2-2. Summary of deviations of temperature and precipitation from monthly means (1924-73) during summer months for the village of Barrow. lative thawing degree-days has been taken by Myers and Pitelka (1979). They demonstrated that 7 of the 26 years from 1950 to 1975 deviated more than 100 cumulative degrees Celsius from the average daily mean temperature. Early spring and late July and August are shown to vary more from year to year than June and early July. The high variability in August is related to the strong contrasts between temperatures associated with winds blowing over the Beaufort and Chukchi Seas or over land. Hence, climatic conditions near the coast are closely determined by at- mosphere-ocean-ice interactions (Barry et al. 1976). The precipitation data, which are subject to measurement problems, have much greater in- terannual variability. The long-term patterns of some individual months indicate that the mean departure for December 1932-41 was +2.0°C, whereas that for December 1953-61 was -2.2°C; this difference is significant at the 2^o level by Student's t. For June, the period 1925-32 had a departure of + 1.3°C; this difference is significant at the 5% level. There are no com- parable shifts for July or August, ahhough July 1959-70 averaged 0.8 °C below, and August 1959-71 0.9°C below, the respective 1924-73 mean values. When running-mean techniques are used on individual summer months at 5-year intervals, there are suggestions that the mid-1950s and the late 1960s were cooler than the long-term means. Rogers (1978), us- ing a linear regression of thaw season degree-days against time, found a summer cooling trend over the 56-year period of air temperature records at Barrow. In terms of the departure of total precipitation from the 1924-73 monthly averages, there was a tendency for more dry months in summer between 1964 and 1972, whereas the 1950s were generally wetter Climate, Snow Cover, Microclimate, and Hydrology 37 IJ 1 1 0-1 cm depth o . A . * tf° * 0) • O r Q ■ • ° o *• A O '. 0 - 1 1 1 5 15 15 Jun Jul Aug - 10 - - 5 - - 0 - 1 Site 1 1 Site 4 1 ' 20 cm deptti • 1970 A 1972 - o 1971 A 1973 — Site 2 ■ 1971 D 1973 - - ai- o S • ■ a 1 o ■ . • ■ 0 _ 1 5 15 15 Jun Jul Aug FIGURE 2-3. Ten-day mean soil temperatures at the Biome research area as measured during 1970-73 and simulated for the extreme years during 1960-73. The shaded area represents the range of the simulated soil tem- peratures for the warmest year, 1968, and the coldest year, 1969. than average (Figure 2-2). A major question in any short-term observational program is how representative the data are with respect to long-term conditions and trends. During the 4 years of Biome measurements the positive depar- tures of temperature from the long-term summer observations were espe- cially pronounced during 1972: -i-2.4°C in July and -i- 1.7°C in August (Figure 2-2, Table 2-1). Precipitation also varied from the long-term averages during the research period: July was much drier than average in 1970 and 1972, and slightly wetter than average in 1971 and 1973. In order to compare the representativeness of soil temperatures dur- ing the Biome years with those in the previous decade when temperatures were not measured, a surface equilibrium temperature model was used to simulate soil temperature and thaw for the period 1960-73. The model, a modification of one developed by Outcalt et al. (1975) to simulate annual snow and soil thermal regimes, predicted daily soil temperatures on a 5-cm grid for the 14-year period. The actual data obtained for the Biome years are within the predicted extremes of simulated soil temperatures (Figure 2-3). Therefore it is reasonable to assume that soil temperatures observed during the 1970-73 Biome period are within the range of varia- tion normally encountered on the tundra at Barrow. SNOW COVER Most plants and animals are small enough to live within the protec- tive blanket of snow, and larger predators and herbivores depend. 38 S. L. Dingman et al. " 40 '/ . r*^ • \: Snow Depth, O o 1 Sep Oct Nov Dec Jan Feb Mar A pr May J un 1970 »i 1971 FIGURE 2-4. Development of the 1970-71 snow cover on the Biome re- search area. Each point represents the average of 15 measurements. (Weller and Benson, unpubl.) directly or indirectly, on food-chain members harbored within the snow cover. Thus a knowledge of snow distribution and structure is an essen- tial part of the understanding of tundra ecology (Pruitt 1960, Formozov 1961). The most detailed work on the nature of the snowpack on the Alaskan Arctic Slope is reported in Benson (1969) and Benson et al. (1975). The major surface processes during the winter are the accumulation of snow, its redistribution by wind, and the transfer of mass and energy within the snowpack. Over most of the tundra at Barrow, the snow builds to a thickness of about 20 cm within 10 to 20 days after freeze-up, following which it continues to increase very slowly in thickness (Figure 2-4). The initial snow cover markedly reduces surface roughness and per- mits more effective snow drifting. Snow is deposited in drift traps and, as the winter progresses, the surface relief becomes more and more sub- dued. With the establishment of the snow cover and the virtual disappear- ance of solar energy input at the surface the net radiation becomes nega- tive. The radiational cooling at the surface sets up fairly steep tempera- ture inversions above the ground, so that both atmosphere and ground supply heat to the surface, one by eddy diffusion, the other by conduc- tion. These conditions induce a flow of heat and moisture from the soil surface upward through the snowpack, driven in part by wind and baro- metric fluctuations (Benson 1969). There is a consequent general drying of the upper soil layers and formation of depth hoar in the snowpack. This results in a net gain of moisture in the snow because of the upward transport of water. Some condensation at the snow surface also adds to the mass of the snow (Weller and Holmgren 1974a). Although fresh snow continues to accumulate throughout the win- ter, the steady winds constantly reshape the pattern of the snow cover. Across the smoothed surface, barchan-type dunes form and move during storm periods. During less windy periods, these dunes stabilize and become drift traps for future storms. Exceptionally high winds may Climate, Snow Cover, Microclimate, and Hydrology 39 TABLE 2-2 Types of Snow Found on the Windblown Arctic Coastal Plain Range of Range of grain size density Snow type (mm) (g cm')* Fresh new snow, variable 0.5 to 1.0 0.15 to 0.20 crystal forms sometimes <0.5 Hard, fine-grained 0.5 to 1.0 0.35 to 0.45 wind slab Medium-grained snow 1 to 2 0.23 to 0.35 Depth hoar: soft, coarse, 5 to 10 0.20 to 0.30 loosely bonded crystals *The density ranges are approximate, but they indicate the differences one may expect between the various layers. remove or reshape the old dunes and even reexpose some of the vegeta- tion. The interaction of wind effects and transfer processes within the snowpack produces four snow types (Table 2-2) and four typical snow structures: 1. A hard, fine-grained, high density, windpacked layer overlying a soft, coarse-grained, low density layer. The upper layer is fre- quently a wind slab; the lower is almost entirely depth hoar. This two-layer structure is the most common, being present over about 80% of the open tundra. 2. The hard wind slab layer alone, found over about 15% of the open tundra. 3. The soft depth hoar alone, covering about 5% of the open tun- dra. 4. A complex, deep snowpack, largely wind slab and medium- grained snow, in natural drift traps such as creek bottoms. The development of the two-layer structural type as seen in Figure 2-5 is a direct consequence of the depositional history of the snow. The rapidly deposited snow layers of September contain few significant wind slabs and are the primary units in which depth hoar develops. As the season progresses, the wind causes very hard snow layers to form as small increments of new snow are added. Admixtures of silt or fine sand from exposed roads or dunes, or of particles of vegetation, strengthen wind slabs. The continual heat flux from the soil below the snow keeps the soil/ snow interface temperature above that of the upper snow surface. This 40 S. L. Dingman et al. 0) o Profile B Ram Hardness 100 0 Profile A Rom Hardness 100 0 Density, g cm" Snow 0.30 040 Chorocteristics I 1 1 -20 -10 Temp,°C ' ')1 Fine groined ^ Medium to coorse_gr^iti_ed_ _ u Depth tioar 1 1 "^iX FIGURE 2-5. Profile of the snow cover showing depth hoar above the vegetation and ice layers, and plots of physical properties. ("Ram hard- ness" is a measure of the snow's resistance to the penetration of a cone under an impact of known energy.) (After Benson 1969.) Climate, Snow Cover, Microclimate, and Hydrology 41 temperature gradient is accompanied by a vapor pressure gradient, which results in upward transfer of water vapor accompanied by recrystalliza- tion of the lower part of the snowpack into depth-hoar crystals. Most of the mass removed from the lower part of the snow cover is redeposited in the colder upper part, where it aids in the development of wind slabs. The rate of mass transfer within snow in interior Alaska, where stronger temperature gradients persist in the snow, has been calculated by Trabant and Benson (1972) and is on the order of 0.02 to 0.03 g cm"^ day"'. The depth hoar layer is so fragile that it often disintegrates with the slightest disturbance, causing the collapsed snow or snowquakes com- monly observed when one walks in an undisturbed area. These snow- quakes are first observed in November and then with increasing fre- quency until April, and are generally restricted to a few thousand square meters in area because of the support afforded by hummocky microrelief. The second structural type, a hard layer of snow without underlying depth hoar, develops after wind erosion entirely removes the snow cover from a small area. Subsequently, a new wind slab forms almost directly on the surface. This process has been observed during a smgle storm, i.e. as the winds shift slightly in direction, an area may change from an ero- sional regimen to one of deposition. Rarely does an area remain denuded of snow for long. The third structural type is rare and occurs only when wind erosion removes the wind slab layers, leaving the depth hoar. The structure may be somewhat stabilized by a thin wind crust, which is usually removed or covered by the next wind event. Almost any irregularity on the surface serves as a drift trap, at least under some wind conditions. The snow depth is generally related to the height of the vegetation. Snow is also caught in the low areas between polygons, which generally become filled by mid-October. Along the coast drifts often exceed 4 m in depth. However, the most extensive drifts accumulate in stream channels incised a meter or more below the tundra surface. For example, on the Meade River at Atkasook drifts are often several kilometers long, up to 20 m wide, and 10 m deep. The large drifts that form on the banks of rivers and lakes are sepa- rated into two groups, one formed by storm winds from the west which bring most of the new snow and the other by the prevaiUng northeasterly winds. The general shapes of the drifts are reproduced each year. The sizes and shapes of the prevaihng-wind drifts are virtually independent of the amount of snowfall. However, the sizes of storm-wind drifts vary significantly with the amount of snowfall. As an example of this process cross sections of drifts on the banks of the Meade River were measured between 1962 and 1973 (Figure 2-6). Drifts caused by storm winds were at 42 S. L. Dingman et al. 0 2 4 6 8 10 FIGURE 2-6. Profiles of snowdrifts on the banks of the Meade River at Atkasook (70°29 'N, 157°24 ' W) as seen from the north. The drifts on the west side are formed by storm winds from the west and vary in size with the amount of snowfall. The drifts on the east side are formed by prevailing winds from the east and are virtually independent of the amount of snowfall. (After Benson 1969.) a minimum in 1964 and a maximum in 1967, while prevailing wind drifts were nearly constant in size (Benson 1969). By early December and until the following spring melt, diurnal changes in air temperature do not appreciably influence the soil thermal regime. The insulating effect of the winter snow causes the amplitude of daily temperature fluctuations to decrease with depth and results in Climate, Snow Cover, Microclimate, and Hydrology 43 •50 -45 -40 -35 -30 -25 -20 a. E ^^'^^'^^^'^^^^'^^TT^m^Tj^m^Tmm^y^ w/Avy/'/'A'^V/'A'v^yxvy/i^ March 1971 FIGURE 2-7. Temperatures in the snow during 1-6 March, the coldest period in 1971. The numbers on the graphs give the distance above the ground surface at which snow temperatures were measured. warmer conditions at the snow/ground interface (Figure 2-7). Although the temperature at the air/snow interface during the 6 days shown in Fig- ure 2-7 ranged between -47 °C and -35 °C, the temperature at the snow/soil interface stayed within a degree or two of -26 °C. The snow temperature phase lag and amplitude attenuation indicate that the effec- tive thermal diffusivity of the snow is larger by a factor of 3 to 4 than that of the organic materials just below the snow/soil interface. Thus, the rather stable and relatively warm thermal environment at the snow/ soil interface is produced by the mass of snow overlying the interface and not by the thermal properties of the snow cover per se. Increasing depths of snow favor a more moderate snow/soil interface environment, with reduced temperature extremes and a higher mean temperature. The typical two-layer snow structure is the most favorable environ- ment for small mammals like lemmings. The loose depth-hoar layer gives little resistance to burrowing, and the hard wind slab provides protection from wind and from predators. At a snow fence site (Slaughter et al. 1975, Outcalt et al. 1975), where the maximum snow depth was increased to nearly 2 m by drifting, there was evidence of much more intense lem- ming activity and nesting than on the surrounding tundra, where the natural snow depth was less than 0.5 m. Other mobile organisms also 44 S. L. Dingman et al. seek out natural deep snow areas to take advantage of the more moderate interface climate. Deep-drifted snow in narrow gullies, between high- centered polygons, and near ponds is a favorable winter habitat. The other two structural types, hard snow and soft snow alone, are relatively unfavorable sites. When the wind slab is very near the surface it effec- tively cuts off the area; no burrowing was found in hard wind slabs. If only a depth hoar layer remains after removal of the wind slab, low tem- peratures occur at the soil/snow interface because of the high air perme- ability of the depth hoar. MICROCLIMATE Definition of Microclimate Seasons A characteristic succession of physical processes is observed each year on the tundra: establishment of the winter regime through large radiant energy losses, modification of the snowpack just prior to melting, gener- ally rapid melting and consequent large runoff, high evaporation from the water-covered tundra after snowmelt, and the relatively dry summer regime followed by freeze-up. This progression is accompanied by a characteristic pattern in the relative magnitudes of the components of the radiation, energy, and water balances that provides a convenient basis for identifying six seasons: winter, pre-melt, melt, post-melt, summer, and freeze-up (Weller and Holmgren 1974a). Although the starting dates and durations of these seasons vary from year to year, it is possible to specify typical values (Figure 2-8). Earlier studies at Barrow have suggested somewhat different bases for dividing the year into seasons: Kelley and Weaver (1969) defined six seasons based on soil-temperature regime, and Maykut and Church (1973) described four seasons based on variations of surface albedo. Figure 2-8 shows the magnitudes and signs of the energy-balance components typical of each season. The radiation balance is negative during the winter, and positive during the rest of the year. The hydro- logic regime is dominated by increasing storage during the freeze-up, winter and pre-melt seasons, by runoff during the melt season, and by evapotranspiration during the post-melt and summer seasons. Radiation Balance The radiation balance describes the partitioning of radiant energy into incoming and absorbed solar radiation, longwave incoming radia- tion, and longwave outgoing radiation at the earth's surface as follows: Climate, Snow Cover, Microclimate, and Hydrology 45 Net Radiation A 1.5 Winter 1 Pre- Melt 5.8 Melt 15.9 Post- Melt 0.0 Summer T. I Freeze -Up Heoting Heating Soil Snow Air MJ m"^ day'' 0.2 1.0 1.2 1.4 0.2 0 1.2 1 0.3 i I.I 2.9 /^ 3.2 0.9 Melting/ Evaporation 0.1 i ^t 0.3 3.5 11.6 6.6 i H 3.2 Typical Dotes I Oct - 31 May I Jun - 10 Jun II Jun -20 Jun 21 Jun -30 Jun I Jul- 31 Aug I Sep- 30 Sep FIGURE 2-8. Heat balances of the coastal tundra at Barrow for six different seasons. The width and direction of the arrows indicate magnitude and direction of energy flux. The numbers at the base of each arrow give typical rates in MJ m'^ day' (1 J = 0.239 cal). Evaporation rates are for open water. (After Weller and Holmgren 1974a.) R„ = Q(l-a) + (0i-0t) where R„ = net radiation (shortwave plus longwave) Q - incoming shortwave radiation or insolation (0.3 to 3 ^m wavelength) a - surface reflectivity (albedo) 0t = longwave outgoing radiation (2 to 100 ^m wavelength) Bi = longwave incoming radiation (2 to 100 ym wavelength). 46 S. L. Dingman et al. TABLE 2-3 Annual Radiation Balance Compo- nents over the Coastal Tundra at Barrow (MJ m'^ yr'^) Component Land Lakes Average* Incoming shortwave radiation, Q 3200 Net longwave radiation, 04 -0t -1000 Net radiation, R^ 450 Albedo, a 0.55 3200 3200 -1010 -1005 540 500 0.52 0.53 ♦Assuming about 50% of the surface is land and 50% is lakes. Source: Maykut and Church (1973). A positive sign indicates a gain of radiant energy at the surface. The energy unit used here is the joule (1 J = 0.239 cal), and balance compo- nents are expressed in MJ m'^ per unit of time. Table 2-3 gives average values of the components of the radiation balance over the coastal tundra. Insolation on the coastal tundra is markedly affected by cloud cover, which averages 68% at Barrow (Figure 2-9). Because of this, most of the shortwave radiation received is diffuse rather than direct (Maykut and Church 1973). The cloud cover increases from 40'yo in winter to 85% in summer, causing a pronounced skewness in the annual insolation curve (Figure 2-10). An effective (weighted average) albedo is about 0.55 for the tundra surface and 0.52 for lakes (Table 2-3). One of the most critical factors in 100 Clear Partly Cloudy ^40-^/10 Cloudy ^40-% FIGURE 2-9. Cloud cover conditions (bars) and transmissivity for coastal tundra. Transmissivity from Maykut and Church (1973), cloud cover from University of Alaska (1975). Climate, Snow Cover, Microclimate, and Hydrology 47 40 1 — I — I — |— 1 — I — I — I — I — I — I — I — |— I — I — I — r FIGURE 2-10. Mean ( — ), maximum (h.) and minimum (^) incoming shortwave solar radiation for five-day intervals over the coastal tundra at Barrow. The data are based on 1962 and 1964-66 from Kelley et al. (1964), Lieske and Stroschein (1968), Weaver (1969, 1970), Maykut and Church (1973), and LeDrew and Weller (1978). determining the surface climate is the large annual variation in albedo that coincides with the establishment and decay of the annual snow cover. Winter albedo values are 80*^0 to 90%, depending on variations in the character of the snow surface caused by fresh snowfalls and wind- packing. Albedo decreases in early June as a result of the progressively thinning snowpack and the appearance of bare patches (Weller and Holmgren 1974a), and drops to approximately 15% within a week or so 48 S. L. Dingman et al. 24j— I— I — I I I I I I — p — I — p— 1 — [— r— 1 — p— I — I I I I I I — [— 1 I — p~i p-p-T o c o o z FIGURE 2-11. MecTAi (—), maximum (k.) and minimum (y) net radiation for five-day intervals over the coastal tundra at Barrow. The data are based on 1962 and 1964-66 from Kelley et al. (1964), Lieske and Stros- chein (1968), Weaver (1969, 1970), Maykut and Church (1973), and LeDrew and Weller (1978). as the snow melts from the land surface. Over lakes the ice cover extends this transition to two or three weeks. During the snow-free period, the average albedo generally varies between 10 and 20^q, with large spatial variations. During freeze-up, the albedo fluctuates between 18% and 60% as snow falls and melts; the permanent snow cover forms by early October. Maykut and Church (1973) found that incoming longwave radiation for three years averaged 7440 MJ m"^ yr'', which is more than twice the annual receipt of incoming shortwave solar radiation (Table 2-3). The longwave input exceeds the shortwave in every month except April and May (Lieske and Stroschein 1968). The longwave radiation balance is at a minimum during the coldest part of the year when cloud cover is at a minimum (February and March), and maximum values are reached dur- ing the summer months when cloud cover increases and cloud tempera- tures are highest. Climate, Snow Cover, Microclimate, and Hydrology 49 The net longwave flux is negative for each of the seasons and for the year (Table 2-3). Positive net incoming longwave flux occurs only occa- sionally when warm stratus clouds are advected over a cooler tundra sur- face. The loss of longwave radiation from lakes is less than that from the tundra during June, because of residual ice and the lower temperatures of the lakes. However, the reverse is true from July to October as a result of higher temperatures due to the greater absorption of insolation. Between October and April the radiation balance is dominated by the negative net longwave radiation; near the end of May the balance be- comes positive (Figure 2-11). After snowmelt, when albedo has decreased dramatically, the net shortwave gain exceeds the net longwave loss by a factor of four (Maykut and Church 1973). Maximum average net energy receipt of about 12 MJ m"' day' occurs during the period 15-19 June. After mid- August, the absorbed radiation gradually decreases with de- creasing day length and solar altitude, becoming negative again in late September-early October. Energy Balance The partitioning of the net radiation at the surface is described by the energy-balance equation: R„ + H + L + G = 0 where R„ = net radiation H = net exchange of heat with atmosphere by conduction/ convection (sensible heat flux) L = net exchange of latent heat with atmosphere (vaporization and latent heat used in melting) G = net exchange of heat with snowpack and/or soil. Estimates of average energy-balance components for the coastal tundra at Barrow were developed from several sources, and are summar- ized in Table 2-4. Differences in the various values probably reflect real differences in the energy balance between 1957-58, 1962-66 and 1971-72, but also include variations due to different measuring techniques as well as procedures for the calculations. The energy used in melting snow is about 35 MJ m"^ yr'', assuming that a snow water equivalent of 106 mm (see below) is melted annually. It is generally assumed that there is no net heating or cooling of the ground, so the average annual value of 0 represents energy used in warming the snowpack. This assumption is consistent with the findings of Kelley and Weaver (1969) at Barrow; however, Lachenbruch and Marshall (1969) s: s: o -s: I I s: s: s: C3 ?1 > s: o o >) r^ s CQ t^ »^ C3 ^ ^ -s: ■$ s: s: o ■%* ^ t^ 0^2 $ ex, U 0Q < O 3 T3 c o '^ X ~ ^ ij P' _ HJ ^ -5 ri "=y rsi > X 5 oo ? in o — w r-- c o 03 y5 .^ r v^ c^ u. O s o £1 H ^ rs| 1 — . _aj 01) E I^ o On =« re 00 ^ * lAi ILI r 1 r r >/-) m ;- (SN 2 o x: =i^ r VO 3 o vO ^ -t rsl 03 x: U (7v 0« >> c on C ;- r n! O D 0! 'J o X) 03 '•-• c o 3 ~ C C >> Ji 00 00 S E c — o o c I •^ -o c c 00 CO 1) -= C13 to 11) '_» -E 3 §5 ~ c =« — S 3 o! j<: ^ 03 >-. ^ 00 00 ir: "^ c 2 C3 re ^ re ^ E 'i CL) L. r^ -1 11 O n » /L CA 50 Climate, Snow Cover Microclimate, and Hydrology 51 showed that average ground temperature has increased 4°C since 1850, so the assumption is not strictly true over several decades. The energy re- quired to warm the snowpack from its typical winter temperature of -30 °C to the melting point is about 7 MJ m"^ The net radiation value of 440 MJ m"^ yr"' measured by Maykut and Church (1973) is probably the most reliable mean value, since it repre- sents measurements over five years. However, it should be noted that the data of Weller and Holmgren (1974a) indicate a radiation balance of about 580 MJ m"^ yr'' and Mather and Thornthwaite's (1958) values are even higher. Most of the radiant energy available at the surface is trans- ferred to the atmosphere by conduction and convection (//). During win- ter, this exchange is positive, i.e. the air is warmer than the snow surface and the snow is warmed. Convective exchange becomes negative as the snowpack ripens in late May and early June, with the rate of energy loss increasing to a peak during the summer and declining again during freeze-up. During the winter, there is a minor net gain of latent-heat energy due to condensation, and some sublimation occurs prior to and during snow- melt. Open-water evaporation rates are estimated to be about ten times larger than transpiration rates, and the rate of evapotranspiration typi- cally peaks in late June when the tundra is still water-saturated and available radiant energy is at its peak. There is a net flow of energy from the ground to the snowpack (G) during the winter, when snow and soil temperatures are higher than sur- face temperatures. As the net radiation increases and surface tempera- tures rise in May, this flow reverses, and most of the available energy is used to warm the snow and soil. After melt, downward heat flow con- tinues as the active layer thaws and warms. Typically over 90% of the thaw occurs by mid- July (Kelley and Weaver 1969). The energy used to warm and thaw the soil in spring must, on the average, be balanced by an upward movement of latent and sensible heat during freeze-up in September. HYDROLOGY Water Balance The water-balance equation at the surface for a specified period of time is expressed as: P = E + R + I+AS where P = precipitation 52 S. L. Dingman et al. E - net evapotranspiration R - runoff / = net infiltration AS = change in surface storage of water in the area considered. As with the energy balance, it is assumed that surface storage is neither increasing nor decreasing, so that the average annual value of A5 is zero. In addition all the water that infiltrates eventually ends up as run- off or evapotranspiration, so that 1 = 0 for long-term average condi- tions. Precipitation and runoff are estimated using standard techniques. For average annual data, evapotranspiration is then found by subtrac- tion. For shorter periods, information on evapotranspiration is based on measurements of evaporation from pans or small ponds, on the energy calculated to be available for evaporation via the energy balance equa- tion, or in a few cases on observations of soil moisture (which may also be used to estimate I). Data on short-term changes in storage are in some cases available in the form of records of the changes in elevation of tun- dra ponds (Hobbie 1980). The average annual precipitation recorded (1941-70) at the Barrow National Weather Service Station is 124.1 mm (Table 2-1), but it is known that this recorded amount is less than the true amount. Compari- sons of the water equivalent of the tundra snowpack (Black 1954, Benson 1969) show that actual snowfall exceeds the amounts recorded at the Bar- row gage because of the effects of wind on the gage catches. A summary of data for six winters (Table 2-5) suggests that, on the average, the true value is 1 .6 times the recorded value. Although data have indicated that summer precipitation measurements do not need to be adjusted for wind effect on gage catch. Brown et al. (1968) found that a correction should be made for the effects of traces. Traces, recorded when precipitation is less than 0.13 mm in a measurement period, have been summed as zero values (Table 2-1). Brown et al. (1968) found that the measured summer precipitation should be muhiplied by 1.1 to give the actual value. Thus, the average annual precipitation value of 170 mm is estimated by multiplying the National Weather Service data by 1.1 for the months June through August and by 1 .6 for the other nine months, when precipi- tation is assumed to be in the form of snow (Table 2-1). The estimate of total precipitation is consistent with most detailed studies of the region's precipitation and water balance, including those of Black (1954), Mather and Thornthwaite (1958) and Brown et al. (1968). Corrected values (Table 2-1) show a precipitation maximum in August and a secondary maximum in January. On the average, about 63% (106 mm) of the an- nual precipitation falls as snow (September through May) and 37% (64 mm) as rain (June through August) at Barrow. At Barter Island, an aver- age of 68% (167 mm) falls as snow and 32% (80 mm) falls as rain. The Climate, Snow Cover, Microclimate, and Hydrology 53 TABLE 2-5 Water Equivalent of Snow on the Nearby Coastal Tundra Compared with Precipitation Records at the National Weather Service Station, Barrow, Alaska Measurements on tundra NWS gage Avg Avg Avg water 1 Sept-31 May depth density equivalent water equivalent Year (mm) (g cm') (mm) (mm) 1962-63 350 0.42 144±17 141 1969-70 170 0.36 61 ± 7 32 1970-71 350 0.36 126 ±14 41 1971-72 300 0.36 108 ±12 30 1972-73 330 0.36 119±13 110 1973-74 340 0.36 122 ±14 68 Average* 113±28 70 ±42 * ± standard deviation. correction factor of 1.6 is a minimum value for Barrow and the remain- der of the Coastal Plain since it is based on the amount of snow remain- ing on the open tundra after some has blown away and concentrated in drifts. Recent experience with Wyoming snow shields indicates that the correction factor may be about 3 rather than our conservative estimate of 1.6. Significant unfrozen zones underlie the Colville River and other large rivers (Williams 1970), and these may be conduits for substantial runoff that originates in the Brooks Range and the northern Foothills. However, because most of the Coastal Plain is underlain by permafrost that extends from a depth of about 0.5 m to depths of several hundred meters, it can be assumed that all runoff from the tundra occurs via sur- face streams. Thus, runoff may be estimated from the discharge records of streams whose drainages are confined to the Coastal Plain. Such data are limited, as regular U.S. Geological Survey stream gaging programs began in the area only in 1970. Data collected for entire water years (Oc- tober to September) and adjusted for year-to-year precipitation varia- tions indicate that average annual runoff from the Coastal Plain is about 1 10 mm. Because of measurement difficulties, discharge data are subject to uncertainties of 15 to 25%. In May, much of the evaporative heat loss is due to sublimation from the snow. During the post-melt season, the greatest proportion of total evapotranspiration is evaporation from open water, while trans- piration increases in importance in July and August. Subtracting t5« s: 5$ to CO s: s: o • ^« o s: s: (3 CO I 09 < § 1 ^ ^ ,? '— -a -a o 3 Q 03 1) ■ — . C . "*^ •■ o 3 c c ° 3 03 c 3 1.; C 0/) o C r c 3 03 r- k« oi 3 >. aj s: E O c ^_, ^M r ,^ u o S u r n* 3 CU D. C/5 O ^ •= E .t; O D. C '" u- il o ^ CU £ 03 3 "o P < c 3 E (1) u« m CO 0^ :? >% U ou _,_^ 03 m r* c aj E 03 m .:^ L- -•^ t/5 OS -^ O 00 Ov ■-D — r0 00 00 00 vD q 00 r^ m r^ v-^ u-\ tri rn r^ ^- '^ '^ '^ O r~- <^ O — ' O Tt ■^ CnI -^ so -^ o <^ <^ o o o vO rsl vO •^ »N >/~i — >/^ »N >o so CTs r^ \0 O 1^ /^ - - - ^ m sO r^ w~i OS oo r- r- OS oo so -rt so r- OS •Tj- sO so ro ■^ r- /-l t^ m OS '^ c 11 V •• o ■i >, (U a> o Di 03 K a> 00 CQ L. Urn CS .^ U 03 > 03 o 03 3 o CO 3 o uligay rudho 3 z 09 03 1/5 oa 03 oa 3 ^ Cu OS r- E c E .2 «N 03 f^ •^ C/5 .2 S — c/o 03 <-• k. CA) 1> JS il t: ^ OS 1> •> C/!) _ t" ■= .2 ■d u. w ca k. nal We land N 3 tj O O "i =« Jn 3 Z 2i u« u. 03 3 C C OS CQ ^ ^m^ \t SO (L» CO SI r- •= X. *^ Os — 00 u— — « "i 03 1 CO u. x: r^ CO x: OS ir; "^ > "7 <1> -70 a 1949- 00 c > Urn 3 C/5 Os -5 3 lo ^i^ irt 'ob T3 O >> CO o sure age ■a > 8 o E a 3 c/6 o k. OJ 3 a> a> LD x: Q. C E o C; Ji O O CO i (1> t/5 ■o CO 00 u o rcenta enthes u. S 1> a. O Q. CO Q. E to CO C o CO "S'^ j: 00 1/5 ii? lU c S .o _a> ii L i u CJ X 5 x: 3 UJ Z H O ♦ C/^ 54 Climate, Snow Cover, Microclimate, and Hydrology 55 estimated average runoff (110 mm) from estimated average precipitation (170 mm) gives a value of 60 mm yr"' for average annual evapotranspira- tion. However, this estimate absorbs the inaccuracies of the other two measurements and therefore has wide confidence intervals. In fact, other data suggest that the actual value of evapotranspiration is substantially higher. Annual Class-A pan evaporation in the Barrow area is about 160 mm (Brown et al. 1968); reducing this by a standard pan coefficient (0.6 to 0.7) gives a range of 96 to 112 mm yr"' for evaporation from a well- watered surface. The energy balance data of Weller and Holmgren (1974a) indicate evaporation rates of 4.8 mm day"' for the post-meh period and 2.7 mm day"' during the summer season; if these rates are considered average, an annual total of 210 mm is calculated. Stewart and Rouse (1976) found that daily evapotranspiration from both wet and dry tundra surfaces can be well estimated from net radiation and air temper- ature. Application of their method using typical values for Barrow sug- gests an annual total of about 140 mm. Interestingly, Stewart and Rouse (1976) found that evaporation from a relatively dry tundra surface averaged 80% of that from the wet surface (standing water) under the same temperature and radiation con- ditions. This is apparently due to the fact that, as noted in Chapter 3, on- ly 14 to 20% of the evapotranspiration from the land is due to transpira- tion from vascular plants. The remainder is evapotranspiration from mosses, which are often wetted by fog and dew and have low resistance to water loss. These considerations therefore indicate that either the estimate of regional average precipitation is too low, or the estimate of runoff is too high, or both. It is likely that failure to account for occult precipitation (fog, dew) is a significant source of error. In any case, it is important to reahze that substantial uncertainties remain in our understanding of arc- tic water balances, even in regions of relatively intensive study. Runoff is concentrated into a short period of time (Table 2-6). Al- though the data are limited, there is a definite suggestion that runoff is more time-concentrated in larger drainage basins. This is the opposite of what would normally be expected, and may be due to the formation and breakage of ice jams on the large streams. Actual data on infiltration are very limited, but it is possible to infer the general nature of the intra-annual variation. In winter, an upward moisture gradient is established, so that there is exfikration in the form of vapor for much of the year. During and immediately after melt, water infiltrates in liquid form, to the extent that soil moisture capacity is virtu- ally reached. Through the summer, most of the water falling as precipita- tion infiltrates, and most of this is subsequently evaporated and transpired. Surface storage increases through the winter as snow accumulates 56 S. L, Dingman, et al. and water vapor transported upward from the soil condenses in the snowpack. Most of the snow is depleted within the few days of the melt period, but part of the snow is converted to liquid water storage in pud- dles, ponds and lakes. This storage, in turn, is gradually reduced by evaporation through the summer. Persistent snowdrifts in stream valleys may contribute small amounts of stream flow well after the general melt is completed. Thermal and Hydrologic Processes During Snowmelt As the radiation balance becomes positive around 1 June, the snow- pack warms or ripens and the underlying soil begins to warm. Incoming shortwave solar radiation reaches its maximum values in May, as atmos- pheric transmissivity is high. However, albedo still exceeds 80%, so that most of this radiation is reflected. About 60% of the available radiant energy is used to warm the snowpack and soil (Figure 2-8). Snowmelt begins at the surface when air temperatures rise above 0°C. Heat is trans- ferred downward in the snow by conduction and as sensible and latent heat associated with liquid water movement. It appears that the latter process is responsible for much of the warming of the snowpack, and also contributes to warming of the soil. The percolating meltwater re- freezes in the colder snow, liberating latent heat and forming a complex network of ice glands, lenses and layers. Benson et al. (1975) calculated that about 1.9 MJ m'^ of heat is transported downward for each I cm of ice thickness formed. With the estimated cold content at this time of 2.3 MJ m"^ the formation of ice layers totaUng a little over I cm thick would suffice to warm the snow to 0°C. Weller et al. (1972) described the 1971 melt in the coastal tundra at Barrow, and reported that melting con- verted an 8 °C temperature gradient across the snowpack to near isother- mal conditions within 2 to 3 days. The soil temperature also rose steeply during this period. Initially, the ice layers tend to form at the top of the depth hoar layer, but as the pre-melt season progresses they are found at the base of the snowpack (Benson et al. 1975). The two-layer structure thus breaks down and the density of the pack becomes vertically uniform. Typically, the density of ripe snow is between 0.4 and 0.5 g cm''. The disappearance of the depth hoar layer and the flooding of low areas as melt progresses are major environmental changes for lemmings. They are forced out of the protective snow cover and become subject to environmental extremes and avian predation, with consequent high mortality (Bunnell et al. 1975). The changes in water content and density of the snowpack during the pre-melt season cause the albedo to decrease from its winter value of Climate, Snow Cover, Microclimate, and Hydrology 57 about 85*^0 to about 75% just before the melt season begins (Maykut and Church 1973). Virtually every aspect of the surface environment changes dramati- cally during the brief melt season. These changes are rapid because of the relatively high insolation rates during the long days and the operation of positive feedback loops affecting the radiation and energy balances. The energy available at the surface increases by a factor of 3.6, and about 60% of this is used in melting (Figure 2-8). Once the snowpack is isothermal at 0°C, the further addition of energy produces meltwater that does not refreeze. Initially, this water fills voids in the snowpack and reduces the albedo, typically to values near 50%. Reduction in the albedo increases the absorption of solar radi- ation, which increases the melt rate. Snowpacks can hold about 5% of their water equivalent as liquid water (Anderson and Crawford 1964). For the coastal tundra this would amount to about 5 mm of water, which can be produced in a period of a few hours at the rates at which radiant energy enters the snowpack dur- ing melt. Once this capacity is filled, the snowpack is ripe and further melt produces runoff at the snow/ground interface or over ice lenses and layers. As meltwater accumulates in low-lying areas and produces slush, the albedo is further reduced and the melting accelerated. Snow-free patches generally appear within a day or so of the onset of melting and initiate the operation of another positive feedback loop. The albedo of the exposed areas is 10 to 15%, so they absorb four to five times as much radiant energy as the snow-covered ground. This addi- tional energy produces local heating of the air and local advection of heat to the surrounding snow, which further accelerates melting (Weller et al. 1972, Weller and Holmgren 1974a). The upper layer of the soil generally begins to thaw a few days be- fore snowmelt is complete. This layer typically has been desiccated by loss of water to the snowpack during the winter, so that some infiltration of snowmelt water occurs. Data of Guymon (1976) indicated that this in- filtration was most significant in poorly drained areas. However, most of the snowmelt water runs off to streams and lakes after ponds and poly- gon troughs are filled. A large fraction of the total annual runoff occurs within a few days (Table 2-6). The spring runoff sequence on the Coastal Plain has been described by Johnson and Kistner (1967), Lewellen (1972), Holmgren et al. (1975) and Hobbie (1980), and for the through-flowing Colville River by Arnborg et al. (1966). In stream channels, the first flow is over ice that is frozen fast to the bottom. The sediment load of this initial runoff is very low. But as the flow increases toward its peak, the channel ice is eroded and melts free from the bottom, generally dislodging large amounts of sediment. In the larger rivers, ice jams frequently occur, 58 S. L. Dingman et al. damming flow until sufficient head builds up to dislodge them and cause sudden catastrophic flooding downstream. There is considerable bank erosion at such times, due to the thermal/mechanical action of the water and the mechanical action of the ice masses. Loss of the winter snow cover proceeds from the Brooks Range northward across the Foothills and Coastal Plain. As indicated on the satellite view of the eastern portion of the Arctic Slope (Figure 2-12), the major valleys of the Brooks Range are seen to be more or less snow-free. But close inspection reveals that many of the gullies on the mountain- sides are filled with snowdrifts that extend to the valley bottoms. In the Foothills the ridges melt out first, leaving snow in the gullies. Meltwater collects in the larger valleys, reducing albedos and accelerating melting there. Thus the major rivers have developed, or are in the process of de- veloping, continuous open-water streams and generally appear as dark bands, probably because of flooded areas and because melting is further advanced. Many lakes appear darker than their surroundings, also be- cause of standing water on the ice. The larger rivers flood their deltas and the sea ice, forming large overflow plumes. Attempts have been made to model the snowmelt runoff process on the Coastal Plain for watersheds ranging in size from 3.8 km^ to 13,890 km^ (Carlson et al. 1974, Dingman, unpubl.). The models used have con- sisted of a snowmelt generator driven by climatic input and a simple stor- age model to transform snowmelt input to streamflow output. The basic form of the storage model is: q. = KiS-So)" where q, - runoff during period t K and n = storage parameters S, = snowmelt during period t So = the amount of melt that is absorbed into "dead" storage (filling lakes, ponds, troughs and soil pores). Using a simple snowmelt model, Dingman (unpubl.) accounted for melt due to absorption of shortwave radiation only. Hourly melt was routed through the storage model to simulate measured runoff in Esatkuat Creek near Barrow (drainage area = 3.8 km^). The parameters K and n were estimated from examination of the measured runoff hydrograph for the area. The value of n was taken as unity, so the stor- age model was effectively a linear reservoir. When modification was made to account for the irregular distribution of snow depths over the watershed, the model appeared to be quite successful (Figure 2-13). Carlson et al. (1974) simulated snowmeh from three large Arctic Slope rivers— the Putuligayuk (456 km^, the Kuparuk (9210 km^), and Climate, Snow Cover, Microclimate, and Hydrology 59 FIGURE 2-12. Landsat satellite mosaic of a portion of the Arctic Slope for the period 27 May to 6 June 1973. Note melting in the river valleys and the meltwater plumes of the Sagavanirktok (A), Kuparuk (B) and Colville (C) Rivers. Early snowmelt induced by the Prudhoe Bay road network can be seen between the Sagavanirktok and Kuparuk Rivers. 60 S. L. Dingman et al. 2.0 o 0.8 0.4 June 1972 FIGURE 2-13. Simulated and measured snow runoff in Esatkuat Creek near the village of Barrow, 13-20 June 1972. (Dingman, unpubl.) TABLE 2-7 Values of Model Parameters for Four Arctic Coastal Plain Watersheds Area (km^) Year K (hr-') L (days) Esatkuat Creek 3.8 1972 0.088 0 Putuligayuk River 456 1970 1971 0.014 0.014 3 4 Kuparuk River 9,210 1970 1971 0.012 0.008 2 2 Sagavanirktok River 13,980 1971 0.008 8 the Sagavanirktok (13,890 km^). Their snowmelt model included energy exchanges due to solar radiation, convection, longwave radiation and other modes. The calculated melt was routed through a linear storage reservoir {n = 1), and then delayed for a specified lag period. Values of the parameter K and the lag period L (Table 2-7) were determined by an optimization procedure to reproduce the runoff measured at the gaging station. The values of K decrease with watershed size, and the values of L Climate, Snow Cover, Microclimate, and Hydrology 61 tend to increase, as would be expected. Based on energy considerations, Weller and Holmgren (1974a) con- cluded that evaporation accounted for only 2% of the total ablation when the snow cover was still complete, but up to 13% when bare patches appeared. They noted also that condensation on the snow surface was likely when air temperature rose above 0°C. Johnson and Kistner's (1967) measured pan evaporation rates of up to 0.47 mm hr"' at midday during snowmelt near Meade River indicate that open-water evaporation begins to become important at this time. Post-melt, Summer Hydrology, and Related Processes Immediately following snowmelt, the coastal tundra is largely cov- ered with water. However, snowdrifts remain in river channels and most lakes are still ice-covered. The albedo of the surface is 10-15%, causing the net radiation to jump almost an order of magnitude from its pre-melt value, and all energy-balance components are at or near their maxima (Figure 2-8). Over the ice-covered lakes the dramatic reduction of albedo extends over a period of two to three weeks. Once the ice cover disap- pears, the lakes have a somewhat lower albedo than the land surface and absorb more solar radiation. During the post-melt season, over 70% of the available energy is utilized in evaporating the extensive surface water. Water balance con- siderations and data on evapotranspiration during the summer season (see below) suggest that an average of 10 mm (= 1 mm day') evaporates during this 10-day period. This value is consistent with the average pan evaporation rates of 2.77 mm day"' and 1 .72 mm day"' measured by Mil- ler et al. (1980) in late June of 1972 and 1973, respectively, if a pan coef- ficient is applied and if less than 100% of the surface is considered to be evaporating. Weller and Holmgren (1974a) estimated considerably high- er rates, 4.2 to 4.6 mm day', on the basis of energy considerations alone. Warming of the soil begins during the pre-melt period, but there is no significant thawing until the post-melt season. Thaw is very rapid ini- tially, and can be expressed by an equation of the type applied to a rela- tively dry upland soil by Kelley and Weaver (1969): Z = L[\ - e\pi-at)] where Z = depth of thaw (cm) / = time in days a = empirical constant (1970 value = 0.067 day"'; 1971 value = 0.047 day"; 1972 value = 0.082 day"') 62 S. L. Dingman et al. L = empirical constant = maximum thaw depth (35 cm in 1970; 37.5 cm in 1971; 63.3 cm in 1972). Note that there is considerable year-to-year variation in the thaw pro- gression; this variation can be correlated with cumulative net radiation during the early summer. There is also marked spatial variation in thaw progression due to soil type. As the summer season begins in late June or early July, net radiation decreases. This is due to the passing of the summer solstice, an increase in cloudiness (Figure 2-9), and an increase of the albedo of the tundra to an average of about 19% as the surface dries. Evaporation still consumes the largest portion of the available energy, but convective heat loss in- creases in importance (Figure 2-8). Diurnal variations in soil temperature are greatest during this period as a result of strong diurnal changes in ra- diation under snow-free conditions (Kelley and Weaver 1969). Summer soil temperatures vary across microtopographic positions and during the summer these differences reflect variations in albedo, mi- croclimate and soil properties, particularly those related to moisture con- tent. Elevated rims of low-centered polygons are sometimes cooler due to wind, while troughs and basins are warmer. However, at other times, in- creased evaporation and transpiration from the more vegetated and/or wetter troughs may result in lower temperatures there (Figure 2-14). Maximum differences occur under clear sky conditions during early af- ternoon with 8°C differences observed at the 1-cm depth between cooler rims and warmer basins and troughs of low-centered polygons (Goodwin 1976). Nighttime difference decreases 2 to 3°C. Although diurnal and seasonal soil temperatures follow closely changes in air temperature, other climatic factors such as cloud cover modify the magnitude of the difference between them. For example, soil temperatures at 1 cm depth for rims, troughs and basins at site 4 averaged 8.7 °C in July 1972 and 5.6°C in July 1973. Average monthly air temperatures for July 1972 were only 1.8°C higher than 1973 (Table 2-1). Increased radiational warming accounted for most of the increased soil temperature. Evapotranspiration rates decrease from the post-melt season be- cause of the decrease in water on the surface and the decrease in available energy (Weller and Holmgren 1974a). Studies have consistently shown a near balance of precipitation with evapotranspiration during the summer (Mather and Thornthwaite 1958, Brown et al. 1968, Guymon 1976). About 80% of the annual evapotranspiration occurs during the 1 July-31 August season. Comparisons of total evapotranspiration and pan evap- oration indicate that water losses from moist vegetated surfaces are ap- proximately the same as those from open water. Koranda et al. (1978) measured overall loss rates from soil as 4.6 to 5.6 mm day', compared with open-pan evaporation of 3 mm day '. Based on Miller et al. (1976) Climate, Snow Cover, Microclimate, and Hydrology 63 12 o _ ^ 8 o. I 4 1 1 - lOcm 1 1 1 1 1 1 1 1 1 ' 1 ' - — ^ • \ r^ — y ^^ =^ -^^^^ — - ^ 1 1 1 1 1 1 1 . 1 1 1 1 — 20 10 20 10 20 Jun Jul Aug FIGURE 2-14. Seasonal course (5-day means) of 1972 summer soil temperature on the trough, rim and basin of a low-centered polygon at I cm and 10 cm depth. (After Bunnell et al. 1975.) a calculated evaporation rate for a standard day (12 July 1973) was 2.3 mm, while average pan evaporation rate was 2.3 mm day"' for 10 to 16 July. In all cases, transpiration made up only a small fraction (7% to 15%) of total water loss. These observations are consistent with those of Rouse et al. (1977) who found evapotranspiration from a shallow tundra lake and a wet sedge tundra essentially identical. Soil temperature-moisture studies (Guymon 1976) showed negli- gible vertical water movement in the mineral soil during the summer. This suggests that virtually all rain falling on the dry tundra infiltrates only into the surface organic layer and is subsequently evaporated and transpired. Runoff originates from rain falling directly on ponds and streams and on adjacent low areas where the water table is at or near the surface. The runoff data of Brown et al. (1968) show zero flow during ex- tended periods of no rainfall, indicating that the thawing active layer is not a source of stream flow. Similarly, records of pond levels (Brown et al. 1968, Hobbie 1980), pond chemistry (Brown et al. 1968), and lake levels (Kane and Carlson 1973) indicate that in the absence of inlet and 64 S. L. Dingman et al. outlet streams, changes in their volume during the summer are caused solely by rainfall and evaporation. The freeze-up season marks the transition from summer to winter. By the time of freeze-up, towards the end of September, the net radiation has decreased substantially (Figure 2-11) as a result of much lower solar elevation and greatly reduced duration of daylight — 13.5 hours in mid- September compared with 24 hours two months earlier (Table 1-1). Light snowfalls, which generally melt, may temporarily reduce the net radia- tion further by increasing the albedo of the tundra; the albedo fluctuates between 18 and 60*^0 before the establishment of the "permanent" winter snowpack. The bulk of the available radiation energy is used in melting these snowfalls, but typically little or no runoff results. The other major physical process of this season is the freezing of the thawed soil layer. As freezing progresses downward, and occasionally upward, a steadily increasing slab or sandwich of soil remains isothermal as the latent heat of fusion is being extracted (Brewer 1958, Nakano and Brown 1972). The result is the zero curtain or the period during which temperatures at a given depth remain at the freezing point. Once the soil is totally frozen the cold wave can penetrate into the permafrost. Diurnal variations of the surface soil temperature decrease because variations in air temperature and insolation are smaller and snow depth is increasing. SUMMARY Data collected prior to and during the Tundra Biome program pro- vide a reasonably complete and consistent picture of the climate, micro- climate and hydrology of the coastal tundra of northern Alaska. The average net radiation at the surface is between 420 and 450 MJ m"' yr"'. Of this, 55% is sensible heat transferred to the air, 36*^0 is used in evapo- transpiration, 1% is used to melt snow, and 2% is sensible heat trans- ferred downward to snow and soil. Two-thirds of the year is characterized by a negative net radiation balance, very low surface temperatures, and a gradually increasing snow- pack subject to substantial drifting. The snow reduces extremes of temp- erature and wind at the ground surface, providing a more moderate microclimate for surface- and near-surface-dwelling organisms. When the net radiation balance becomes positive in late May, the snowpack, upper soil, and air temperatures approach the freezing point. Surface melting of the snow redistributes water and heat downward, causing the first in a series of rapid changes in the immediate surface en- vironment. Profound changes occur over the few days when the snow- pack melts and the upper layers of the soil thaw. During this time there is a rapid increase in net radiation, which is accelerated by decreasing snow- Climate, Snow Cover, Microclimate, and Hydrology 65 pack albedo and then by the absorption of heat by the exposed ground surface. In a few days, the snowpack has disappeared except for larger drifts. Most of the meltwater runs off, and streams are in flood condition as 50% or more of the annual flow volume is discharged in a few days. The ground is covered with extensive areas of shallow surface water. Net radiation is at a maximum during this post-melt period, and most of the energy is used in evaporation. The short summer season is subject to relatively small interannual variability in temperature but large variability in precipitation. About two-thirds of the net radiation is used in evapotranspiration, and the rest in heating the air. The depth of thaw in the soil approaches its maximum near the beginning of August. There is a near balance between precipita- tion and evapotranspiration, but significant runoff may occur in wetter years. Only about 10<^o of the evapotranspiration is transpiration. The remainder is evaporation from soil and interception from plant surfaces. Significant evaporation takes place from lakes and ponds, as well. Net radiation is still positive but small in September. Air tempera- tures drop consistently below 0°C and the soil begins to freeze from below and above. Precipitation is largely in the form of snow, but inter- mittent melting often occurs. By the end of September, the net radiation becomes negative, the soil active layer may be completely frozen, and the permanent snowpack is becoming established. Biophysical Processes and Primary Production p. C. Miller, P. J. Webber, W. C. Oechel, and L. L. Tieszen INTRODUCTION During most of the year the arctic tundra is covered with snow, and the exchange of heat at the earth's surface through radiation, convection and evaporation involves only physical components of the environment. However, during the short snow-free period the vegetation becomes a significant exchange surface. The vegetation influences the partitioning of incoming energy into evaporation, convection and soil heat conduc- tion. The vegetation also accumulates the biomass on which the plants themselves and the other ecosystem components depend for energy. This chapter discusses the rates of primary production, the standing crop, and the partitioning of incoming energy by the vegetation. The interactions between diverse environmental factors and specific canopy and plant properties affecting plant temperatures are integrated in the energy budget equation (Gates 1962, 1965, Parkhurst and Loucks 1972, and others): Q^ + Oa = d+G + H + LE + M where Q„ = incoming shortwave radiation absorbed by the plant da = longwave radiation absorbed by the plant d = longwave radiation emitted by the plant G = net heat flux into the soil or moss layer H = heat exchanged by convection LE = evaporative loss of energy by transpiration from vascular plants or the evaporative loss of energy from moss sur- faces M = metabolic term to account for energy used in photosyn- thesis or produced in respiration. 66 Biophysical Processes and Primary Production 67 The energy exchange processes are significant in determining the rates of primary production, the temperatures of the plants and soil, and the rates of water loss from vascular plants and mosses. The metabolic term is small relative to the total energy exchanged by physical processes, but is important in maintaining all biological processes. The effect of the energy exchange processes is mainly through absorbed solar energy for photosynthesis and through influences on plant temperatures. Tempera- tures at the Biome research area, even during the growing season, are almost always below the optimum for physiological processes in most plants. Therefore, physiological adaptation of plants to low tempera- tures, and morphological adaptation that increases plant temperature, should be more evident in the vegetation of the coastal tundra at Barrow than in temperate regions. PRIMARY PRODUCTION IN THE BARROW WET MEADOW TUNDRA Standing Crop and Primary Productivity Primary productivity in the tundra at Barrow has been estimated by the harvest method for the aboveground vascular vegetation (Webber 1978), by cuvette measurements for vascular plants (Tieszen 1975, 1978b) and mosses (Oechel and Sveinbjornsson 1978), by photosynthesis models for vascular plants (Miller et al. 1976) and mosses (Miller et al. 1978a), and by the aerodynamic method for the total ecosystem (Coyne and Kelley 1975). Based on these different measurements, gross above- and belowground primary production for the Biome research area in 1972, a year of nearly normal temperature and precipitation, was 465 gdw m"\ including 358 gdw m"^ for vascular plants, 106 gdw m'^ for mosses, and about 1 gdw m"^ for lichens. Net primary productivity, which is gross productivity minus the esti- mated respiratory costs for plant maintenance and growth, was about 230 gdw m"^ yr"', including 162 gdw m"^ yr"' for vascular plants, 66 gdw m"^ yr"' for mosses, and less than 1 gdw m'^ yr"' for lichens. For these calculations growth respiration was calculated as 0.3 of the new biomass for both vascular plants and mosses. Maintenance respiration was calcu- lated as 0.0054 gdw gdw"' day"' for 35 days for aboveground vascular material, 0.0027 gdw gdw"' day"' for 60 days for belowground vascular material, and 0.003 gdw gdw"' day"' for 60 days for mosses. The daily maintenance cost was calculated using protein percentages of 9, 4.5 and 4.8 for above- and belowground vascular and moss material respectively (Penning de Vries 1974, Chapter 5). The overall value is composed of the separate productivities of different plant growth forms in eight vegetation to s: <3 'Si. ^ ?3 s: I O a, J4J to O s: Go O B s: o o n ^^ *" tn M S It CO 2 o 00 o « u •a c to > m < .g C 5 ? o ^^ ^ T3 ~ ■s: > s E o :^ 00 ■5 "I _ ' 5 * 6 =^0 i-5 = ^^ e2 " f ~ r*-i r*^ \0 rs| O \0 ^ — r- r- p*-i — o — fN — — fN w^ — m ^- rsj f^ I 00 O •* > rvj rsj o ^ < — O 00 r- r.-, O VO 3 m r~ ■* — Tt o f^ -3 • (N 1^ — — »/-i O 00 IN »/-i »/^ rn *o rn o^ r^ ^ — vO r-~ o\ vo — 00 Tt 1/1 >/-l — . r*-t ■ — w-i rNl 00 — O O^ •* 1 ■»■ > 0^ o CTs r- oosC— oo*^— • — r-oor-v-1 ^ r*^ rn OS r^ 1^ ^ »^ <^ rsi vo ON — O — — (N ^ 5 I U flQ < c "o. 3 O T3 ^.2-0 ;S C 3 c o ^ (0 C-Q. l/l > ■o - c o to 3 2^ 3 O 00 E o ZZ o i; u 00 ■t3 ? X> O to -O o :5 ,0 15 H ^ 2 u c to 3 O 3 o CO b> o o fc ^ o C 3 O 00 (^ > u o ^ g = is CO D-D. _C0 C5 O > > &2 o J- > > CA — o o s _C0 [5 a-Q. ^ JO 1/) ■= =* > ^ > 13 -a CO G § = = 3 S -^ t-" ^ *- u, CO O 00 00 > 1/5 u- OJ > OJ o ^ > — c/) -— CO ul CO — ^ * o o ''^ "^ 2 -o-S 55 O u to .:- — ;/! > 1» 0* ?*^ W a> 4> = 00 — tU io •■5 Si c 3 •- o •t- 00 68 Biophysical Processes and Primary Production 69 types recognized at Barrow, weighted by the surface area of each type (Table 3-1). Net aboveground primary production of vascular plants at Barrow in 1972, estimated by the harvest method and averaged according to the relative surface area of the different vegetation types, was 42 gdw m"^ (Webber 1978, Table 3-1). Aboveground production of vascular plants ranged between 18 gdw m"^ for Luzula heath and 115 gdw m"^ for Arcto- phila pond margins. Eighty-four percent of the area had aboveground vascular plant production of 39 to 51 gdw m"^ Although the average productivity within a tundra region is frequently near the low end of the range of productivities of the vegetation types in that region (Beschel 1970, Webber 1971), the average in the coastal tundra at Barrow is near the middle of the range for the vegetation types near Barrow because this area is dominated by reasonably productive vegetation types. The least productive vegetation type, the Luzula heath, occupies 3% of the entire area. Aboveground vascular plant productivity increased along a mois- ture gradient from the tops of high-centered polygons to pond and stream margins. The wetter areas have reduced soils and moderately high phos- phate levels and are dominated by graminoids. The productivity of forbs was highest in dry areas with moderate levels of phosphate and more ox- idized soils, but productivity rarely exceeded 6 gdw m"^ yr"'. The produc- tivity of woody dicotyledons was also highest in dry oxidized areas with moderate levels of phosphate, and was about 20 gdw m"^ yr'. Bryophyte productivity, calculated as 56% of the green biomass or about 66 gdw m'^ yr"' (Oechel and Sveinbjbrnsson 1978), differed widely in the differ- ent vegetation types, with the highest rates in the mesic meadow where many acrocarpous mosses such as Dicranum elongatum and Pogonatum alpinum are abundant. The principal factor controlling the distribution of bryophytes appears to be slight differences in microrelief, moderated by the vascular plant canopy, which influence the moss and soil moisture regimes (Webber 1978, Oechel and Sveinbjornsson 1978, Stoner et al. 1978b). Belowground productivity estimated from the belowground biomass (Webber 1978, Table 3-2) and from longevities of belowground plant parts (Shaver and BiUings 1975, Billings et al. 1978, Chapter 5) was 120 gdw m'^ yr"'. The belowground productivities ranged between 47 and 217 gdw m'^ yr"' from the Luzula and Salix heaths to the Dupontia meadow. The ratio of above- to belowground productivities varied from about 1:1 to 1:4 and averaged 1:2.9. Dennis (1977), using a regression approach, estimated that belowground productivity was 143 gdw m'^ yr"' in 1971. Total standing crop of live and intact dead plant material was 5292 gdw m"^ weighted by the relative area of the different vegetation types (Table 3-2). The standing crop was dominated by graminoids, bryophytes 1^ s: o t 60 s: CO 60 ^ 2 03 -^ ■ QQ < c2 « s ^^ •- rt > y, es to o ^ ^ g oa < •= ~ |o- ■3; -o = ■s: IS) ^ c '5b "J _ I? c o a 2 ^ >. ^ b 00 \0 2 UJ 2 S c o VO a^ o "2 > ^ lo 3 ^|o >0 On l^« > d .ts o S P~l ^ S E O a ^? — o 1 " "™ 5- ,Q E O ^ 3 Tj- \o */"> r- oo >/^ O — »i-^ — O O On 00 o (N r ^ — „^ o o r- — in ■* ^^ »o On r- (N m > S -5 a. < 1/5 '2 = c O CO ;§ a 8 = >/> ._ 'J o ^ s ?. >. > £ T3 _ Q. O CO O O S =^ ^ |_ GQ C a ■a c 3 o u CO 00 ^ > K . > T3 § ^ § iS - C CO o XI -c "to "ra OO P (U 1) ^ "T* < rr -i ^ c/l CO > C 3 O CO — CO 1- o c/> 00'- 00 > «. -^ ^ o CO £ i_ o u E -2 ^ 2 f .2 m iS 2 -^ ij E O CO S > o ra o 4/ "O ■a CO to — •T3 to 3 3 00 C O O C CO o ,-s o o ^ - I- h- -ii c/l -J oa 1/ "* o - > o °o 0"^ u 1) u p i: > > CO — „ _Q _Q c/o _i — Q < < CD b- u 00 •-r n CO OO > _3 CO > IN o -o p 3 O SO EC/) -, o to j: E S ° s T3 — O >- -■§ •5 ni CO 0/ CO X) S .12 ^ CO ON E C o ^ 00 XI •T3 X3 o ■* * i! ° 3 l> o CO 1/2 70 Biophysical Processes and Primary Production 71 and dead material. At peak season the average dry weight of above- ground plant material was: graminoid biomass 35 gdw m'\ forb biomass 4 g m'\ woody dicotyledon biomass 16 g m"^ bryophyte biomass 117 g m'\ lichen biomass 28 g m'\ vascular standing dead 36 g m"^ vascular litter and prostrate dead 91 g m"^ and belowground intact dead 4116 g m'^ The aboveground vascular standing crop averaged 54 gdw m"^ and ranged between 19 and 119 gdw m"^ depending on the vegetation type. Primary productivity in the Carex-Oncophorus meadow vegetation type (Biome research site 2) was estimated from cuvette photosynthesis measurements on vascular plants and mosses, by canopy photosynthesis models for vascular plants and mosses, and by the aerodynamic method. In this vegetation type, gross primary productivity, estimated by the har- vest method and by respiration costs, was 450 gdw m"^ yr"', including 414 gdw m"^ yr"' for vascular plants and 36 gdw m~^ yr"' for mosses. Net pri- mary productivity was 209 gdw m"^ yr"', including 187 gdw m"^ yr"' for vascular plants and 22 gdw m"^ yr"' for mosses. The respiratory cost for growth of above- and belowground tissues was 57 gdw m"^ yr"', and the respiratory cost for maintaining these tissues was 170 gdw m"^ yr"'. The growth and maintenance costs for mosses were 6.6 gdw m"^ yr"' and 7.2 gdw m"^ yr"', respectively. The gross primary productivity for vascular plants of 414 gdw m"^ yr"' was equivalent to a carbon dioxide uptake of 609 g CO: m'^ yr"'. The conversion was made using glucose as a base to be consistent with the calculation of growth respiration (Penning de Vries 1974). Based on his cuvette measurements, Tieszen (1975) estimated carbon dioxide up- take at 602 g CO2 m"^ yr"'. Using the canopy photosynthesis model Miller et al. (1976) estimated 610 g CO2 m'^ yr"'. The simulated gross primary productivity for mosses was equivalent to a carbon dioxide uptake of 53 g CO2 m"^ yr"' (Miller et al. 1978a), which is only slightly lower than the cuvette measurement of Oechel and Sveinbjbrnsson (1978) of 57 g CO2 m"^ yr"'. The gross primary productivity for the community of vascular plants and mosses was 667 g CO2 m"^ yr"', which was similar to the estimate by the aerodynamic method (Coyne and Kelley 1975) of 632 g CO2 m"^ yr"'. The general agreement of these estimates for Biome re- search site 2 gives support to the calculations for the other vegetation types and for the coastal tundra at Barrow as a whole. Seasonal Progression of Primary Productivity The seasonal progression of primary productivity was estimated in the Carex-Oncophorus meadow (site 2) by periodic harvests of above- ground material, periodic photosynthesis measurements in cuvettes, simulations based on photosynthesis, light and temperature relations, 72 P. C. Miller et al. and the aerodynamic method (Dennis 1968, Tieszen 1972a, b, 1975, 1978b, Coyne and Kelley 1975, Miller et al. 1976, Dennis et al. 1978, Oechel and Sveinbjornsson 1978). Estimates of carbon dioxide incorporation with the cuvette and the simulation model showed a constant increase in carbon dioxide uptake from the beginning of the season to mid-July (Figure 3-1). The above- ground harvests for the first 30 days of the season showed a constant rate of carbon incorporation that was approximately equal to the initial rates estimated by the cuvette and by the model (Figure 3-1). During the re- mainder of the season the rate of carbon dioxide uptake by vascular plant tops declined. The comparison indicated that during the first 30 days, when net photosynthesis by the canopy was increasing, the rate of photosynthate allocation to aboveground biomass productivity was con- stant and the allocation to belowground parts was increasing. In spite of this allocation to belowground parts, the weight of the belowground parts decreased because of respiratory costs associated with maintenance (Chapter 5). During the second half of the season, when green tissues were gradually senescing, both net photosynthesis by the canopy and aboveground production rates were decreasing at about equal rates, and allocation to belowground parts continued. During this period below- ground weights were increasing (Chapter 5). Net photosynthesis in the latter part of the season remained greater than the rate required to main- tain aboveground biomass at peak season levels. The reduction in aboveground biomass must have been triggered by intrinsic controls. By 4 August, the standing crop of live aboveground biomass had begun to decline from the peak season level. Aboveground production was becom- ing negative, indicating mobihzation of aboveground organic and/or in- organic nutrients and translocation of these nutrients to belowground parts. Mosses show a more or less constant rate of seasonal CO2 incor- poration (Oechel and Sveinbjornsson 1978). High early season carbon uptake makes mosses active at a time when vascular plants are not highly productive (Figure 3-1, Miller et al. 1978a). The seasonal course of atmospheric carbon dioxide flux, estimated by the aerodynamic method, showed a fairly constant rate of carbon di- oxide removal from the atmosphere during the first 20 days of the sea- son, indicating that most of the carbon dioxide incorporated in net pho- tosynthesis was counterbalanced by plant root and soil respiration. From about 10 July to 25 July, atmospheric carbon dioxide flux increased, sug- gesting that respiratory sources of carbon dioxide were insufficient to maintain the observed increase in net photosynthesis. During August, when net photosynthesis declined, ecosystem respiration sources gradu- ally assumed greater importance and carbon dioxide flux from the at- mosphere declined. Net photosynthesis, and consequently primary production, was lim- Biophysical Processes and Primary Production 73 14-24 24Jun 4-14 14-24 24Jul 3-13 13-23 23Aug Jun 4 Jul Jul Jul 3 Aug Aug Aug 2 Sep FIGURE 3-1. Seasonal course of carbon incorporation in the moist meadow tundra estimated by independent approaches during 1971 or 1973. A — vascular plant carbon incorporation estimated by a canopy simulation model (Miller et al. 1976). B — vascular plant carbon incorporation estimated from cuvette measurement of pho- tosynthesis, corrected for shading (Tieszen 1975). C — community net carbon incorporation estimated by the aerodynamic method (Coyne and Kelley 1975). D — vascular plant carbon incorporation estimated from periodic harvests (Tieszen 1975). E — moss carbon incorporation estimated from simulation model for 1973 (Miller et al. 1978a). F^green moss carbon incorporation estimated from cu- vette measurements of photosynthesis for 1973 season. (After Oechel and Sveinbjornsson 1978.) 74 P. C. Miller et al. c 0 •- 20 - FIGURE 3-2. Seasonal courses in 1971 of incoming solar irradiance, 300 to 3000 nm (B) and the ratio of simulated net photosynthesis to incoming irradiance (A), the ratio of net photosynthesis to intercepted irradiance (A '), and the ratio of intercepted to incoming irradiance (C). Photosynthesis was calculated in terms of energy units, and the ratios are dimensionless. ited strongly by the availability of photosynthetic tissue and intercep- tion and absorption of solar radiation throughout most of the season (Figure 3-2). Carboxylation data (Chapter 4) suggested a maximum photosynthetic capability around mid-July, which, combined with a maximum live foliage area index around the first day of August, should have given peak community photosynthesis around or preceding the first day of August. This trend was predicted by the simulations. VERTICAL DISTRIBUTION OF BIOMASS AND CANOPY STRUCTURE The canopy structure affects the microclimate and soil temperature by intercepting and emitting radiation and decreasing the vertical trans- Biophysical Processes and Primary Production 75 port of heat and water vapor. Soil surface temperatures affect the air temperatures immediately above the surface and the deeper soil tem- peratures. Thus there is an interacting system comprising 1) the vertical profiles of leaf area index, leaf inchnation, leaf width, leaf absorptance, leaf conductance to water loss, and stem area index; 2) the absorptance of the soil and the properties of the soil affecting heat conduction; 3) the profiles of the processes of energy exchange; and 4) the profiles of plant temperature. During the course of the growing season, the foliage area index of the canopy of the Carex-Oncophorus meadow develops from zero at the beginning to between 0.8 and 1.2 by the last half of July (Caldwell et al. 1974). (The foliage area index was calculated as the total of the leaf and stem area indices, m^ plant surface per m^ ground, one side of the leaf considered.) In years without intensive lemming grazing the standing dead material with a foliage area index of 0.3 to 0.5 was present throughout the season but was most conspicuous early in the season. Lit- ter with an area index of 0.3 to 0.5 was concentrated within 2 cm above the moss surface. Standing dead material and litter included dead mate- rial in various stages of decay from several previous years. From mid- July to the end of the growing season leaf material produced in the cur- rent year senesced and was added to the crop of standing material. New leaves grow from the stem base, located in the moss layer, and must grow through the shade cast by the dead material. From above, the appearance and albedo of the canopy were dominated by the light brown dead mate- rial until early June when the darker green live material began to pre- dominate in the upper levels. However, the conversion from light to dark color was not complete because the current year's growth senesces and turns light brown before the new leaves dominate the canopy. The graminoids (Dupontiafisheri, Carex aquatilis and Eriophorum angustifolium) developed the highest foliage area index, about 0.2-0.4 (Figure 3-3). However, foliage area indices and mean inclinations vary systematically across microtopographic units and were highest near the wet end of the gradient (Caldwell et al. 1974, Dennis et al. 1978). In standing water, foliage area indices of pure stands of Dupontia and Arc- tophila fulva were as high as 5.2 and 8.5, respectively. At the other ex- treme, the centers of low-centered polygons, occupied predominantly by Carex, had foHage area indices of 0.1 to 0.2. The seasonal progression of aboveground biomass increased with the foliage area and was 80 to 1(X) g m"^ by late July in the moist meadow studied. The foliage areas were not uniformly distributed vertically in the canopy. Foliage was concentrated in the lowest 5 cm of the canopy (Fig- ure 3-4) and in late July was stratified by species and growth form, with the grasses and sedges growing above the dicotyledonous plants. Most of the leaves and stems were steeply inclined, with angles 60° to 90° from 76 P. C. Miller et al. -^---^ -| FIGURE 3-3. Seasonal progres- sion of the development of the foliage area index of leaves plus stems o/Eriophorum angustifo- lium, Dupontia fisheri, Carex aquatilis, other graminoids, and all dicotyledonous plants in 1970 and 1971. the horizontal (Figure 3-5). Carex was the most steeply inclined, with most leaves inclined 80° to 90° from the horizontal, followed by Dupon- tia with inclinations between 50 ° and 90 °, and Eriophorum with foliage distributed almost equally through all leaf inclinations. The dicotyledons and understory plants, for example, Salix pulchra and Petasitesfrigidus, usually have more horizontally incHned leaves. By late July the canopy of the grasses and sedges was made up of the leaves and stems of the in- dividual tillers, each with four to six new leaves and the dead leaves of past years. The density of individual tillers was 2600 to 4800 m"^ (Dennis et al. 1978). Belowground live biomass was 5 to 10 times that above ground and was concentrated near the soil surface. Over 50% was in the upper 5 cm of soil and over 80% in the upper 10 cm (Dennis 1977, Dennis et al. 1978) (Figure 3-6). In the upper 5 cm, stem bases made up 8 to 22% and rhi- zomes 9 to 25% of the belowground biomass. The remaining biomass Biophysical Processes and Primary Production 77 20 h E u H 10 h ? 0) X 20 Jun 1970 I Jul II Jul Kl^ 21 Jul k 3 Aug Area Index, m^ m ^ 201- Z lOh I 23 Jun 1971 3 Jul 17 Jul 27 Jul 9 Aug 0. Lti 22 Aug 1 I ^^ I ■ —r^^ T 1 ^ < 0 .2 .4 0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6 .8 0 .2 .4 £ 24 Aug 0 .2 .4 0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6 0 .2 .4 .6 .8 0 .2 .4 .6 Area Index, m^m"^ Ql-eaf BSheoth and Stem FIGURE 3-4. Seasonal progression of leaf area index, sheath and stem area index, and inflorescence area index during 1970 and 1971. In- florescence area index is shown by the shaded area to the left of the ver- tical axis. 0.6 o.ep I Petasites 0.4 - Salix -| — ij _L _1_ 25 45 65 85 Inclination, degrees FIGURE 3-5. Relative frequency of leaf inclinations (degrees from the horizontal) for five common species in wet meadow tundra: Dupontia fisheri, Eriophorum angustifolium, Carex aquatilis, Petasites frigidus and Salix pulchra. 78 P. C. Miller et al. E u 10 o 20 - ■ Stem base 0 Rhizome D Root Thaw Depth 899 g m" 15 Jun If 570 g m' 3 Aug 100 0 2 623 g m^ 24 Aug FIGURE 3-6. Seasonal progression of the vertical distribu- tion of belowground biomass as percentage of total in stem bases, rhizomes, and roots at different soil depths. The total belowground biomasses are given for each date. The dashed line indicates the seasonal progression of the depth of thaw. (After Dennis 1977, Dennis et al. 1978.) was composed of roots. Between 5 and 10 cm most of the biomass con- sisted of roots, since stem bases did not occur below 5 cm and rhizomes made up less than 1% of the biomass. Although roots were concentrated in the upper 10 cm, they occurred to a depth of 25 cm. The relative pro- portion of the different belowground parts at a given depth varied sea- sonally. The percentage of the total belowground biomass that was stem bases and rhizomes was lowest in early July, but by late August the per- centage had increased to the early season levels. Thus most of the below- ground biomass was in thawed soil early in the season, even though the total soil volume available for exploitation by the plants remained constrained by the underlying permafrost. Species differences in rooting patterns are discussed in Chapter 5. INFLUENCE OF THE CANOPY ON THE PHYSICAL ENVIRONMENT Interception and Absorption of Radiation A discussion of radiation in the tundra canopy must include several wavelength bands, depending on the biological processes being consid- ered. Radiation in the 300- to 700-nm band, "photosynthetically active radiation" or PAR, provides energy for photosynthesis. Radiation between 300 and 3000 nm (insolation) and 9000 and 1 1 ,000 nm (infrared) provides energy that warms the plant above air temperature, and affects the rates of metabolism, growth, and development. The canopy affects the spectral composition of the radiation (Lemon 1963). Within the Biophysical Processes and Primary Production 79 canopy solar irradiance decreases because of interception and absorption by the leaves, stems and dead material, while infrared irradiance com- monly increases because the leaves usually radiate more than the sky does. Most of the photosynthetically active radiation is absorbed and not reflected or transmitted (Stoner et al. 1978a). Thus, from the point of view of the irradiance absorbed for photosynthesis, only the solar irradi- ance penetrating to a leaf without interception is considered. But from the point of view of the total energy exchanged by a leaf and an analysis of leaf temperature, the reflected solar and infrared radiation must be included. Simulation models for irradiance in vegetation canopies have been developed for lower latitudes for vegetation with well developed, homo- geneous canopies (deWit 1965, Anderson 1966, Duncan et al. 1967) and have been applied to the tundra (Miller et al. 1976, Ng and Miller 1977, Stoner et al. 1978c, Tieszen 1978c). Stoner et al. (1978a) showed that the simulation model used previously (Miller et al. 1976, Ng and Miller 1977) predicted the vertical distribution of photosynthetically active radiation well in northern latitudes, so that the basic equations appear valid. For canopy energy exchange, incoming shortwave radiation was divided into downward streams of direct beam and diffuse, and an upward stream of reflected radiation. Infrared radiation was divided into downward and upward streams. Canopy properties affecting the interception and penetration of solar radiation in the canopy included the inclination of the leaves from the horizontal, the distribution of leaf and stem area, and the reflectivities of leaves, stems and dead material. In addition the alti- tude of the irradiating source affected the interception and penetration of radiation. The simulation models were used to estimate the partitioning of solar and infrared radiation in the canopy. Of the incoming direct solar beam, about 86% was intercepted in the canopy; the rest passed through to the soil or moss surface (Figure 3-7). The canopy appeared more trans- parent to diffuse solar radiation because of the scattering and downward reflection of direct beam radiation. The diffuse radiation reaching the soil-moss surface was 36% of the incident diffuse above the canopy be- cause of the additional loss to scattering. About 18% of the incoming solar was reflected back, most of the reflected amount coming from the canopy rather than the soil-moss surface. Some absorbed solar radiation was emitted as infrared. Infrared radiation was lost from the canopy both upwards and downwards. However, the canopy received more in- frared from the soil-moss surface than it lost to the sky. The net radia- tion in the canopy was about twice that of the soil-moss surface. Of the net radiation absorbed by the canopy, 80 to 90% was lost by convection and 10 to 20% was lost by evaporation. Bowen ratios — convectional heat loss divided by evaporative heat loss — for the canopy were 4 to 9. Transpiration was low because of the nearly saturated air 80 P. C. Miller et al. SOLAR Direct Diffuse Reflected 2.9 1 17.0 ■ -3.6 Canopy 2.5 10.9 -2.4 ' 0.4 6.1 -1.2 INFRARED Down Up 23.9 -3.9 -28.5 0.8 27.8 11.7 7.9 -29.3 3.8 Soil Rr Canopy 11.7 CONVECTION 8.9 7.9 777777777777 Soil 777777777777777777777 EVAPORATION CONDUCTION 2.8 4.6 ' 2.5 3.8 0.8 1.8 2.0 ninnnnnninnnni/i/nnnninnnj 0.04 FIGURE 3-7. Partitioning of incoming solar and infrared irradiance (MJ m'^ day'^) by the canopy and soil in the meadow vegetation type. Con- vectional loss is divided into that lost from standing dead material (4.6 MJ m~^ day') and that lost from green leaves (2.5 MJ m'^ day'). (After Stoner et al. 1978b.) with relative humidities of 90 to 100% within the canopy. The high relative humidity was caused by the high rates of evaporation from the moss, and by the convectional loss of radiation that was intercepted by standing dead material. At the moss surface 30 to 50% of the absorbed radiation was lost by convection and 50 to 70% by evaporation. Conduc- tion accounted for less than 1% of the incoming solar radiation. Of the total water lost by evapotranspiration from the wet meadow 14 to 20% was lost by transpiration from the vascular plants. The remainder was lost by evapotranspiration from the moss understory (Miller et al. 1976, Ng and Miller 1977, Stoner et al. 1978b). This partitioning of water loss was confirmed in field studies with tritiated water (Koranda et al. 1978). During the growing season the fraction of incoming radiation inter- cepted by the Carex-Oncophorus meadow canopy increased as the foli- age area index increased, while the intercepted radiation per unit of foli- Biophysical Processes and Primary Production 81 1.4 - 1.2 E xj . — < fe O.R a> ■*- ?s 0.6 - 0.4 FIGURE 3-8. Seasonal progression of total foliage area index (live plus dead, Af) and the fraction of incoming beam irradiance intercepted by the canopy (o) and by a unit foliage area (mj. age area decreased because of increased self-shading (Figure 3-8). On 21 June, when solar irradiance was potentially the greatest, the fraction of incoming radiation intercepted by live leaves in the Carex-Oncophorus meadow was almost zero because of the small foliage area index, but the irradiance per unit foliage area was large. The irradiance per unit foliage area decreased dramatically during the season when the standing dead material was absent, but less dramatically when a constant standing crop 2 4 6 Foliage Area Index FIGURE 3-9. Ratio of inter- cepted to incoming beam ir- radiance per unit ground sur- face ( — ) and per unit of foli- age surface ( — ) at three solar altitudes (fi) with differ- ent foliage area indices. 82 P. C. Miller et al. 15 10 1200 2400 0 Hour 5 2.5 1200 Hour 10 2400 0 1.0 W FIGURE 3-10. a) Isopleths of solar irradiance (300 to 3000 nm) absorbed by leaves (J m'^ s'^) at different heights in the canopy through the day on about 15 July 1971. b) Isopleths of leaf temperatures through the day for 15 July, c) Vertical profiles of live (m) and dead (o) foliage area indices (Af). of dead material was present. Interception by evergreen shrubs was more constant through the growing season than was interception by grasses and sedges because shrub leaf and stem areas were more constant. Thus photosynthesis was possible earlier in the season in evergreen shrubs. On 21 June, the interception efficiency of a canopy with leaves incHned 65 ° and a foHage area index of 1.0 was about 0.6. With similarly inclined leaves and foliage area index of 2.0, interception was about 0.96. The fraction of incoming direct beam radiation intercepted by the canopy increased as solar altitude decreased; thus interception was high with relatively low foliage area indices (Figure 3-9). On 21 June at solar midnight with the sun 5° above the horizon, interception was almost complete with foliage area index of 0.5. At solar noon with the sun about 40° above the horizon, interception was only 0.3 with the same foHage area index. At this time complete interception required a foliage area in- dex of about 4.0. The Dupontia and Arctophila stands, after developing foliage area indices of 5 and 8 respectively, should intercept all incoming solar radiation. Leaf absorptances were lower in regions with higher solar irradiance than in regions with lower solar irradiance, and were higher in the alpine than in the Arctic (Billings and Morris 1951, Mooney and Billings 1961, Mooney and Johnson 1965). In the simulation models for vegetation of Biophysical Processes and Primary Production 83 30 - 20 - "TyyyrTy ''77>>r-rr777ZZZ:rr: __^e„„ - -^^^^""^.^ "down ^ * net loss 1 1 1 ■ 111! - 20 10 20 10 20 Jun Jul Aug 10 FIGURE 3-11. Seasonal progression of incoming solar irradiance (Q), quantity reflected up from the canopy (Q,.), quantity intercepted by the canopy (Qj), quantity absorbed by the soil (Q^), quantity absorbed by the canopy (QJ, and quanti- ty absorbed by the photosynthetic tissue (QJ and of downward and upward infrared irradiance (Q). The net infrared loss is partitioned by source. the coastal tundra at Barrow (Miller and Tieszen 1972, Miller et al. 1976, Ng and Miller 1977), absorptances of 0.5 for live leaves and stems and 0.4 for dead leaves were used. Chlorophyll concentrations in the vegeta- tion at Barrow were similar to those of temperate plants (Tieszen 1972b). Absorbed solar radiation per unit leaf area was highest at the top of the canopy throughout the day and through the growing season (Figure 3-10). However, even at the top of the canopy absorbed irradiances were near or below light saturation for photosynthesis, indicating that photo- synthesis was usually light-limited (Chapter 4). One might anticipate greater leaf areas at canopy levels where photosynthesis was greater. But most of the live leaf area was concentrated at the bottom of the canopy and most of the solar radiation on a ground area basis v/as absorbed near the ground surface. 84 P. C. Miller et al. The incoming solar radiation was partitioned between the canopy and the soil for the Carex-Oncophorus meadow during the growing sea- son. The incoming solar radiation was relatively high in June and became lower as the growing season progressed. The fraction of incoming solar radiation (300 to 3000 nm) absorbed in the canopy varied because of the different live and dead leaf area indices, but increased through the season as the Hve leaf area increased (Figure 3-11). The lowest interception oc- curred in the coldest and wettest year (1973) because of the small leaf area index in that year. The reflected fraction decreased through the growing seasons as the darker colored live leaf area extended above the lighter colored dead leaf area. The ground surface absorbed 50 to 60% of the incoming solar radiation at the beginning of the season and 40 to 50% by the end of the season. Under an evergreen shrub canopy, with its more even seasonal course of leaf area, the ground surface would have a more uniform seasonal course of radiation absorption. Air Temperature, Humidity, and Wind Profiles Air temperatures within the vegetation canopy differ from air tem- peratures above it because of the heat exchanged between the air, soil surface, leaves and stems. The vertical profile of air temperatures through the canopy depends upon the absorption of radiation vertically through the canopy, which depends on the solar altitude and the profiles of leaf and stem area. With low solar altitudes, when the solar radiation is intercepted mostly near the top of the canopy, the air is warmest at the top of the canopy. With high solar altitudes, solar radiation penetrates to the ground surface and the air near the ground surface is warmed more than the air higher in the canopy. These patterns are consistent with measurements of Weller and Holmgren (1974a) at Barrow and Larcher et al. (1973) in the Austrian alpine tundra. Because net radiation is positive through the 24-hour arctic summer day, surface temperatures and air temperatures near the ground are usually higher than air temperatures above the canopy or near the top of the canopy, in contrast to the diurnal patterns of air temperatures in lower latitudes. The difference between surface and above-canopy air temperatures can be up to 20 °C in the Carex-Oncophorus meadow and up to 30 °C on drier beach ridges (Kelley and Weaver, unpubl.). Similarly, air humidities within the canopy differ from those above it. Evaporation from the ground surface increases the humidity near the ground surface and transpiration from leaves increases the humidity of the air within the canopy. The exact profile has not been measured be- Biophysical Processes and Primary Production 85 cause such measurements involve difficult logistic problems in the low canopy of the coastal tundra vegetation. The humidity difference should be close to the difference between the saturation vapor density at the temperature of the ground surface and the humidity of the air. Heat and humidity of the canopy air are exchanged vertically by processes of turbulent transfer, which are related to wind and to foliage area in a complex manner (Monteith 1974). The Carex-Oncophorus meadow canopy, with its small leaf area at the top and larger leaf area at the bottom, has a rapid exchange of air near the top of the canopy and slower exchanges near the ground. Weller and Holmgren (1974a) meas- ured wind profiles through the canopy with a hot wire anemometer. Foli- age area indices were not measured at the same time, but assuming that the foliage area index was between 1 and 2, and that wind decreases ex- ponentially with the foliage area index, an extinction coefficient for wind of 1 to 1.5 (foliage area index)"' is appropriate. Wind has been shown to affect stomatal opening, photosynthesis, and transpiration (Caldwell 1970a,b, Grace and Thompson 1973), and has been imphcated in reduc- ing growth by lowering plant temperatures (Warren Wilson 1966a). Effect of Plant Properties and Environmental Factors on Leaf Temperatures The influence of plant form on plant temperature was suggested pre- viously. Krogg (1955) found that willow catkins with transparent hairs reflected solar radiation to the inner surface while trapping the infrared radiation. The inner surface was dark, increasing the absorption of solar radiation. As a result, catkin temperatures were several degrees above ambient temperatures. Hocking and Sharplin (1965) noted that flower shapes are sometimes parabolic, focusing the sun's rays into the center of the flower. The warmer center may then attract pollinators or speed de- velopment of reproductive parts. The influence of plant properties on plant temperature can be simulated by defining the environmental vari- ables and solving the energy budget equation for plant temperature. Such simulations indicated that leaf temperatures may increase 0.07 °C per percent change in leaf absorption, 0.2 °C per mm change in leaf width, and 2 to 3 °C per s cm'' change in leaf resistance at low leaf resistances (Figure 3-12). Leaf temperatures can be expected to rise with decreased wind speed at low wind speeds, but be relatively unaffected by changes in air humidity. 86 P. C. Miller et al. 10 0.5 1.0 Absorptance a> 1- 0 Leaf 25 50 Resistances cm"' 5 20 15 _ D. r Wind. Icm s' 10 -r~ Wind, 20cm s'' r> f Wind.SOOcms' i 1 1 0 25 50 Leaf Resistance, s cm"' FIGURE 3-12. Effect on simulated leaf temperatures on about 10 July (5-day mean) at solar noon of changes in A) leaf absorptance, B) leaf width, C) leaf resistance to water loss at two ambient vapor densities, and D) leaf resistance to water loss at three wind speeds. The standard envi- ronmental conditions used were: total solar irradiance 560 J m'^ s~\ infrared radiation from the sky 280 J m~^ 5"', air temperature 6.0°C, wind speed 1.0 m s'^ and vapor density of the air 7.9 g m'\ Influence of the Canopy on the Soil Thermal Regime The moss or soil surface temperature is the interface between the aerial and soil thermal regimes. In natural and modified vegetation cano- pies, thaw depth was shallower with higher foliage areas (Brown et al. 1969, Linell 1973, Ng and Miller 1977). In field experiments (Ng and Mil- ler 1977), thaw depth on 1 August decreased 3 cm with an increase in foli- age area index of 1.0 on the control plots. These shallower thaw depths reduce the available minerals and water for the plants, and lower temper- atures of the roots. Biophysical Processes and Primary Production 87 Temp, "C 10 Jul 20 0 20 1 . • i - '/ / 40 - \ 60 1 /i 1 1 1 1 1 -20- -4 -2 -60 Temp, "C 22 Jul Temp, "C 4 Aug FIGURE 3-13. Profiles of air and soil temperature at different times of the year in the moist meadow canopy. Points show measured values of daily mean temperature. Curves give results of simulations. (After Miller et al. 1976, Ng and Miller 1977.) Ng and Miller (1975, 1977) presented a model of canopy processes, ground surface heat exchange, and soil heat conduction based on micro- climate and foliage area profiles measured in 1973. The model predicted the seasonal course of soil temperature well, except during a period of snow (Figure 3-13). The simulation indicated that in the Carex-Onco- phorus meadow, evaporation from the wet moss-soil surface can ac- count for over 80% of the latent energy lost with low and ambient foliage area indices (0.5 and 1 .9), but that evaporation can decrease to 40 to 75% of the energy lost with a foliage area index of 4.6. The amount of the decrease depended on the extinction coefficient for turbulent transfer; the coefficients used, 1.0 and 0.5 respectively, were within the range of measured values. With a dry ground surface and foliage area indices of 0.5 and 1.9, convection accounted for over 90% of the energy lost from the ground, but at the high foliage area index convectional loss was 77 to 84%, depending on the turbulent exchange extinction coefficient (1.0 and 0.5, respectively). Under normal or wet conditions with a turbulent exchange extinction coefficient of 0.5, convection added energy to the surface with all foliage area indices. But using an extinction coefficient of 1 .0, convection added energy only with the low and ambient foliage area indices. Using the extinction coefficient of 0.5, both the energy con- ducted into the ground and the depth of thaw decreased with the higher foliage area index. However, the converse occurred with an extinction coefficient of 1.0. These results have the common basis that decreasing the extinction coefficient for turbulent transfer increases the turbulent transfer at the soil surface and increases heat loss due to evaporation and 88 P. C. Miller et al. convection. However, convection is more sensitive to turbulent exchange than is evaporation. Increasing the foliage area index decreases solar ir- radiance at the surface, increases the absorbed infrared, and diminishes turbulent transfer. Decreased soil moisture results in higher surface tem- peratures and decreases evaporative heat losses. Simulations of the seasonal course of the depth of thaw with differ- ent environmental conditions, using 1973 environmental data as stand- ard conditions, indicated decreasing sensitivity, i.e. centimeter change in thaw depth at peak season per unit change in the environmental variable, in the order (Figure 3-14): diffuse solar radiation, total solar radiation, infrared radiation, vapor density, air temperature, and turbulent diffus- ivity. The thaw depth at peak season increased with increased solar ir- radiance, infrared irradiance, and air temperature, but decreased with turbulent exchange and air humidity. The seasonal course of thaw depth with different values for ecosystem properties indicated that the thaw depth was most sensitive to changes in thermal conductivity of the or- ganic layer, air resistance near the ground, thickness of the organic layer, and leaf inclination, and was least sensitive to the reflectance of the sur- face under the vascular canopy. Expressed in terms of the expected ac- curacy of the instruments used to measure the variables, the sensitivities were in the order (most sensitive to least sensitive): thermal conductance of the organic layer, thickness of the organic layer, vapor density, diffuse solar radiation, total solar radiation, infrared radiation, leaf inclination, air temperature, ground air resistance, turbulent diffusivity, leaf area in- dex, and ground surface reflectance. Changing the vertical distribution of the foliage area index had little effect on the air temperature and humidity profiles. The general trends in sensitivity were the same in another set of sim- ulations involving a plot from which the canopy was removed. The depth of thaw was more sensitive in the clipped plot than in the control plot to solar radiation, ground surface reflectance, boundary layer resistance at the ground surface, and the thickness of the organic layer. Thaw was less affected by infrared radiation in the clipped plot than in the control and showed about the same sensitivity in the two plots with regard to air tem- perature and thermal conductivity. These changes in sensitivity relate to the attenuation of solar radiation and turbulent transfer by the canopy. The thaw development under horizontal leaves, as contrasted with more vertical leaves, decreased because of the decreased penetration of solar radiation with horizontal leaves. The thaw deepens with increasing boundary layer resistance or vapor density because both suppress evapo- ration, and increase surface temperature and conduction. The simulations indicate that as the vascular canopy develops, and standing dead or live and dead moss material accumulates, the depth of thaw decreases and the potential volume that can be exploited by the Biophysical Processes and Primary Production 89 E u o. Q o 38 -5 18 0 0 50°/ 1 1 22 g. \ \ - 26 - V - 30 - > \- 34 - 1 1 -100 ■50 0% 600% FIGURE 3-14. Simulated sensitivities of the depth of thaw on 8 August to changes in various canopy parameters. The parameters were changed from the standard case values by the percentage indicated on the abscissa, except for the boundary layer resistances of the soil surface and the resistance of the soil to evaporation. The canopy parameters are a) leaf inclination, b) foliage area index, c) extinction coefficient for turbu- lent exchange, d) soil surface reflectance, e) relative humidity of the ef- fective evaporating surface of the soil, f) boundary layer resistances of the soil surface, g) conductivity of the organic layer, and h) depth of the organic layer. In (c) the sensitivity of the thaw depth to the extinction coefficient for turbulent exchange was calculated with midseason foliage area indices of 1.56 (o) and 4.68 (•). Otherwise, where two lines are given, ▲ indicates sensitivities run with the foliage area index of a canopy from which the foliage was removed. Dashed lines indicate end points off scale. (After Ng and Miller 1977. J 90 P. C. Miller et al. TABLE 3-3 The Effect of a ±20% Change of Initial Conditions after 100 Simulated Years, Expressed as Percentage Change from the Standard Case Effect on variable after 100 years (%) Variable changed initially Organic mat thickness Thaw depth Annual phosphorus release Live aboveground standing crop Foliage index Organic mat thickness Thaw depth Live aboveground standing crop ±16 ± 6 0 ± 3 ±17 0 + 2 to -4 ±17 0 ± 5 ±28 0 + 4 to -7 ±26 0 Source: Miller (1978). roots of plants for water and minerals decreases. The depth of thaw on 1 August decreased about 1.3 cm per unit increase in foliage area index in simulations, compared to 1.5 to 0.1 cm per unit foliage area index in field measurements. It is possible that without a periodic removal of the standing dead material by lemmings the vegetation composition would shift to shallow-rooted species such as mosses, lichens and Dupontia or to evergreen forms requiring less nutrients. Decreased thaw depth and in- creased moss has occurred in long-standing exclosures in the vicinity of the Biome research area. Some of the interactions between foliage area index, thaw depth, phosphorus availability, plant growth and accumulation of soil organic matter were explored in a simplified ecosystem model (Miller 1978). Co- efficients relating the annual change in each compartment to the state of the system were estimated from the understanding at the time of the pro- cesses involved. The initial values for the organic mat thickness and thaw depth influenced several variables after 100 simulated years, but changes in initial values for standing dead and live vegetation had little effect (Table 3-3). The simulations support the notion that subtle changes in the tundra may persist for many years, although in its grosser features the system appears unchanged. Parameters defining phosphorus cycling were critical in influencing the system. Changing the initial amounts of standing dead and live plant biomass by ± 20% had no effect but chang- ing the initial thickness of the organic mat by ± 20% caused long-lasting changes in several variables. Changing the initial thaw depth by ±20% changed the peak season thaw depth, phosphorus release, standing crop, foHage area index, and organic mat thickness. The state of the system after 100 years was influenced by changing parameters defining trans- fers, especially those affecting phosphorus release. Periodic clipping of aboveground standing crop by lemmings stabilized the system, although varying the period between clippings from 3 to 5 years had little effect. Biophysical Processes and Primary Production 91 CONTROL OF EVAPOTRANSPIRATION BY PLANTS AND PLANT-WATER RELATIONS The control of water loss by the plants affects both surface energy exchange processes and plant physiological processes. The maintenance of turgor is essential for photosynthesis, respirative growth, and develop- ment (Hsaio 1973). Various aspects of the plant water relations are af- fected at temperatures well above those typical at the Barrow research sites, including water absorption (Kuiper 1964), plant growth and osmotic potential of the leaves (Kleinendorst and Brouwer 1970) because of decreased root permeabihty. The relative water content and growth of alpine and subalpine species were reduced by soil cooling, although net photosynthesis and transpiration were unaffected (Anderson and McNaughton 1973). Stomatal opening in temperate plants was inhibited at leaf temperatures lower than 10 °C (Kuiper, cited by Ketellapper 1963, Courtin and Mayo 1975). The flow of water was viewed in a simple model of the soil-plant- atmosphere continuum (Stoner and Miller 1975, Ehleringer and Miller 1975, Miller et al. 1976). Water flows through the roots and stems to the leaves because of a difference in the water potentials of the soil and leaves, and is lost to the air because of a difference in the vapor densities of the leaves and air. The flow of water through the plant is restricted by the resistance of the root-soil system to water uptake and the resistances of the leaves and leaf air boundary layer to water vapor diffusion. Mosses were viewed similarly except that water uptake is mainly on the surface of the moss, Hquid water is absorbed through the surface, and water loss from the tissue is suppressed when a surface film of water is present (Miller et al. 1978a, Stoner et al. 1978b). Water Relations of Vascular Plants The partitioning of absorbed energy into convection and evapora- tion is controlled by stomatal closure, which occurs to prevent low water contents and detrimental water potentials, either of which may be harm- ful (Jarvis and Jarvis 1963). Leaf resistances increased abruptly with rel- ative water contents below about 91% in Arctophila, 80 to 89% in Du- pontia. Potent ilia hyparctica and Salix pulchra, and 72% in Eriophorum angustifolium (Figure 3-15) (Stoner and Miller 1975). Species in other regions are similar, i.e. between 80 and 89% for Caltha leptosepala and Bistorta bistortoides (Ehleringer and Miller 1975), 85 and 90% for alpine plants (Anderson and McNaughton 1973), and 90% for Populus tremula 92 P. C. Miller et al. (Relative Water Content) FIGURE 3-15. The relationships between relative water content and leaf resistance and between relative water content and leaf water potential for Arctophila fulva (A.f.), Dupontia fisheri (DJ.), Carex aquatilis (C.a.), Eriophorum angustifolium (Em.), Potentilla hyparctica (P.h.), and Salix pul- chra (S.p.). (After Miller et al. 1978b.) and Betula verrucosa (Jarvis and Jarvis 1963). The similarity of these levels in different environments and species indicates that most vascular plants maintain leaf water contents of more than 80 to 90% of their turgid water content. The relation between water potential and water content of the leaves also differed among the species in the Carex-On- cophorus meadow (Stoner and Miller 1975). Arctophila showed the highest rate of change of leaf water potential with a change in relative water content (Figure 3-15). Eriophorum had the lowest. Leaf water potentials of grasses, sedges and soft-leaved forbs were near 0 bars with low transpiration rates during the night, even though the sun is above the horizon 24 hours a day. These leaf water potentials were higher than those of well-watered fell-field plants maintained in satu- rated air in the laboratory, including Dryas integrifolia (-10 to -15 bars), Saxifraga oppositifolia (-10 bars), and Dryas octopetala (-13 bars) (Courtin and Mayo 1975). Soil water potentials within the root zone were always greater than -0.5 bar. During the relatively warm, dry summer of Biophysical Processes and Primary Production 93 1972, midday water potentials reached -28 bars in Dupontia, -25 in Carex, and -13 in Poa arctica. But in the relatively wet, cold summer of 1973, water potentials were above -15 bars (Stoner and Miller 1975). In other studies water potentials of soft-leaved forbs were similar, above -10 bars in Thlaspi alpestre (Rochow 1967) and above -11 bars in Caltha leptosepala (Kuramoto and Bliss 1970), but those of evergreen dwarf shrubs and trees were lower, -35 bars in D. integrifolia (Courtin and Mayo 1975), -60 bars in D. integrifolia, -54 in Picea englemanii, -35 to -54 in S. oppositifolia, -62 in spruce, and -54 in Diapensia lapponica (Courtin 1968). Leaf resistances of arctic and alpine plants were also similar to those of temperate zone plants (Miller et al. 1978b). Minimum leaf resistances in the coastal tundra at Barrow were 1 to 3 s cm"'. Cuticular resistances were 12 to 39 s cm"'. Leaf resistances decreased as temperatures rose to 15 °C. The light response curve for stomatal opening indicates that mini- mum leaf resistances of 1 to 2 s cm"' were approached at 140 J m"^ s"' (400 to 700 nm) in plants grown at 5 °C in the laboratory (Figure 4-4). In the field open stomata occurred at 70 J m"^ s"' (400 to 700 nm). Ehler- inger and Miller (1975) reported minimum leaf resistances at 140 to 210 J m"^ s"' solar irradiance (300 to 3000 nm) in the alpine tundra on Niwot Ridge for Caltha leptosepala and Bistorta bistortoides. Courtin and Mayo (1975) reported high minimum resistances for Dryas integrifolia. With reasonable conversions between incoming shortwave and photo- synthetically active radiation, it can be demonstrated that tundra plants at Barrow open stomates at lower light intensities than do alpine plants. The root-soil resistance to water uptake of an entire plant included the root's permeability to water uptake, and the total root mass. Root re- sistances measured in situ were 0.6 to 1.7x10* bar s cm"'. Root resis- tances decreased in the order Dupontia, Carex and Eriophorum, which had rooting depths of 15, 20 and 25 cm, respectively. Field measurements on coastal tundra species and laboratory measurements on alpine species indicated that root resistances were independent of root temperatures but related to transpiration rates (Stoner and Miller 1975, Caldwell et al. 1978). Typha latifolia, from high elevations, showed no change in rela- tive water content with changes in root temperature between 20° and 30 °C, while plants from low elevations showed decreases (McNaughton et al. 1974). Taken as a whole, these data support the idea that species from colder climates have lower root resistances at low temperature. Water Relations of Mosses Water relations of mosses differ from those of vascular plants, part- ly because mosses show little control over tissue moisture status. Mosses 94 P. C. Miller et al. lack roots and, except for genera such as Polytrichum and Pogonatum, most mosses lack the functional equivalent. The moisture supply in the soil is largely unavailable, and moss tissue water contents can vary widely. The movement of liquid water occurs primarily as a result of cap- illarity on the outside of the plant (Bowen 1931, 1933, Magdefrau 1937, Anderson and Bourdeau 1955), and is probably limited to within 1 to 2 cm above the water surface (Anderson and Bourdeau 1955). Water ab- sorption from the capillary stream can take place at the leaf bases, leaf traces, and especially at the thin-walled cells at the plant tip (Bowen 1933). Rapid absorption is aided by the lack of cuticle in most species (Vaizey 1887, Czapek 1899, Kressin 1935, Patterson 1943), a feature which also facilitates desiccation. Water vapor is absorbed less than li- quid water. In a saturated atmosphere mosses generally only reach water contents (g water gdw'') of 30 to 60%. At 95% relative humidity, water contents of 50% were reported (Patterson 1943), whereas the water con- tent of mosses placed in liquid water for a few minutes may be 300 to 700% (Muller 1909). Anderson and Bourdeau (1955) found that both Atrichum and Polytrichum failed to become turgid in relative humidities of up to 100% and that turgid mosses wilted at 95% relative humidity. However, the mosses became turgid minutes after liquid water was added. It appears, therefore, that mosses less than 1 to 2 cm high may acquire water through capillary movement from the soil surface or through the addition of liquid water in the form of rain, dew or fog. Taller mosses must rely almost solely on aerial transport of liquid water in the form of rain, dew or fog for active growth and photosynthesis. Possible excep- tions are the Polytrichaceous species which, especially at relatively high humidities, may maintain turgidity through the transport of water from the underlying organic or soil layer, within and along the stem. In the tundra at Barrow, moss shoots were usually less than 2 cm in height above the soil surface. The frequent saturation of the soils and the high incidence of rain, fog and dew suggested that both capillary movement of water from the soil surface and the application of liquid water directly to the moss surface were important in supplying moisture to the moss. Transpiration rates in mosses are potentially much higher than in vascular plants because of low resistance to water loss. Calliergon sar- mentosum is a mesic to hydric species and showed little resistance to des- iccation. Pogonatum alpinum occurred in xeric to mesic locations and showed numerous xerophytic adaptations, including a well-developed cuticle, the abihty to roll the tissue margins over the photosynthetic lamellae during desiccation, and the ability to fold the leaf tissue against the stem. Oechel and Sveinbjornsson (1978) showed that with water con- tents of 400%, a temperature of 24 °C, relative humidity of 20%, and wind speed of 1.7 m s"', Calliergon lost 0.23 g H2O gdw"' min"' and Po- gonatum only lost 0.03 g H2O gdw"' min"'. Both air resistances and tissue Biophysical Processes and Primary Production 95 resistances of the moss canopy were low when expressed on an areal basis. Air resistances were similar for C sarmentosum and Dicranum elcngatum and ranged from 0.9 to 1.0 s cm"' respectively at a wind speed of 4.5 m s"'. P. alpinum had an air resistance of only 0.3 s cm"' at 4.5 m s"', presumably as a result of the very rough canopy (Alpert and Oechel, unpubl.). All moss species showed little tissue resistance to water loss (less than 0.5 s cm'') at the field water contents commonly observed (Figures 3-16 and 3-17). As tissue desiccated C. sarmentosum increased resistance to water loss only slightly. At the moisture compensation point for pho- tosynthesis (75% w.c), C sarmentosum showed a leaf resistance of only 0.84 s cm"'. P. alpinum, on the other hand, had a higher resistance of greater than 3.0 s cm"' at its moisture compensation point for photosyn- thesis (60% w.c). At a water content of 200%, the resistance in Pogonatum was the lowest of all species, 0.3 s cm"'. However, resistance increased to 3.2 s cm"' at 75% water content. The low resistance at high water contents of Pogonatum alpinum may partially be the result of the photosynthetic lamellae present on the leaf surface. As desiccation pro- ceeds, the rolling of the tissue margins over the photosynthetic lamellae and the appression of the leaf tissue against the stem increased the leaf resistance. The more responsive nature of resistance to drought in Pogo- natum and Polytrichum species is presumably an important adaptation to the more xeric sites where they often occur. The colony growth form has been found to increase the rate of water uptake over that achieved by individual shoots. Pogonatum represents the less dense turf growth form, and water uptake rates were dictated pri- marily by the response of single shoots. Calliergon, on the other hand, develops a carpet growth form with high shoot density. In Calliergon sar- mentosum water is held on the tissue surface and the colony form is much more important in controlling water uptake and loss than in Pogo- natum (Gimingham and Smith 1971). However, it appears that appreci- able water can be taken up via the tissue bases on both species. These rates of water uptake vary considerably. Pogonatum, for example, had a low rate of uptake, as shown by the length of time required to recover 50% of the total water content (WCso). The rate of water uptake from air-dried status to WCso was 0.24 g H2O gdw"' min"' when the bases were immersed in water to a depth of 2 mm and 0.01 g H2O gdw"' min"' when the apexes were immersed to 2 mm. Carpets of Calliergon took up water much more rapidly and to a larger extent. Water was taken up via the apexes at a rate of 4.05 g H2O gdw"' min"' and WC50 was reached in 1.2 minutes. Water uptake via the bases was faster at a rate of 12.9 g H2O gdw"' min"' with WCso achieved in only 0.3 minute (calculated from Gimingham and Smith 1971). The rates of water uptake for these two moss species approximate the range of extremes found at Barrow. 96 P. C. Miller et al. _ 2.0 t 1 1 'E Po. u - - in -- 16 - I - 0) u \ c — \ \ - o \ \ \ \ .!C 1.2 - \ \ - m \ \ 0) ■^ _ q: C?s>v \ o) 0.8 - ^\^ \ XD.e. - =) ^^ \ tn ^^^ (/) - " H 0.4 - - >e* - 1 ^"^^^ 100 200 300 Percent Water Content 400 FIGURE 3-16. The relationship between water content and leaf resistances in three moss species: Pogonatum alpinum (P. a.) from Eagle Creek, Alaska (Alpert and Oechel, unpubl.); Dicranum elongatum (D.e.) from Barrow; and Calliergon sarmentosum (C.s.) from Barrow. (After Oechel and Sveinbjornsson 1978.) FIGURE 3-17. The seasonal course of the tissue water contents in four moss species: Calliergon sarmentosum (C.s.), Dicranum elongatum (D.e.), Dicranum angustum (D.a.), and Pogonatum al- pinum (P. a.). (After Oechel and Svein- bjornsson 1978.) Biophysical Processes and Primary Production 97 The low irradiance, low temperature, and low evapotranspiration helped maintain the turgidity of moss tissue. Pogonatum had favorable moisture levels throughout some summers, 150% to 200^^0 water con- tent, which were sufficient to maintain photosynthesis at near optimal levels (Oechel and Collins 1976, Sveinbjornsson 1979). Calliergon oc- curred primarily in troughs and wet meadow areas and was often sub- merged or inundated with water early in the year, resuhing in water con- tents greater than 1400%. As the season progressed, water levels dropped and water contents declined to 400% (Figure 3-17). Even at this level, photosynthesis was at near-maximum rates. Locally, however, tissue moisture relations affected species distribu- tion and survival. The moisture regimes of rims of low-centered polygons appeared to restrict the growth of Calliergon in those areas, and the wet nature of many Calliergon habitats would depress photosynthesis in Pogonatum if it occurred there (Oechel and Collins 1976). Local populations of all species, and especially of Calliergon, became desiccated at various times during the summer. Populations on sides and tops of high-centered polygons drained free of water early in the summer and became desiccated. Individuals growing on the sides of polygon troughs that were water-filled early in the season were especially prone to desiccation. Also, periods of drought had a major influence on bryophyte growth and survival. An exceptionally warm, dry period of several weeks in 1972 resulted in the death of numerous individuals of Pogonatum and Calliergon. Damage was delayed and less extensive in Pogonatum than in Calliergon. At the end of the drought period, large areas of Calliergon mats recovered by initiating new shoots from existing material (Oechel and Collins 1976). Because of water uptake through the deep stems of Pogonatum, this species can maintain turgidity in the absence of standing water or precipitation for much longer periods than the other species examined (Oechel, unpubl.). SIMULATION OF PLANT-WATER RELATIONS The dynamics of vascular plant-water relations were simulated for the June- August period in 1970-73 (Stoner and Miller 1975). Simulations for 1973, which was cold and wet relative to the long-term average, indi- cated that leaf resistances of Arctophila and Potent ilia decreased in the early morning, then increased in midmorning to a plateau which contin- ued through the day. Dupontia resistances increased to a maximum at midday and decreased in the afternoon. Salix and Eriophorum showed slight increases in leaf resistance at midday, while Carex leaf resistances were low throughout the day. The increase in leaf resistances in the mid- dle of the day reflected the sensitivities of leaf resistance and leaf water 98 P. C. Miller et al 0 0600 1800 0600 0600 Solar Time 1800 0600 Solar Time FIGURE 3-18. Diurnal courses of leaf water po- tential for 29-30 July 1973 for Arctophila fulva (A.f), Dupontia fisheri (D.f.), Carex aquatilis Cm.), Eriophorum angustifolium (E.a.J, Potentil- la hyparctica (P.hJ and Salix pulchra S.p.). Meas- ured (o = individuals in troughs, • = individuals on polygon tops) and simulated ( — ) values are given. Vertical lines are the 95% confidence inter- vals. (After Stoner and Miller 1975.) potential to leaf water content. The simulated courses of leaf water potential generally followed the observed courses (Figure 3-18). Simula- tions for 1972, which was relatively warm and dry, indicated stomatal closure occasionally in Carex and frequently in Dupontia (Figure 3-19). Complete stomatal closure occurred in Dupontia for short periods. In 1973, Dupontia usually was partially stressed; Carex, Eriophorum and Salix had little or no stress while Arctophila and Potent ilia showed com- plete stomatal closure. Biophysical Processes and Primary Production 99 u 1 P.h. t I -10 : ;J7 \/ 1 1 ICD IJ> 1 1 20 1 1 1 0 - -20 1 I s.p. I A ■ ^ A B 1 1 1 15 15 15 15 15 15 Jun Jul Aug Jun Jul Aug FIGURE 3-19. Seasonal courses of midday leaf water potentials for Carex aquatilis (C.aJ, Dupontia fisheri (DfJ, Arctophila fulva (A.f), Eriophorum angusti- folium (E.a.), Potentilla hyparctica (P.h.) and Salix pulchra (S.p.) in 1972 and 1973. Water potentials at which leaf resistances are three times the minimum (A) and infinite (B) are given. (After St oner and Miller 1975.) Measurements of photosynthesis in the field (Tieszen 1975, 1978b) indicated a midday depression in carbon dioxide uptake, which was relat- ed to a slight stomatal closure (Tieszen 1978b). Thus water stress may limit vascular plant production at Barrow by increasing stomatal resis- tance and decreasing photosynthesis. The role of water in limiting pro- duction via its effect on growth has not been studied. The species, in de- creasing order of their sensitivity to water stress, were Arctophila, Du- pontia, Carex, Salix, Potentilla and Eriophorum. The water relations of the mosses Dicranum and Calliergon were simulated in the open or under a vascular plant canopy along a substrate moisture gradient (Stoner et al. 1978b). The canopy tended to increase 100 p. C. Miller et al. available water by reducing the vapor density gradient between the mosses and the air but also tended to decrease the available water by in- tercepting 30% of the summer precipitation. The significance of the counteracting tendencies varied with the substrate water potential. Both Calliergon and Dicranum showed similar seasonal courses of water con- tent with and without the canopy and with different substrate water po- tentials. With substrate water potentials of -1 bar both species had a higher water content under the canopy than in full sun. The effect of the canopy on plant water potential was much reduced at a substrate water potential of -5 bars. During the late-summer dry period interception of precipitation reduced the water reaching the surface, and moss in the full sun had higher water contents than did moss under the canopy. The water contents of Dicranum had the same pattern as Calliergon, although the effect of the canopy in reducing evaporation was less be- cause of the higher resistances of Dicranum to water loss. SUMMARY In summary, in 1972, a year of near normal temperatures and pre- cipitation, calculated gross primary production above and below ground, averaged for the whole Barrow region, was about 465 gdw m"% including 358 gdw m"^ for vascular plants, 106 gdw m"^ for mosses, and 1 gdw m'^ for lichens. Net primary production was about 230 gdw m"^ including 162 gdw m"^ for vascular plants, 66 gdw m"^ for mosses and less than 1 gdw m"^ for lichens. The average net primary production above ground was 108 gdw m"^ Belowground production was about 120 gdw m'^ In the Carex-Oncophorus meadow vegetation type, in which most research was concentrated, gross primary productivity was 450 gdw m"^ yr"', including 414 gdw m'^ yr"' for vascular plants and 36 gdw m"^ yr"' for mosses. Net primary productivity was 209 gdw m"^ yr"', including 187 gdw m"^ yr"' for vascular plants and 22 gdw m"^ yr"' for mosses. The res- piratory cost for maintaining the above- and belowground vascular bio- mass was 170 gdw m'^ yr"' and the respiratory cost for growing new bio- mass was 57 gdw m"^ yr"'. For mosses the maintenance and growth costs were each 7 gdw m"^ yr"'. The gross primary productivity of vascular plants was equivalent to a carbon dioxide incorporation of 609 g CO2 m"^ yr-'. The plant canopy of the Carex-Oncophorus meadow interacts with various biophysical factors to affect production and water loss. Several features of the Carex-Oncophorus meadow canopy structure act to in- crease plant temperatures, which are usually below optimum for physio- logical processes in these tundra species. The steeply inclined leaves of the grasses and sedges and the accumulated standing dead material increase Biophysical Processes and Primary Production 101 the interception of radiation in the canopy, and by convecting this energy warm the canopy air. The increased leaf temperatures increase trans- piration, photosynthesis, biosynthesis and leaf expansion. Even in these wet soils, the species show water stress, indicated by high leaf resistances, and patterns of distribution related to soil water. Increased interception in the canopy causes decreased absorbed energy at the moss and ground surface, producing lower moss temperatures, lower soil temperatures and shallower thaw depth. The latter two effects should reduce growth at the stem base, growth of roots, and the amount of nutrients made available by decomposition. Reduced nutrient uptake and reduced activi- ty at the stem base should reduce leaf growth. This sequence of events should have a stabilizing influence on foliage area index. The low foliage area early in the growing season in the Carex-Oncophorus meadow, relative to an evergreen shrub canopy with the same foliage area at peak season, should increase soil thaw and nutrient availability and partially compensate for the increased energy required to thaw waterlogged soils. Photosynthesis L. L. Tieszen, P. C. Miller, and W. C. Oechel INTRODUCTION The interception of solar radiation and the conversion of that energy by photosynthesis into stable organic forms is essential for the mainten- ance and growth of plants as well as for their vegetative or sexual repro- duction. Accumulating information on photosynthesis of tundra plants suggests that this process is highly adapted to the extreme conditions of the tundra. This chapter describes photosynthesis in the coastal tundra at Barrow and the sensitivity of carbon dioxide assimilation to abiotic and biotic factors. Response patterns and internal and external controls over photosynthesis in vascular plants and mosses are described in an attempt to quantify those factors that govern rates of carbon dioxide uptake. The objectives of Chapter 4 are to understand the controls over photosynthe- sis, analyze the sensitivity of the system, and estimate community pro- ductivity as reviewed in Chapter 3. Photosynthesis is a photochemical, diffusion, and enzymatic pro- cess with a rate controlled by intrinsic and extrinsic factors. The process is basically similar in all vascular plants and mosses, although variations in component dark reactions have evolved and are most notable in the dis- tinctions between C3 and C4 plants (Hatch et al. 1971). Tundra vegetation consists mainly of C3 plants (Tieszen and Sigurdson 1973), and no signif- icant differences would be expected in the basic mechanisms between Cj plants in the Arctic and d plants in more temperate climates. However, component reactions, e.g. at the enzyme level, have probably evolved and could be manifest as quantitatively different response patterns. The amount of carbon dioxide assimilated is a function of the maxi- mum capacity (rate) for carbon dioxide uptake, which may be related to intrinsic factors such as component enzyme levels (Treharne 1972), the concentrations of ribulose-l,5-diphosphate, nutrient status, innate leaf growth, or development patterns. The extent to which this maximum 102 Photosynthesis 103 capacity is realized is a function of the microenvironment within the can- opy (see Chapter 3.)- At the cell and leaf level, responses to light, temper- ature and water are most crucial in determining the rate of photosynthe- sis. Nutrients can limit leaf carbon dioxide uptake at the cell level by affecting the internal capacity and at the plant and canopy level by affecting the allocation for the production of more photosynthetic tis- sues. Similarly, grazing will alter this pattern directly by removing estab- lished tissues at various developmental stages. Most of these interactions are discussed in this chapter and have been incorporated into a canopy photosynthesis model (Miller et al. 1976). INTRINSIC FACTORS AFFECTING CARBON DIOXIDE EXCHANGE Maximum Rates and Growth Forms The maximum photosynthetic rates for expanded blades range be- tween 7 and 31 mg CO2 dm"^ hr"' (Table 4-1) or nearly as widely as those of vascular plants in other Biomes (Tieszen and Wieland 1975). Within the tundra, however, similar species, e.g. Carex spp. at Barrow and Devon Island, Canada (Mayo et al. 1977), show very comparable rates. Furthermore, photosynthetic rates show a distinct relationship to growth forms. The rates are highest in graminoid types and forbs (~30 mg CO2 dm"^ hr"'), sHghtly lower in some of the deciduous dwarf shrubs ('^^20 mg CO2 dm"^ hr"'), except for Salix species which tend to be higher, and still lower among the evergreen dwarf shrubs (7 mg CO2 dm"^ hr"') such as Cassiope tetragona, Ledum decumbens, and Vaccinium vitis-idaea (Johnson and Tieszen 1976). Photosynthesis rates in vascular plants are equivalent to rates of similar growth forms in more temperate zones (Table 4-2), suggesting that these species are adapted genetically or physi- ologically to the low ambient temperatures. As expected, photosynthetic rates are much lower in mosses than in vascular plants (Table 4-2), rang- ing between 1.0 and 4.4 mg CO2 gdw"' hr"' (Oechel 1976, Oechel and Collins 1976, Oechel and Sveinbjornsson 1978). Although the rates of different moss species vary widely, they are similar to those of temperate, Photosynthesis, mg COg dm' ^hr~' 8 10 12 14 16 FIGURE 4-1. Photosynthetic ~\ — ' — I — ' — 1 — I — I — ' — I rates of various plant parts of Leaf Sheath Exposed Culm Dupontia fisheri near mid-sea- son. The rates were determined ^infiorescence ^'^^^ ^he '^C system. N = 8. Enclosed Culm (After Tieszen and Johnson 1975.) 104 L. L. Tieszen et al. TABLE 4-1 Maximum Photosynthetic Rates of Field-grown Tundra Plants at the Tundra Biome Research Area under Ambient Light and Temperature Regimes Species Leaf area basis (mg CO2 dm"' hr" Dry weight basis (mg CO2 gdw' hr') Barrow Graminoids' Alopecurus alpinus 16 Arctagrostis latifolia 14.7 Arctophila fulva 19.6 Calamagrostis holmii 12.6 Carex aquatilis 18.5 Dupontia fisher i 17.1 Eriophorum angustifolium 20.9 Elymus arenarius 30.8 Hierochloe alpina 7 Poa arctica 11.5 Poa malacantha 10.1 Forbs' Petasites frigidus 13.4 Ranunculus nivalis 18 Deciduous dwarf shrubs' Salix pulchra Mosses' Pogonatum alpinum Calliergon sarmentosum Polytrichum commune Dicranum angustum Dicranum elongatum 28 35 37 34 33 24 25 33 12 14 17 21 4.4 2.7 2.9 1.0 1.3 Niwot Ridge Graminoid' Deschampsia caespitosa 21.3 Forbs^ Geum rossii 26.7 Kobresia myosuroides 21.3 'Tieszen (1973, 1975, unpubl.) and Tieszen and Johnson (1975. 'Oechel (1976), Oechel and Collins (1976), and Oechel and Sveinbjornsson (1978). 'Johnson and Caldwell (1974). Photosynthesis 105 TABLE 4-2 Maximum Photosynthetic Rates of the Major Plant Growth Forms Among all Biomes Leaf area basis Dry weight basis Species (mg CO: dm-' hr') (mg CO2 gdw' hr') Herbaceous plants Cultivated with C, pathway' 20-35 30-60 Herbs from sunny habitats' 15-60 30-90 Herbs from shaded habitats' 4-16 20 Tundra graminoids^ 7-31 18 Tundra forbs' 13-18 15 C4 plants' 30-70 40-120 Succulents' 4-12 8 Submerged macrophytes' 4-6 Woody plants Deciduous broad-leaved trees' Sun leaves 10-25 15-30 Shade leaves 6-15 Tundra deciduous dwarf shrubs. 13 15 average^ Tundra evergreen dwarf shrubs, 7 5 average^ Evergreen broad-leaved trees' Sun leaves 10-16 6-10 Shade leaves 3-8 Semi-arid sclerophyllous shrubs' 4-12 4-6 Evergreen conifers' 4-12 3-15 Mosses Tundra mosses'" 0.1-4.4 Temperate mosses* 1.1-3.5 Temperate epiphytic mosses* 0.6-1.5 Lichens" 0.3-3.9 'Sestak et al. (1971). Tieszen et al. (1981). 'Kallio and Heinonen (1973), Oechel (1976), Oechel and Collins (1976), Oechel and Sveinbjornsson (1978). 'Kallio and Karenlampi (1975). subarctic and antarctic mosses (Stalfelt 1937, Hosokawa et al. 1964, Rastorfer 1972, Kallio and Karenlampi 1975). The proportion of shrubs decreases and that of graminoids increases with increasing latitude in tundras. This may reflect the higher ratio of potentially productive to supporting tissue (e.g. stems) in the graminoid growth form. In Dupontia the leaf is obviously the most important pho- tosynthetic component; however, other components (Figure 4-1) are photosynthetically active and contribute to the total amount of carbon 106 L. L. Tieszen et al. dioxide incorporated. Mosses and lichens represent an extreme develop- ment of this trend since nearly all tissues are photosynthetic. Thus at high latitudes plants are selected which either have little nonphotosynthetic tissue or are highly opportunistic in their CO2 uptake. Enzyme Levels and Component Resistances The maximum rates of carbon dioxide uptake among all vascular species are highly correlated with specific leaf density or thickness (r = + 0.83) and with carboxylation activity {r = +0.76, N = 54) (Tieszen 1973). Chabot et al. (1972) noted an acclimation response of Oxyria digyna that resulted in higher carboxylation levels at low temperatures, and Treharne (1972) suggested a causal relationship between carboxyla- tion activity and photosynthesis. Their data suggest that the range in car- bon dioxide uptake potential is determined by differences in carboxyla- tion activity. Further support is provided by data from the Biome research area, which showed high correlations between photosynthesis near light saturation and carboxylation activity among all leaves throughout the season {Dupontia, r = +0.74, p > 0.97; Carex, r = + 0.81, /7 > 0.99; Eriophorum, r = +0.75, p > 0.99). Therefore, species differences and seasonal patterns are directly related to carboxylation ac- tivity. Since ribulose-l,5-diphosphate carboxylase is a substantial por- tion of total cell protein (Huf faker and Peterson 1974), this enzyme also accounts for the major changes of nitrogen content through the season. The high correlation of maximum photosynthesis with carboxyla- tion activity further suggests that differences in photosynthetic rates are related more to differences in some component of the mesophyll resis- tance than to leaf resistance. In the field, minimum leaf resistances for Dupontia are generally less than 2 to 3 s cm"', whereas minimum meso- phyll resistances are rarely below 7 s cm"' and are often well above 12 s cm"'. Similar values for Carex, Eriophorum angustifolium, Salix pulchra and Petasites frigidus support this contention. This trend is even more pronounced in mosses, where leaf resistances are generally less than 1 s cm'' but mesophyll resistances are large (Oechel and Sveinbjdrnsson 1978). Growth Rate and Developmental Stage Photosynthetic competence is a function of leaf development, in- creasing as the leaf elongates or expands until a mature stage is attained. The leaf usually remains at full competence until senescence occurs and carbon dioxide uptake ability decreases as proteins and other materials Photosynthesis 107 are hydrolyzed and mobiHzed. Obviously, the dynamics of leaf photo- synthesis will vary with plant growth forms as patterns of leaf develop- ment and retention vary. In Dupontia exsertion is followed by an elonga- tion period of 20 to 22 days, followed by a shorter period of 8 to 10 days during which the growth rate is near zero. At the end of this period, the leaf initiates senescence and in about 25 to 30 days it is dead. Thus, in comparison with other growth forms where the mature phase may last more than one growing season, Dupontia has a short period of maximal photosynthesis (Johnson and Tieszen 1976). Carex and Eriophorum an- gustifolium have somewhat longer mature periods than Dupontia, while moss tissue may remain photosynthetically active for at least 3 years (Collins and Oechel 1974). Photosynthetic activity of vascular plants does not occur beneath the winter snow even though substantial carboxylation activity is present (Tieszen 1974). Thus photosynthesis begins concurrently with growth, which is initiated within one day of snowmelt. This has now been con- firmed in the Arctic not only for graminoids but also for Dryas (Mayo et al. 1977), which remains inactive until snowmelt. This is not unexpected since the plant temperatures beneath late-winter snow may approximate the permafrost temperature, thereby presenting a distinct contrast with conditions that may occur in mid-latitude alpine areas. As mehwater per- colates through the snowpack, however, temperatures abruptly approach 0°C (Tieszen 1974). Following snowmelt, leaf expansion and growth of Dupontia occur rapidly and are accompanied by the development of photosynthetic com- petence. The first leaf elongates and exserts some chlorophyllous tissue produced the previous season. This tissue never becomes very active al- though it does make a positive contribution to the carbon balance. By about 19 June, however, the second and third leaves have elongated and they are soon active (see Chapter 5). Although the sequential pattern of photosynthesis is somewhat obscured by the short growing season, suc- cessive leaves become more active as the season progresses. This general ontogenetic leaf pattern is similar to that of other graminoids, and results in a sequence of developing photosynthetic competence as leaves elon- gate or enlarge, a period of maximal photosynthetic competence associ- ated with maturation, and a subsequent decline in photosynthetic com- petence as senescence develops (Johnson and Caldwell 1974, Johnson and Tieszen 1976, Tieszen 1978b). In a short growing season a sequential leaf pattern seems costly since it requires a large investment in synthetic and growth processes (see Chapter 5). Although it does replace leaves at successively higher posi- tions in the canopy in more favorable radiant flux (but less favorable thermal) environments, this pattern must have other selective value, e.g. as a mechanism for withstanding acute or chronic grazing pressures. 108 L. L. Tieszen et al. o E o J3 O O Live ond 1.0 - /\D.f. Deo^t..-- - ^A« \ 0.8 : / 0.6 0.4 0.2 - // / / r/' // // V. \ 0 1 1 1 rf^ / 1 1 . 1 . 1 10 20 10 20 10 20 Jun Jul Aug FIGURE 4-2. Seasonal progression of the carboxylation activity for entire tillers (summation of leaf activity times leaf area) in 1971 of: Dupontia fisheri (D.f.), Carex aquatilis (C.a), Eriophorum angusti- folium (E.a.), and the total foliage area index (A,) of the community. The foliage area indices of the species are given in Figure 3-3. Ab- solute rates not directly comparable among species. (After Tieszen 1978b.) On both a daily and seasonal basis, leaf photosynthetic rates are highly correlated with carboxylation activity (Tieszen 1978b). The inte- gration of carboxylase activity among all leaves suggests that on a tiller basis the greatest potential for photosynthesis occurs well before the time of maximum standing crop or leaf area index (Figure 4-2). Although the greatest conversion efficiency on a green leaf area basis should occur on 20 July, the increased canopy or tiller density later in the season results in a greater efficiency on a land basis. Mosses begin each season with a high proportion of chlorophyllous tissue which may equal 50% of the maxi- mum for the season. This tissue is photosynthetically competent under the snow, with potential in situ photosynthesis rates of about 25% of the normal seasonal maximum (Tieszen 1974, Oechel and Sveinbjbrnsson 1978). In contrast to vascular plants, the photosynthetic moss tissue does not decrease in activity during the growing season (Oechel 1976), but net photosynthesis is reduced as the tissue ages. Early in the season photo- synthesis is carried out by tissue from the previous 1 or 2 years. This pat- tern permits early season photosynthesis but at rates for 2- and 1 -year- old tissue of only 40% to 75% of new tissue, respectively (ColHns and Photosynthesis 109 Oechel 1974). As the season progresses, new tissue is produced with high photosynthetic capacities. The moss growth pattern has the potential for significant late season photosynthesis since no end-of-season senescence is observed. However, mortality of older age classes is high, and the amount of older tissue decreases markedly at ages greater than 1 year (Collins and Oechel 1974). Nutrients Nutrients can limit photosynthesis at the leaf and plant level if the allocation for photosynthetic structures exceeds the support capabilities of available nutrients. Under field conditions Dupontia appears to con- trol allocation to produce a complement of photosynthetic structures operating at near optimal capacities. The main response to chronic and intense fertilization (Schultz 1964) was an increase in productivity due to the stimulation of greater plant density (Dennis et al. 1978) and a two times greater leaf area index. Although fertilizer stimulated a slight in- crease in leaf width in Dupontia, there was no significant difference in carboxylation activity and presumably no difference in leaf photosyn- thesis. The short-term responses at site 2 were similar (Dennis et al. 1978) and resulted in statistically significant, but small, increases in plant phos- phorus and potassium, but not nitrogen (Chapin et al. 1975). In an attempt to document the spatial variability of photosynthesis and to determine the extent to which large changes in production were associated with changes in photosynthetic rates, a study was made along a productivity and growth form gradient (Tieszen, unpubl.). Although aboveground production ranged from 21 g m"^ in the basins of low- centered polygons to 215 g m"^ in a disturbed vehicle track, there were no significant (p = 0.95) correlations among photosynthesis and soil or leaf potassium, nitrogen and phosphorus (Tieszen 1978b) (Table 4-3). This TABLE 4-3 Range of Nutrient Concentrations in Leaves (%) in which Photosynthesis was Independent of Leaf Nutrient Concentration (P = 0.95) Dupontia Carex Eriophorum fisheri aquatilis angustifolium Nitrogen 1.83-3.28 2.74-3.28 1.50-3.21 Phosphorus 0.07-0.24 0.07-0.40 0.15-0.31 Potassium 0.64-1.59 0.55-1.55 0.45-1.14 110 L. L. Tieszen et al. independence of photosynthesis and potassium, nitrogen and phosphor- us over a large range of field concentrations provides strong evidence for precise control of allocation in response to available nutrients. Plants of the coastal tundra at Barrow do not appear to produce additional leaf area unless they can operate at near maximal capacity. Thus, under field conditions, plants seem to avoid nutrient limitations of photosynthesis by Hmiting the amount of photosynthetic tissue within the support capa- bilities of the available nutrients (see also Chapter 5). Ulrich and Gersper (1978), however, show that these plants are always on the borderline of being nutrient-limited; and the addition of phosphorus and nitrogen clearly stimulates production (Chapin et al. 1975, Dennis et al. 1978). EXTRINSIC FACTORS AND THE RATE OF PHOTOSYNTHESIS Light Arctic tundras have often been described as light-limited ecological systems. Daily totals of irradiance can be high at Barrow, but instantane- ous irradiances are generally low because of the low sun angle and fre- quent cloudiness. Light response curves for vascular plant species are TABLE 4-4 Irradiance (300 to 3000 nm) Required for Photosynthesis to Equal Respiration as Determined from Field Measurements at Temperatures near 0°C Irradiance Species (J m'^ s"') Vascular plants Dupon tia fisher i 1 6 . 7 ± 2 . 1 Carex aqua t His 9. 1 ± 0.7 Eriophorum angustifolium 9.1 ±0.7 Salix pulchra 14.0 ±2.1 Vascular plant mean 12.6 ±3.5 Mosses Pogonatum alpinum 9.1 ±2.1 Calliergon sarmentosum 1 0. 5 ± 3 . 5 Dicranum angustum 5.6 ± 2.8 Dicranum elongatum 10.5 ±4.2 Polytrichum commune 1 1 .2 ± 1 .4 Moss mean 9.1 ±2.1 Photosynthesis 1 1 1 similar (Tieszen 1973) and tend to approach saturation at 280 to 350 J m"^ s"' (400 to 700 nm). These saturation requirements are sufficiently high that leaves are rarely light-saturated in situ, which suggests that the entire canopy might be responsive to increased irradiance. Individual leaves of tundra plants require very low light for carbon dioxide compensation, 5.6 to 7 J m"^ s"' (400 to 700 nm) (Tieszen 1973, Mayo et al. 1977). Under field conditions, whole shoots possessed simi- larly low compensation requirements (Table 4-4), especially at low tem- peratures. Although the respiratory capacities of these tundra plants are high, the combination of efficient photosynthesis and low daily tempera- tures often resulted in the maintenance of a positive carbon budget for 24 hours (Tieszen 1975). The close coupling of the daily course of carbon di- oxide uptake to irradiance implies a direct dependence even during mid- day hours. This is further documented by the significant positive regres- sion between daily carbon dioxide uptake and daily irradiance which is discussed later. This light dependence may be mainly a vascular plant phenomenon 100 200 300 Irradiance, J m s ' 100 200 -? -I Irradiance, J m s FIGURE 4-3. The response of photosynthesis to irradiance (400-700 nm) in Dupontia fisheri (D.f.), Pogonatum alpinum (P. a.), and Calliergon sarmentosum (C.s.). The curves for D. fisheri are from the field (o) and from plants grown in the laboratory at 5°C (%). The curves for the mosses are from field-collected samples measured in the laboratory. Note the different vertical scales. Standard errors are shown by the ver- tical bars. (After Tieszen 1974, 1975, Oechel and Collins 1976.) 112 L. L. Tieszen et al. since arctic mosses tend to reach light saturation at lower radiant fluxes than vascular plants— 98 J m'^ s"' (400 to 700 nm) (Figure 4-3). Mosses are generally light-saturated for most of the midday periods. The tend- ency for light intensities above saturation to reduce the rate of photosyn- thesis in Pogonatum alpinum may be a result of photo-inhibition or photo-oxidation of the photosynthetic apparatus (Oechel and Collins 1976). This pattern is in contrast to graminoids, which increase photo- synthesis to radiation levels approaching full sunlight, and represents a major response difference between Pogonatum alpinum, especially populations from low light environments, and the graminoids. The dif- ferent light saturation requirements result in different daily responses be- tween mosses and vascular plants. During the season, mosses show only a slight daily dependence of photosynthesis on total daily irradiance (Oechel and Sveinbjornsson 1978), in marked contrast to vascular plants (Tieszen 1975). Temperature An effective photosynthetic system at low ambient temperatures is essential for maintaining a positive carbon balance. Although mean am- bient air temperatures during the growing season are less than 4°C and graminoid leaf temperatures are closely coupled to air temperatures, de- tailed studies at Barrow (Tieszen 1973, 1978b) and in other areas (Mayo et al. 1977) have shown a temperature optimum for leaf photosynthesis between 10 and 15 °C and significant carbon dioxide uptake at 0°C. Photosynthesis in Dupontia is generally active until the leaf freezes, which may not occur until -4 to -7°C. However, the destruction of en- larging cells was observed at a temperature of -4°C, which usually repre- sents the lower limit of photosynthesis. Very low temperatures are infre- quent in July and do not appear to affect carbon dioxide uptake as sig- nificantly as other growth processes. The underlying physiological and biochemical bases for the temper- ature response curve are not clear. All resistances in the graminoids, in- cluding leaf resistance, remained low down to 5 °C (Figure 4-4). Leaf re- sistance did not increase at 0°C. At higher temperatures there was a slight increase in leaf resistance but the mesophyll components of resist- ance became significantly more important, indicating an internal diffu- sion or carboxylation limitation to photosynthesis. A substantial increase in light respiration may account for the net photosynthesis decrease at temperatures greater than 15 °C. Mosses also have relatively high photosynthetic rates at low temper- atures (Figure 4-5). Temperature responses are similar to those observed in vascular plants with temperature optima between 10 and 19°C (Oechel Photosynthesis 113 £ ^ 30 (/) c o o~ 25 o ^ 20 o u> In 0= 15 10- 5- 300 -2 -l Irradionce, J m s ' 1 1 1 1 1 - r- • • • • _ / / .• / _ •• / // .•/ .•/ - .•/ / ^ — /• ^ — _ / / ^y"^ ^-____ T 1 20 Temperature, °C 30 FIGURE 4-4. Relationships of the resistance to CO2 transfer to irradiance and temperature. The resistances include leaf (stomatal and cuticular) ( — ), cell wall ( — ) and carboxylation (...). The plants were grown in growth chambers at 5°C. (Tieszen, unpubl.) 1976, Oechel and CoUins 1976). Rates are only slightly decreased at 5°C. Pogonatum photosynthesizes at 55% of the maximum rate at 0°C. The high rates of photosynthesis at low temperatures are of obvious adaptive significance, since tissue temperatures frequently drop to between 5 and 0°C during the growing season. Continuous sunlight results in positive net photosynthesis during these periods. Temperature optima from 10 to 19°C seem high; however, during periods of high irradiance, moss tissue temperatures exceed air tempera- ture. In 1973, midday tissue temperatures were above 5°C 87% of the time and above 10° 44% of the time. In 1972, which was warmer and 114 L. L. Tieszen et al. E ■o (\i O o E 0) £L C >> M O 10 20 Temperature, "C 6 - 'E O 4 o 5 - E IV c I - £ 0 - - r/ .-o- ■^ c.s. - P.O. c/ X ^ -A- c ^ ,4. ■ - ^ C.S. .^-'^ •^ \ : , — 1_ 1 1 1 1 1 1 . 30 6 ^ E c* o o h- o. a> a. 0 10 20 Q. Temperoture, °C FIGURE 4-5. The response of photosyn- thesis ( — ) and dark respiration ( ) to temperature in Dupontia fisheri (D.f.), Po- gonatum alpinum (P.a) and Calliergon sar- mentosum. (C.s.). The curves for D. fisheri are from the field (o) and laboratory (m). The curves for P. alpinum and C. sarmen- tosum are for field-grown samples meas- ured in a field laboratory. (After Tieszen 1973; unpubl.; Oechel and Collins 1976.) drier, temperatures were above 10°C at midday 73% of the time and reached as high as 30 to 35 °C. However, in 1974 tissue temperatures above 20 °C were seldom measured (Oechel 1976, Oechel and Collins 1976). The broad temperature responses of arctic bryophytes make them well adapted to the wide range of tissue temperatures encountered in the Arctic. However, at least in the case of Dicranum, simulation modeling indicates that the relatively high values for temperature optima for pho- tosynthesis result in a seasonal depression of photosynthesis of about 25 x: c o Z I I I I I I I I I J I I L J L 12 16 20 24 0 4 8 Alaska Standard Time 12 16 20 24 FIGURE 4-8. Diurnal patterns of CO2 flux for alternate 10-day periods through the 1973 growing season simulated for Pogo- natum alpinum (P. a.), Calliergon sarmentosum (C.s.), Dicran- um elongatum (D.e.), anc^Dicranum angustum (D.a.). The envi- ronmental input was the 10-day average for the hour simulated. Periods began on 24 June (A), 14 July (B), 3 Aug (C), 23 Aug (D). and 12 Sept (E). (After Miller et al. 1978a.) elation of the linear equation relating daily totals of carbon dioxide up- take to radiation suggests that for the three graminoids slightly less than 4.2 MJ m"^ day"' is required to compensate for daily aboveground respir- atory carbon dioxide losses. The value is somewhat greater than the com- pensation points actually measured during night runs, and may suggest a higher respiration rate during daytime than at night. Late in the season the combination of shorter photoperiods, reduced irradiances, a developing senescence, and self shading results in a de- crease in the daily incorporation of carbon dioxide. Thus, by 25 August, daily photosynthetic totals for some graminoids are well below 50 mg CO2 dm-^ day-'. A multiple linear regression analysis (Tieszen 1975) suggests that in all species there is a highly significant change in photosynthesis which is independent of the seasonal changes in radiation for the entire plant, which could be caused by an increase in the proportion of supporting or other non-chlorophyllous tissues, developing senescence, or other phen- omena. The overall seasonal trend of photosynthesis is one of decreasing Photosynthesis 119 diurnal ampHtude and of decreasing photosynthetic input. However, photosynthetic efficiency computed on a land area basis remains high and attains its maximum between the middle of July and the first week in August, which agrees with tiller carboxylation data (Figures 4-2 and 5-16). Maximum efficiencies are above \% for the graminoids and attain 2.3% for Salix pulchra. Early in the season, while light intensity is high and often above saturation, mosses under simulated canopies show rates of carbon diox- ide incorporation similar to those of mosses growing in open areas. Dur- ing this period, protection from photoinhibition and higher rates at mid- day under reduced sunlight offset the effects of reduced levels of carbon dioxide incorporation during the evening. However, as the light intensity decreases, especially during the period around solar midnight, and as dark respiration increases, relative rates of carbon dioxide incorporation by mosses under the canopy decrease. When midday radiation values are high, advantage is conferred through shading, but if midday radiation values are below saturation, there is a lowering of photosynthetic rates at midday in response to shading. During evening hours photosynthesis is also lowered by shading, often below the compensation point, and the period of dark respiration is increased as a result. The mosses differ from vascular plants in their levels of energy cap- ture. The mosses are much lower in overall efficiencies than are vascular plants, except under periods of low radiation when the percentage of energy capture increases. They also differ from vascular plants in that mosses show no decreasing efficiencies at the end of the season resulting from senescence. Moss photosynthesis shows less seasonal variation than does vascular plant photosynthesis, but both have equally marked diur- nal changes. Light intensity as well as water status are important control- ling factors in moss photosynthesis. Other Factors Other factors could potentially influence photosynthesis and alter the daily and seasonal courses just described. Plant pathogens, for exam- ple, commonly inhibit photosynthesis by damaging chloroplasts and/or by destroying proteins. Pathogens are not obvious on plant species of the tundra at Barrow. One of the most striking impressions given by the veg- etation is the absence of leaf lesions. Root nematodes are present and fungi become active after the leaf senesces, but neither of these relation- ships affects photosynthesis directly. Grazing may also influence the photosynthetic response patterns, but mainly by altering the relative number of young, mature and senescent leaves. These phenomena and the role of accUmation are discussed in the following sections. 120 L. L. Tieszen et al. SIMULATION ANALYSIS OF PHOTOSYNTHESIS AND VASCULAR CANOPY INTERACTIONS Models Vascular Plant Previous discussions of photosynthesis have considered the physio- logical responses of single leaves, tillers or moss mats to independent biotic and abiotic factors. In plant communities these factors do not operate independently. They result from complex feedbacks involving regional cHmate, canopy structure, and the process of carbon dioxide up- take. In an attempt to quantify these relationships and to determine their relative and absolute importance, they have been incorporated into an in- teractive model called Stand-Photosynthesis which is basically an out- growth of models discussed earlier (Miller and Tieszen 1972, Miller et al. 1976, Stoner et al. 1978b). The stand photosynthesis model is based on the fluxes of carbon dioxide, water, and heat for a single leaf located in the canopy in profiles of direct and diffuse solar and infrared radiation, wind, air temperature, and vapor density. The canopy consists of horizontal strata of live and dead leaves, stems, and reproductive structures. The vegetative canopy produces profiles of solar and infrared radiation, by intercepting, ab- sorbing, and emitting radiation, which are calculated for each stratum. Similarly, the canopy effects on wind, air temperature, and humidity are V 1.2 o o 0) o o 0.8 0.4- -0.4 0400 0800 1200 1600 2000 2400 0400 0800 1200 1600 2000 2400 Solor Time FIGURE 4-9. Comparison of net CO2 exchange from field cuvette (o), simulation (x), and aerodynamic estimates (%). The aerodynamic data represent flux from the atmosphere only. The cuvette measurements and simulations are for vascular plants. Cuvette data from Tieszen (1975), aerodynamic data from Coyne and Kelley (1975), simulations from Miller et al. (1976). Photosynthesis 121 calculated as described previously. The energy budget is solved for each stratum, and the partitioning of energy exchange by convection and transpiration is determined for stems and leaves. Transpiration is the re- sultant relative saturation deficit which affects leaf water potential and thereby leaf resistance and the carbon dioxide diffusion pathway. Photo- synthesis on a leaf area basis is calculated for sunlit and shaded leaves and stems in each stratum. In this model the internal resistances depend on solar radiation and temperature. Solar Absorbed, J IT) s Net Photosynthesis, rrig COj dm^ hr'' 0600 1200 laoo 0600 1200 1800 Leaf Water Potentiol, bars Leaf Resistance, s cm' Meon Leaf Temperature, °C 0600 1200 1800 0600 1200 1800 Solar Time 0600 1200 1800 FIGURE 4-10. Isopleths showing the simulated daily course of various plant responses to environmental conditions through the canopy for a 5-day period beginning 15 July 1971. (After Miller et al. 1976.) 122 L. L. Tieszen et al. The output for the canopy model was vaHdated with production data (Miller and Tieszen 1972) and more recently with photosynthesis data from both field cuvette experiments and an assessment of commun- ity carbon dioxide exchange (Miller et al. 1976). Daily courses were gen- erally similar (Figure 4-9) as were the estimates of seasonal incorpora- tion. Production data were simulated for various periods throughout a growing season and, in some cases, for a variety of seasons. Sensitivity analyses of several environmental parameters were made with a standard day (Figure 4-10) that represents the mean input data for the 5-day period beginning 15 July 1971. The standard day represents midseason conditions in 1971; mean temperatures and solar radiation used in the model are near the means for the four years of the field program. Moss The moss simulation model is similar in concept to the vascular plant model (see Miller et al. 1978b). Photosynthesis and transpiration follow from the solution of the energy budget equation, with the inclu- sion of appropriate physiological relations. The input climatic data con- sist of solar and infrared irradiance, air temperature, air humidity and wind speed. The vascular canopy is composed of leaves, stems and stand- ing dead material of different species, each defined by inclination and by vertical profiles of area per unit area of ground. Solar and infrared radia- tion from the sun and sky are intercepted by the canopy and produce profiles of direct, diffuse reflected solar, and reflected infrared radiation within the canopy. The air temperature and humidity above the canopy at the moss surface interact with the canopy structure, wind profile, and radiation profiles to produce profiles of air temperature, humidity and leaf temperature. At the moss surface the receipt of net radiation is balanced by heat exchanges due to convection, evaporation and conduction. The convec- tional heat exchange occurs by turbulent exchange of air from the sur- face across a surface boundary layer and across a bulk canopy air layer to a reference height in the canopy. Surface evaporation is related to the turbulent exchange of water vapor across the surface boundary layer and bulk canopy air layer. Moss photosynthesis is related to solar irradiance, tissue tempera- ture and water status. Solar and infrared irradiance, air temperature, hu- midity, and wind velocity affect the plant water status through their ef- fect on leaf temperature and transpiration. The plant water status influ- ences the rates of transpiration and photosynthesis through their com- mon resistance to water and carbon dioxide diffusion. Water, in the form of precipitation and dew, that is not intercepted by the vascular Photosynthesis 123 canopy is added to the moss surface water film. Water flows between the green moss surface and the non-green moss and peat layers below. Sur- face water can be evaporated directly or absorbed into the moss tissue, to be lost by transpiration later. The seasonal progression of microchmate and production was simu- lated using a standard set of climatic conditions and deviations from the standard. The standard input cUmate is based on climatic and microcli- matic data collected at the Biome research area during summer of 1973. Climate data were adjusted using long-term records to produce two other sets of conditions which were each 3 standard deviations above or below the standard temperature conditions. The two contrived chmates are hereafter referred to as hot and cold. The microclimatic data from 1973 have been used in other studies (Ng and Miller 1975, Stoner and Miller 1975, Miller et al. 1976, Ng and Miller 1977, Stoner et al. 1978b) and were used here as the standard case to aid in interpreting the results. Simulation Results Temperature Relationships Analyses suggest that the vascular plant photosynthetic system has adapted to function at near maximal capacities under existing tempera- tures of the coastal tundra at Barrow while maintaining a leaf tempera- ture optimum above mean ambient temperatures (Figure 4-11). Large amounts of leaf area occur in positions of the canopy where tempera- tures are ameliorated at times of the day when irradiances are high enough for carbon dioxide uptake to respond to temperature. Further- more, since within-season temperature changes are small, the results sug- gested that seasonal acclimation responses, i.e. compensatory shifts in the response curve, are not necessary to maintain high daily photosynthe- sis rates. Strong accUmation responses have not been seen in the Barrow tundra plants (Oechel and Sveinbjornsson 1978, Tieszen 1978b). As the temperature optimum for photosynthesis increases, the stra- tum which supports the highest photosynthetic rates on a leaf area basis shifts to lower levels in the canopy (Figure 4-12). However, the total daily photosynthesis rate of each stratum remains similar, because strata with high irradiance and potentially high photosynthetic rates have less leaf area than do strata at the base of the canopy. The model simulations sug- gest that dwarf shrubs and cushion plants should be characterized by higher optimum temperatures for photosynthesis than the graminoids, which are more closely coupled to ambient air temperatures. In part, plants are capable of effective photosynthesis because the response curves are broad enough that with an optimum of 15°C rates 124 L. L. Tieszen et al. 0 2 4 6 e 10 12 Mean Daily Air Temperature, "C FIGURE 4-11. The simulated effect of various temperature optima for photosynthesis on daily net photosynthesis of Dupontia fisheri and Di- cranum elongatum at varying mean daily air temperatures. Also presented is the frequency of days with the indicated mean temperatures during 15-30 June, July and August, 1970-73. (After Oechel et al. 1975 and Miller et al. 1976.) Photosynthesis 125 0600 1200 1800 0600 1200 1800 Solor Time 0600 1200 1800 FIGURE 4-12. Isopleths of simulated net photosynthesis (mg CO2 dm~^ hr^J through the day at different levels in the canopy, using the stand photosynthesis model with different temperature optima. (After Miller et al. 1976.) are still positive at 0°C. The significance of the breadth and shape of the response curve is illustrated by simulations in which the curve was en- hanced or depressed at O^C. When the response at low temperatures was increased so rates at 0°C were increased lOO^o, the daily increase was on- ly 7%. When the rates were decreased by 50% at 0°C, daily photosynthe- sis was reduced only 3%. If, however, the curve was depressed so carbon dioxide uptake at 0°C was zero, daily uptake was reduced 27%. Thus a photosynthetic capability at 0°C is very important, and it appears that the Dupontia response curve is well-adapted to prevailing temperatures. Carbon dioxide uptake is much more temperature-sensitive in moss- es than in graminoids. In simulations where the temperature optimum for photosynthesis is varied and the shape of the response surface held constant, a temperature optimum of 5°C yields the highest uptake rates for the most frequently observed temperatures 1-3 °C (Figure 4-10). Up- take is suppressed only slightly at a temperature optimum of 10 °C under these conditions. However, temperature optima of 15 °C and higher result in large depressions in carbon uptake at ambient temperatures be- low 10 °C. The high temperature optima typical of the mosses result in significant losses of carbon. For example, the observed temperature opti- ma of 11 to 19 °C in Dicranum result in a seasonal carbon uptake 25% lower than that possible with lower optimum temperatures (Figure 4-13) (Oechel et al. 1975). Other moss species at the Biome research area show similar patterns (Oechel 1976, Oechel and Sveinbjbrnsson 1978). The temperature optimum in Dicranum elongatum does not acclimate season- 126 L. L. Tieszen et al. FIGURE 4-13. The simulated effect of different tem- perature optima on the seasonal course of net photo- synthesis in Dicranum elongatum. The environmental input data are from 1973. (After Oechel et al. 1975.) ally in a manner that would maximize carbon uptake, despite the fact that the low temperature optima necessary to maximize carbon uptake in Dicranum have been observed in Dicranum fuscescens in the Subarctic (Hicklenton and Oechel 1976). The controls on low temperature acclima- tion are not understood. The carbon uptake benefits of such acclimation or genetic adaptation are known, but the costs are not. Because of the temperature sensitivity in mosses, they should be af- fected more by climatic temperature changes than are the vascular plants. Compared to the standard climate, photosynthesis in the simu- lated cold climate described above decreased from 22% in Dicranum to 72% in Calliergon. Under the hot climate photosynthesis increased slightly in Dicranum, relative to the standard year. Photosynthesis de- creased in Calliergon, because of water stress induced by the higher tem- peratures. Although mosses are more temperature-limited than are vas- cular plants, increased temperatures may reduce the success of certain species of moss by creating an unfavorable moisture balance (Stoner et al. 1978b). Maximum Rates and Competency Simulations show that Dupontia is very sensitive to the light- saturated rate of photosynthesis (Figure 4-14). Doubling the rate of light- saturated photosynthesis results in a 58% increase in daily uptake of car- bon, whereas a reduction of the saturated rate to 25% results in a reduc- Photosynthesis 127 max •> 200 % of control FIGURE 4-14. Simulations of the effect of the light-saturated rate of photosynthesis (Pmax) on daily photosynthesis and maximum simulated rates using conditions for the standard day. tion to 3297o of the control rate. As was indicated earher, one of the com- pensatory adjustments of these vascular plants is the maintenance of photosynthetic rates comparable to temperate region plants. The model assumes maximum competency in all photosynthetic leaves, an assump- tion which leads to at least a sHght overestimate of uptake. The justifica- tion for this assumption was that developing leaves with incomplete competency are located near the base of the canopy where irradiance is reduced; the error resulting from light-saturated responses is therefore minimized. Irradiance The sensitivity of daily photosynthesis to daily irradiance (Figure 4-15) illustrates that the mean daily photosynthetic rate approaches satu- ration near 25 MJ m"^ day"'. Dupontia, therefore, approaches maximum daily photosynthesis rates at the upper range of daily intensities received. For major portions of the season the leaves are light-limited and would 128 L. L. Tieszen et al. 140 16.7 25. Radiation FIGURE 4-15. The simulated response of daily total net photosyn- thesis to daily irradiance and saturation irradiance (indicated below curve). Also indicated is the fre- quency of daily irradi- ancesfor 15-30 June, Ju- ly and August 1970-73. (Other input = standard day.) Simulations of the effects of varying re- quirements for satura- tion are also shown. (After Miller et al. 1976.) remain so even if they saturated at 140 J m"^ s"'. In the alpine similar response curves and equivalent daily total irradiance result in less carbon dioxide uptake because at solar noon, when the irradiances are substan- tially greater, plants become saturated, and water stress often develops. Vapor Density, Soil Water Potential, and Root Resistance Cuvette data and the simulations indicate that water stress develops to a sufficient extent to cause some stomatal closure and therefore occa- sional reductions in the photosynthesis of vascular plants. However, the Photosynthesis 129 field data and vascular plant simulations suggest that such water stress occurs infrequently and only when leaf temperatures and/or irradiances are significantly higher than the mean. Dupontia may have allocated suf- ficient resources to root absorptive tissue to meet the demands of the evaporative leaf surfaces. The water vapor density gradient is one of the factors determining the rate of water loss. The gradient is normally small because the air is nearly saturated and the leaves are close to air temperature. Simulations, in which air water vapor density was varied from -AQ^o to + 30% of am- bient, indicate a slight sensitivity of -9% to +9% change in photosyn- thesis to changes in the water vapor density gradient. This effect was related to lower leaf water potentials and higher leaf resistances as the water vapor density gradient increased. Transpiration losses also in- creased as the gradient increased, to the maximum simulated, resulting in an increase in water loss of 42*^0 in the upper part of the canopy, 26% in the center, and 21 % at the bottom. The greater increase at the top of the canopy was because the gradient of water vapor density from the air to the leaf became more important than the higher leaf temperatures at the bottom of the canopy. The slight increase in leaf resistance and the associated slight decrease in photosynthesis when transpiration changes are large suggests that root resistance in Dupontia is small relative to the water re- quirements. Increasing root resistance by 50% results in less than a 6% reduction in daily photosynthesis. The reduction is caused by a midday decrease of -2.6 to -3.8 bars in leaf water potential at the top of the can- opy and a decrease of -3.7 to -4.9 bars at the canopy bottom. However, leaf water potentials increase to standard day values near solar midnight as the plant regains its water deficit. These simulations strongly suggest that Dupontia is sensitive to periods of high water demand but that it can normally supply the amount of water required to keep stomates open. Although this is in agreement with data for Carex from both Barrow and Devon Island, it contrasts sharply with Dryas (Mayo et al. 1977), which shows decreased rates of uptake, associated with low water potentials, at high temperatures. Calliergon is more sensitive to a decrease in soil moisture levels than are the other moss species analyzed. For Dicranum elongatum, Dicran- um angustum and Pogonatum alpinum, simulations indicate a 25% decrease in photosynthesis associated with a 10-bar decrease in soil water potential (from 0 to -10 bars). Under the same conditions, Calliergon undergoes an 80% decrease in photosynthesis. Calliergon appears to be much more dependent than the other species on a liquid water film and on standing water to maintain a beneficial moisture status (Figure 4-16). Calliergon takes up water poorly from depth and displays low resis- tances to water loss when compared to the other species. In the Dicranum 130 L. L. Tieszen et al. e E X o 5 c 01 u 0. 75- 50- 25- ■V- — 1 r t -■■ - -\ ^~Ro~~*"~~- 1 1 1 1 -o 1 ^0 -10 -20 -30 Soil Water Potential, bors FIGURE 4-16. Simulated annual net photosynthesis at different soil water potentials expressed in absolute terms and as a percentage of maximum for four arctic moss species: Pogonatum alpinum (P. a), Calliergon sarmen- tosum (C.S.), Dicranum elongatum (D.eJ, and Dicranum angustum (Dm.). (After Miller et al. 1978a.) species examined, the mat growth form plays an important role in water retention by increasing the apparent air resistance to water loss and in aiding water uptake by maintaining a nearly saturated environment at the base of the photosynthetically active zone. In Pogonatum, xeromor- phic adaptations of tissues that are deeply rooted in the substrate and are efficient in translocating moisture are important in maintaining advan- tageous moisture balances under xeric conditions (Miller et al. 1978a). Calliergon 's susceptibility to xeric conditions is shown by the effect of a hot season. At soil water potentials of -5 bars and less, Calliergon shows net carbon dioxide loss in the hot climate. By comparison, Dicran- um and two vascular plant species show little suppression by photosyn- Photosynthesis 131 Calliergon Dicranum sarmenfosum anqustum 14 -1 ba r 7 0 -6 • 'E -12 14 1 1 \ 1 CM o o r -5 bar a> in 0) 7 0 -6 -12 14 -^^^^ c >. in o o 1 1 . 1 1 Q. -lObarj- Z 7 - 0 -6 1 0 — — — » 1 > - '^C ;old Std. Hot Cold Std. Hot Conditi ons FIGURE 4-17. Annual net photosynthesis for Calliergon sarmentosum and Dicranum angustum at different substrate water poten- tials for three simulated seasons and condi- tions ("cold, " standard, "hot") for moss in full sun ( — ) and under the canopy ( — / (After Stoner et al. 1978b.) thesis in the hot climate as compared with the standard climate (Figure 4-17). Canopy Architecture Effects Previous simulations and field data have shown the importance of physiological parameters and the dependence of photosynthesis on cer- tain environmental variables. Since these variables are influenced by structural features of the plants as well as by physical features of the can- opy, their potential influence on photosynthesis needs to be understood. 132 L. L. Tieszen et al. The graminoid leaf, especially in the single-shooted growth form, is closely coupled to air temperature. The temperature correspondence is mainly a result of the low boundary layer resistances associated with nar- row leaves and the generally turbulent wind conditions of the Biome research area. As leaf width increases, leaf temperatures should increase and more closely approach the temperature optimum for photosynthesis. However, at the same time, the water vapor density gradient between the air and the leaf, and the transpiration rate, increase, resulting in a poten- tially large water deficit and in a decrease in leaf water potential. Altering leaf width from 4 to 15 mm under standard conditions has no effect on carbon dioxide uptake. The radiation load is low; therefore the leaf tem- perature remains close to the air temperature. The rate of photosynthesis for any given leaf will also vary as a func- tion of its position in the canopy, since irradiance and leaf temperatures are markedly influenced by the canopy. Leaves at the top of the canopy protrude above the standing dead material and are occasionally light- saturated although they are usually at low temperatures. The trends are reversed for the leaves positioned at the bottom of the canopy. Because of the effect of canopy density on thermal and radiance properties, the foliage area index will determine the range of photosynthesis rates by all leaves. Increasing live foliage area will reduce available light, since the absorptivity of visible wavelengths by live leaves is high. Mean photosyn- thetic rates should decrease, although stand photosynthesis should in- crease up to a maximal foliage area index. Beyond this point self-shading should result in a decrease in carbon dioxide uptake. Mean leaf photo- synthetic rates decrease as the live foliage area index exceeds 0.74 (Figure 4-18). With a live foliage index of 1.0, a common upper value, mean leaf rates are decreased to 14% of the open canopy, a decrease which results principally from the absorption of radiation by the live leaves, resulting in high light extinction in the canopy. Thus, with a foliage area index of 1.0 or higher, relatively few leaves in the 10 to 15 cm stratum are saturated and then for only 1 hour around solar noon. Although mean photosynthetic rates on a foliage area basis continue to decrease, com- munity uptake increases up to a foliage area index of from 3 to 6. With a foliage area index of 6 as much as 34 g CO2 m"^ day"' is assimilated. In terms of carbon dioxide exchange, a foliage area index as high as 8, which has been measured (Dennis et al. 1978), can be supported by the graminoid vegetation. With such a high foliage area index the lower leaves are in a negative carbon balance. Canopy architecture becomes in- creasingly important in affecting photosynthesis at high foliage areas. In the standard canopy with a dead area index of 1.24, photosynthesis in- creases at all times of the day as leaf inclination increases. One of the characteristic features of graminoid canopies, especially in the absence of lemmings, is the accumulation of standing dead mate- Photosynthesis 133 400 o c a itu — 1 1 — 1 — 1 — 1 — 1 — 1— 1 — 1 — 160 1 \ 80 ^x::^b^ _ 1 1 [ 1 1 1 1 I 1 2 4 6 8 Foliage Area Index 10 FIGURE 4-18. A simulation il- lustrating the effect of increasing live (upper graph) and dead (lower graph) foliage area index on photo- synthetic rates per unit ground area (a) and per unit leaf area (b). Other input is for the standard day. In the upper graph, ground area rates are the sum of all vascular plants. In the lower graph ground rates are for Dupontia only with a foliage area index of 0. 125 in the presence of a total live foliage area index = 0. 744 plus a dead foliage area index given. (After Miller et al. 1976.) rial (Tieszen 1972b). As an intercepting and emitting component in the canopy, the effect of the accumulation of standing dead material on photosynthesis and soil thaw could be quite significant. But because of the decreased absorptivity of the dead material its effects on photosyn- thesis are less than that of live material (Figure 4-18). In the simulations a dead fohage area index of 5.0, 2 to 3 times that usually measured in the field, results in a 64% reduction of photosynthesis. Accumulation of dead material has a depressing effect on photosynthesis, which is mini- mized as long as winter snows and some decomposition reposition this potentially intercepting material at the bottom of the canopy. At the bot- tom of the canopy, however, the effects of dead material will be greater in Dupontia than most other species, since Dupontia has more leaf area near the base of the canopy. Canopy relationships also have profound effects on mosses. A vas- cular plant canopy under moist conditions reduces the carbon dioxide uptake by mosses from the levels simulated in the open. The canopy reduces radiation and temperature at the moss surface, which would tend to decrease photosynthesis, especially during periods of low radiation. The canopy also reduces turbulence and results in a lowered vapor densi- ty gradient between the moss and the air. Therefore the reduction in car- 134 L. L. Tieszen et al. high c o o 3 O o 3 C C < low With Canopy low high FIGURE 4-19. A diagram- matic summary of the effect on moss production of the vascular canopy along a moisture gradient. (After Stoner et al. 1978b. f Available Water bon dioxide uptake due to the lowered irradiance and temperature may be more than compensated for by the increased photosynthesis because of the improved tissue moisture status. The result of an interplay between the environment in the vascular canopy and the moss-water relation is a changing relationship of moss production and canopy cover along a moisture gradient (Figure 4-19). The simulations indicate that at high levels of available moisture, there is no water limitation, and mosses are most productive in the absence of a vascular plant canopy under full ambient radiation. Production is limited by light, and plant resistances to gas exchange are important. At moderate moisture availability, a vascular plant canopy increases moss production by decreasing evaporation by more than the amount lost through interception. At low levels of available moisture, mosses are most productive in the absence of a canopy. In this situation, a canopy intercepts much of the precipitation available and mosses are seldom hy- drated. Under these conditions, brief periods of light precipitation are not effective in penetrating the canopy to hydrate the moss tissue (Stoner et al. 1978b). These results should be viewed as speculative because of uncertainties in the interception model and in the water uptake relation- ships in mosses. The water relationships of mosses are poorly understood compared with the level of understanding of moss photosynthesis. Grazing Grazing is an important biotic interaction that affects the photosyn- thetic rates and patterns described above. In the absence of a peak in the lemming population, the grazing pattern in the coastal tundra at Barrow results in the removal of some vegetation, usually as tiller units or shoots, near the moss surface. Thus, photosynthetic tissues- at various stages of photosynthetic competency are removed and consequently seasonal pro- duction is reduced. Heavy grazing, especially in late winter or early spring, removes the canopy, including standing dead and live material. Thus, the microenvironment surrounding the photosynthetic leaves Photosynthesis 135 FIGURE 4-20. Simulated responses of a Dupontia fisheri tiller to grazing. Leaf photosynthetic rates were estimated for 1971 environ- mental data and 1972 leaf growth data. Grazing was simulated by the complete removal of photosynthetic tissue. The numbers repre- sent the percent change in seasonal CO2 uptake result- ing from four grazing events. is altered dramatically. Leaves near the moss surface are exposed to more intense radiation and have lower temperatures, since convectional losses are greater (Figure 4-20). Early season grazing by lemmings has little effect on photosynthesis in Dupontia because photosynthetically competent tissues are not avail- able to be harvested and because mean photosynthetic rates of new tis- sues increase in a more open canopy. Midseason grazing, however, is very detrimental to seasonal carbon dioxide uptake. In plants grazed in midseason, carbon dioxide uptake was 42% less than in ungrazed plants. This results mainly because photosynthesis is limited by the available leaf area, and grazing at midseason removes photosynthetic tissue at a time when its contribution is greatest. In addition to the major effect on eco- system carbon balance, grazing reduces storage reserves (Tieszen and Ar- cher 1979), which will affect plant performance for one or more growing seasons. We have not assessed the occasional severe grazing pressure which can result in rhizome and stem base destruction. This may have dramatic effects on population structure. Seasonal Course of Carbon Dioxide Uptake Primary production varies both spatially and from season to season. Since the rate of net photosynthesis is not strongly depressed because of the low temperatures, other factors must account for major portions of the variation. The primary factors limiting primary production or can- opy development appear to be: 1) the length of the growing season (see Figures 3-1, 3-2), which is dictated by the duration of snow cover and related to topography and seasonal radiation patterns, and 2) the alloca- 136 L. L. Tieszen et al. 240 « 50 60 70 80 90 100 no >. ^ Year Leng 00 - - o § " ' S>S 4 a> CO - - 3 1 S S 0 ^if 5 I 1 , 1 1 1 I, , 1 0 60 70 80 90 00 no FIGURE 4-21. The simu- lated effect of season length on annual net photosynthesis for Dupontia fisheri in the canopy of the Carex-Onco- phorus meadow. The fre- quencies of yearly thaw sea- son lengths of periods with above 0°C mean daily tem- peratures for the years 1922 to 1973 are given also. Other input is for the standard season. (After Miller et al. 1976.) Season Length, days tion of carbohydrates for leaf production which, in the coastal tundra at Barrow, is controlled to a large extent by the availabihty of soil nutrients, principally nitrogen and phosphorus. The spatial pattern of above- ground primary production reflects these topographic and nutrient rela- tionships; and the magnitude of production depends upon climatic fac- tors. Simulating the seasonal courses of carbon dioxide uptake for Dupontia fisheri can be used to assess the response of the graminoids to some of these variables. Season length was simulated by initiating the season earlier or later than the 15 June date used for 1971. Environmental factors and rate of canopy development were assumed to be the same as in the 1971 simula- tion. Under these conditions, there is a marked effect of season length on total carbon uptake (Figure 4-21), primarily because plants can more fully utilize the high levels of radiant energy at the time of the solstice. During the simulated long season, Dupontia attained a foliage area index of 0.22 on 4 August, nearly 30% greater than the standard season. Con- versely the foliage area was reduced by 30'yo during the short season. Un- fortunately, long-term field observations covering this range of season lengths are not available for comparison with productivity estimates. Production and carbon dioxide uptake are increased in the short season Photosynthesis 137 if the canopy delays senescence and continues to develop until a standard canopy is attained. The simulations and an analysis of the seasonal course of CO2 uptake (Figures 3-1, 3-2) suggest that production is very sensitive to the rate of allocation to leaves. We have estimated (Tieszen 1978b) that a 25% increase in leaf production results in a 45% increase in CO2 uptake. Thus, it appear that both the date of initiation and the pat- tern of canopy development are among the most important factors con- trolling CO2 uptake and primary production. Variations in the amount of radiant energy received during a season of constant length affect seasonal uptake and production. An increase in incoming solar radiation equivalent to one standard deviation results in a 13% increase in uptake. The overall sensitivity of the carbon dioxide uptake system to tem- perature variations is much less significant. Decreasing temperatures by one standard deviation results in a mean temperature of -l-3.0°C, or 2.3 °C lower than the long-term mean, but reduces carbon dioxide uptake by only 1 .4%. Early in the season, low temperatures inhibit carbon diox- ide uptake on a leaf basis. However, the small foliage area at this time makes its effect on the total season budget nearly negligible. Increasing temperatures by one standard deviation increases carbon dioxide uptake by only 0.4% (Tieszen 1978b). In order to determine the effect of several factors interacting simul- taneously, two seasons were simulated, one warmer and brighter than the mean and one colder and with less irradiance than the mean. The warmer season with higher irradiance resulted in an 1 1 % increase in carbon diox- ide uptake whereas the colder season with less irradiance reduced carbon dioxide uptake by 17%. If these patterns are associated with changes in rates of allocation, the effect is great. When the warmer, brighter season was combined with a 20% increase in live foliage area index, net photo- synthesis increased 53%. When the colder, darker season was combined with a similar decrease in the foliage area, net photosynthesis decreased by 41%. The carbon dioxide uptake system, however, appears relatively insensitive to temperature, is quite dependent on radiant energy, and is very dependent on allocation to photosynthetic tissue. The sensitivity of allocation to various environmental factors must now be understood because photosynthesis, per se, functions quite well at the prevailing am- bient temperatures. SUMMARY The photosynthetic rates of plants of the coastal tundra at Barrow show consistent patterns among growth forms that are comparable to similar plant types in more temperate zones. Maximal rates of carbon 138 L. L. Tieszen et al. dioxide uptake are greatest in leaves of short duration, for example grasses and forbs, and are lowest in the evergreen dwarf shrubs, mosses and lichens. The dominant graminoids attain rates around 17 to 21 mg CO2 dm"^ hr"' and the mosses 1 to 5 mg CO2 gdw' hr^'. Leaves rapidly develop photosynthetic competency following snow- melt, and maximal rates are highly correlated with carboxylation activ- ity. The continuous irradiance and a relatively open canopy result in high daily rates of carbon dioxide uptake. Net rates for whole plants are high- est in early July, 200 to 400 mg CO2 dm"^ hr"', and decrease progressively until early September as senescence progresses, self-shading increases, and the irradiance decreases following the summer solstice. Efficiencies of energy conversion are above 1 % for the graminoids and are as high as 2.3% for Salix pulchra. The net seasonal incorporation is 602 g m"^ a field value corroborated by the aerodynamic assessment of carbon diox- ide exchange and the simulation model. Approximately two-thirds of the seasonal incorporation occurs after the canopy has developed and has re- plenished belowground reserves. Plants are well adapted to prevailing tundra environments. The vas- cular plants and mosses have similar, low light compensation require- ments (5.6 to 15.8 J m'^ s"', 400 to 700 nm), but differ with respect to light saturation. Grasses saturate around 279 J m"^ s'' and mosses satu- rate around 98 J m"^ s"'. On a daily basis vascular leaves are rarely light- saturated, but mosses may be inhibited by high irradiances, especially in open canopies. Temperature optima for leaf carbon dioxide uptake are commonly around 15 °C or well above mean ambient temperatures. The high uptake efficiencies on a daily and seasonal basis suggest that this optimum allows plants to function effectively under climatic conditions of the coastal tundra at Barrow. Simulations confirm that a temperature opti- mum of 15 °C allows vascular plants to take up carbon dioxide efficiently across the range of temperatures experienced. This occurs in part because the leaf area is concentrated at the base of the canopy where leaf temper- atures are higher and because the leaves often function on the light- dependent portion of the light response curve. Seasonal temperature ac- climation is not apparent. Mosses, however, show a greater sensitivity to temperature changes and a greater vulnerability to water loss. Thus, they show more frequent water stress than the vascular plants, which are rarely water-stressed even though leaf resistances are low. This results from a low evaporative demand and a high soil water potential. Annual carbon dioxide uptake and net primary production are mainly limited by the availability of photosynthetic leaf area. In a typical season, photosynthesis on a land area basis is strongly limited because the canopy is not well developed until late July when solar irradiance is already decreasing. Season length, or more precisely, the date of snow- 1 Photosynthesis 139 melt, is an important factor dictating the extent of canopy development and therefore the uptake of carbon dioxide. Spatially, the uptake of carbon dioxide is under a similar control. The productivities of the vegetation types are largely a function of the plant densities or foliage area indices, which are determined by the mois- ture and nutrient (mainly phosphorus) gradients. Wetter and more fertile areas are more productive because the plants composing the vegetation types in these areas allocate more carbon for photosynthetic tissues. The photosynthetic rates on a leaf basis of a given growth form remain com- parable across a wide range of microtopographic units. Control of Tundra Plant Allocation Patterns and Growth F. S. Chapin III, L. L. Tieszen, M. C. Lewis, P. C. Miller, and B. H. McCown INTRODUCTION There is no significant difference in photosynthetic potential among populations in the three major graminoid species in the coastal tundra at Barrow (Chapter 4), and all populations receive the same low solar irra- diance. Nevertheless, there is as much as a ten-fold difference in standing crop among these same populations. Clearly, growth and primary pro- duction are limited by more than plant photosynthetic potential or energy available for photosynthesis. This chapter seeks to explain how tundra plants control allocation patterns and growth to successfully ex- ploit the cold-dominated, nutrient-limited environment of Barrow. Carbon and minerals may be allocated to: 1) leaf and stem growth, which exerts a direct positive feedback on canopy photosynthesis by aug- menting the quantity of photosynthetic tissues; 2) root production, which increases the capacity of the plant to absorb water and nutrients; 3) rhizome production, which leads to vegetative reproduction and addi- tional photosynthetic tissue and provides for carbon and nutrient stor- age; and 4) inflorescences, which can lead to the dispersal and establish- ment of genetically distinct individuals. The balance and timing of these alternative allocation pathways determine the pattern of plant growth and reproduction and the relative supply of carbon, water and nutrients that the plant acquires. We place special emphasis on Dupontia fisheri, which typifies the non-caespitose or single graminoid growth habit that constitutes the major part of the coastal tundra vegetation at Barrow (Chapter 3, Tieszen 1972b). 140 Control of Tundra Plant Allocation Patterns and Growth 141 GROWTH PATTERNS OF TUNDRA GRAMINOIDS Growth Form of Dupontia fisheri Dupontia, like most grasses, consists of a prostrate, subterranean branched stem (rhizome), aerial shoots, and roots produced from below- ground nodes (Figure 5-1). Each new tiller is initiated from an axillary bud of a leafing node. As the bud develops and the rhizome elongates be- low ground, the tiller is termed a "VO," a tiller containing only rhizome phytomers. At Barrow a VO tiller seldom reaches the stage of leaf exser- tion in its first season of growth, whereas mid-latitude grasses generally exsert leaves in the same season that rhizome growth begins (e.g. Koller and Kigel 1972). During the second growing season, when the shoot ap- pears above ground, the tiller is designated "VI," a tiller in its first sea- son of shoot production. The tiller then produces three or four new leaves each year. Tillers in their second, third and fourth years of growth above ground are designated "V2," "V3," and "V4," respectively. A second VO and sometimes a third or fourth may be produced by any tiller in its first year of leaf production. These subsequently pro- duced tillers are designated as "primes" (for example VO ', VI ') to indi- cate their age class and relationship to other members of that class. Sub- FIGURE 5-1. Typical tiller system o/ Dupontia fisheri showing age class- es of component tillers. The letter V with associated number indicates the age class of the tiller (see text). (From Allessio and Tieszen 1975a.) 142 F. S. Chapin III et al. sequent development of "primes" may be slower, and they may take one year longer to complete their life cycles. Production of prime tillers is the means by which a tiller system branches, although prime tillers often do not produce daughter tillers and become "dead ends" or side branches of the main rhizome axis (Figure 5-1). After production of approximately 18 nodes, the apical meristem typically differentiates into an inflorescence. The last four nodes and the inflorescence normally develop in the third (V3) or fourth (V4) season of aboveground growth, although induction and initiation occur at the end of the preceding season (Mattheis et al. 1976). The last growing unit de- velops into the flag leaf and the elongated lower internode of the culm. After flowering, the shoot dies. The tiller rhizome and roots continue to live for several years and may occasionally produce VO tillers (AUessio and Tieszen 1975a, Shaver and Billings 1975). The tiller system comprises all the vegetative offshoots of the orig- inal seedling or tiller, and in the tundra at Barrow consists of more than 20 to 30 tillers. Living tillers are often interconnected by dead rhizomes. Seasonal Growth Patterns Shoot Growth The general pattern of leaf turnover in Dupontia is similar to that of temperate grasses (Evans et al. 1964, Langer 1966, Milthorpe and Moorby 1974) and does not represent any unique adaptation to arctic conditions. Dupontia produces leaves continuously through the growing season, although more rapidly early in the season (Figure 5-2) (Mattheis et al. 1976). Leaves that are not fully exserted by season's end lie quies- cent until spring and then resume growth. Laboratory studies suggest that leaf growth is largely supported by reallocation of carbohydrate and nutrients from simultaneously senescing old leaves (McCown 1978), so that leaf production may represent a large sink for carbohydrates and nutrients only in the spring of the first year of aboveground growth. This hypothesis is supported by '^C labeling studies (Allessio and Tieszen 1975a, 1978) and computer simulations (Miller et al. 1978c) but lacks documentation of nutrient reallocation patterns. Flowering tillers differ from vegetative tillers in their pattern of leaf turnover in that all leaves of flowering tillers senesce relatively early in the season. The accelerated senescence presumably represents a develop- mentally programmed redistribution of materials to reproductive struc- tures, as in temperate graminoids (Williams 1955). Thus the pattern of shoot growth of Dupontia may maximize reutilization of nutrients from senescing leaves. Control of Tundra Plant Allocation Patterns and Growth 143 VO Moss Surface . ■■ ,J|| I Ml ^.^.■)M.y,,„^.„y.yj^^l_;^;.,i^ "• ■' '"' ^ ^ 1 T 2 Jul 12 Jul 22 Jul I Aug II Aug 16 Aug 20 Sep Aug II Aug 16 Aug FIGURE 5-2. Seasonal growth patterns of Dupontia tillers of different age classes. Leaves are represented by lines inclined at 62 °, the mean leaf inclination for Dupontia; the solid portion of the line represents green leaf length and the dashed portion represents senesced length. The most recently exserted leaf is inclined 90° until fully exserted. The inflores- cence is represented with an arrow. [Drawn from data collected for shoots (Mattheis et al. 1976), roots (Allessio and Tieszen 1975a, Shaver and Billings 1975), and VO tillers (Lawrence et al. 1978) in 1973.] 144 F. S. Chapin III et al. 0.12 0.10 0.08 o "o. 0.06 o cr. o 0.04 o «, 0.02 -0.02 - -0.04 n — I — I — I — r~ (o)Carex aquatilis (•)Dupontia fisheri 1 J L A M J J L T \ I I I I I r (o) Agropyron smithii (•) Bouteloua gracilis 0 N D J FIGURE 5-3. Mean relative growth rates of graminoids observed in situ in a) tundra (calculated from Tieszen 1972b), and b) temperate grass- land at Cottonwood, S.D. (Lewis et al. 1971). The most striking feature of plant growth at Barrow is the high pro- duction rate at low temperatures. Graminoids in the Biome research area have relative production rates (g g"' day"') comparable to those of some dominant grasses of a mid-latitude grassland, in spite of a 15 to 20 °C dif- ference in average air temperature during the growing season (Figure 5-3). The principal difference in their patterns of production is not the rate of growth but the short period during which growth occurs in the tundra. The capability of tundra plants to grow effectively at low tem- perature is further seen in controlled environment experiments where graminoids from Barrow exhibit maximum rates of leaf initiation, elongation, and hence growth at 15°C (Tieszen, unpubl.). This is com- parable to the optimum temperature for growth of some alpine grasses (Scott 1970), but some 10 to 15° cooler than the optimum growth tem- perature of temperate zone grasses (e.g. Evans et al. 1964, Warren Wil- son 1966a). However, the 15 °C temperature optimum for growth of tun- Control of Tundra Plant Allocation Patterns and Growth 145 dra graminoids is 5 to 10 °C higher than average summer shoot tempera- ture in the field. Transplant studies show that Barrow graminoids and other arctic species do grow faster in warmer climates (Warren Wilson 1966a, Chapin and Chapin, pers. comm.) than in their natural environ- ment, indicating that arctic plant growth is limited in part by temperature despite adaptations that permit rapid growth at low temperature. The similarity of relative production rates between tundra and mid- latitude grasses in the field suggests that the metabolic cost associated with this production (i.e. growth respiration) may also be similar, assum- ing comparable production efficiencies. This hypothesis is supported by laboratory studies showing that arctic plants have a higher respiratory rate than mid-latitude plants when measured at some standard tempera- ture, but that the respiration rates of various populations at their respec- tive habitat and growth temperatures may be comparable (Mooney and BiUings 1961, Billings et al. 1971). High rates of mitochondrial oxidation are the cause of high respiration rates measured in intact plants (Klikoff 1966). Because respiration is temperature-dependent, a high respiratory capacity would be required for arctic plants to maintain their observed growth rates at low ambient temperature. The high respiratory capacity of arctic plants is determined both genetically and environmentally, al- though genetic factors appear more important than acclimation in ex- plaining this temperature compensation (Klikoff 1966, Billings et al. 1971). Many authors (e.g. Bliss 1962a, Billings and Mooney 1968) have commented upon rapid spring shoot growth of tundra species. However, Warren Wilson (1966a) found lower growth rates in the high Arctic than in England. Further critical studies of relative growth rates of arctic plants are needed. Rhizome Growth Simulations suggest that the rapid early-season leaf growth of Dupontia is correlated with a corresponding decrease in the biomass of the rhizome and to a lesser degree of the stem base (Figure 5-4). This has been corroborated in measurements of temperate and upland tundra sedges (Bernard 1974, Chapin et al. 1980). Later in the season there is substantial allocation of biomass to belowground organs and probably a retrieval of materials from senescing leaves to the rhizome and sheath/stem base in preparation for the following season. This agrees with conclusions of the carbon dioxide budgets (Chapter 12) and '"C and '^P autoradiography (Allessio and Tieszen 1975a, Chapin and Bloom 1976). The main growth phase of new VO tillers is from mid-July onward in contrast to the mid-June onset of leaf production. The delay of below- 146 F. S. Chapin III et al. FIGURE 5-4. Simulated seasonal biomass pattern of different plant parts o/Dupontia fisheri tillers of differ- ent ages. (After Stoner et al. 1978d, Lewis, unpubl.) ground growth behind aboveground growth is common to temperate and other arctic graminoids and presumably is a consequence of programmed reallocation of resources from one plant part to another (Evans et al. 1964, Auclair et al. 1976, Callaghan 1976, Chapin et al. 1980). The soil reaches maximum temperature later in the season than does air, and this may have further selected for asynchrony of above- and belowground growth. Simulations suggest that rhizome weight in flowering tillers con- tinues to decline through the entire season, coincidental with the growth of the culm and inflorescence. Although the overall seasonality of rhi- zome growth is probably genetically programmed, the absolute growth rate is subject to environmental modification. In contrast to the temper- ate graminoids investigated, Dupontia is capable of active rhizome growth at soil temperatures at least as low as 4°C (McCown 1978). I Control of Tundra Plant Allocation Patterns and Growth 147 Root Growth Primary root primordia begin accumulating new photosynthate in the spring while the soil is still frozen (Dadykin 1954, Allessio and Ties- zen 1975a, 1978), but growth and elongation do not begin until after soil thaw. In Dupontia elongation of primary roots is complete by late July, whereas other major graminoids of the tundra continue primary root elongation throughout the growing season (Shaver and Billings 1975). Dupontia initiates two to four roots per rhizome node. Autoradiography shows that these roots remain functional throughout the life of the tiller, frequently remaining alive even after the shoot dies (Allessio and Tieszen 1975a). A few shorter, slender roots are initiated in later seasons from the nodes of older leafing phytomers and elongate upward between the dead sheaths; they may be important in retrieving leached nutrients from stem flow. Primary roots of Dupontia do not elongate after their first season (Shaver and Billings 1975). Toward the end of the first season lateral roots are initiated with a larger surface-to-volume ratio (Shaver and Billings 1975). As with primary roots, these roots actively accumu- late new '"C photosynthate much earlier in the season than they com- mence visible elongation (Allessio and Tieszen 1975a). Root biomass in the tundra at Barrow shows more variation both among and within microtopographic units than between sample dates. Therefore the seasonality of root production is difficult to determine by the harvest method (Dennis and Johnson 1970, Dennis 1977, Dennis et al. 1978). '"C translocation studies (Allessio and Tieszen 1975a) and di- rect observations of root elongation (Shaver and Billings 1975) suggest that much of the root production for the wet meadow tundra occurs in July and August, after the early flush of leaf production. It appears that early in the season new photosynthate is allocated primarily to new shoot growth, and that the root growth that does occur at this time proceeds largely at the expense of rhizome carbohydrate reserves acquired in pre- vious seasons. In the community as a whole, approximately 25% of the root biomass (i.e. 100 g m~^) may turn over each year (Shaver and Bill- ings 1975). This percentage is considerably lower than that found in many temperate ecosystems such as the eastern deciduous forest (Harris et al. 1977), but is approximately the same as that reported for grassland communities (Dahlman and Kucera 1965). The seasonality of root loss through senescence is not known, but loss is assumed to occur largely during winter. Because of the large ratio of belowground to aboveground biomass, roots and rhizomes constitute a major carbon and nutrient in- put to the saprovore food chain and are probably a relatively more im- portant carbon-nutrient source than in temperate ecosystems (Chapter 12). 148 F. S. Chapin III et al. Roots of tundra plants, of necessity, grow at temperatures below 5 °C and can resume active elongation even after being temporarily fro- zen (Billings et al. 1977). Graminoids at Barrow have lower optimum temperatures for root initiation, elongation, and hence production than do comparable temperate species, when grown in the growth chamber (Chapin 1974a). In part due to low temperature sensitivity of root growth, root production in the field is not correlated with root tempera- ture but appears to be controlled largely by photoperiod (Shaver and Billings 1977). For example, elongation of primary roots in Dupontia oc- curs during the first half of the growing season when root temperatures are lowest. Secondary root production predominates in August (Shaver and Billings 1975, 1977). In Eriophorum angustifolium most of the root tips are close to the retreating permafrost table, so the zone of most rapid elongation occurs in the coldest soil. The primary roots of all graminoids at Barrow have large diameters and contain aerenchyma. These roots transport sufficient oxygen to pro- duce an aerobic rhizosphere (Barsdate and Prentki, unpubl.) in a soil that is frequently oxygen-deficient. Because tundra plants are effective in transporting oxygen to roots, it is unlikely that metabolism of primary roots is ever limited by an inadequate oxygen supply. The root diameter probably results from a balance between selection for large diameter for adequate oxygen transport and selection for large surface-to-volume ratio for effective nutrient absorption (Chapin 1978). Both Dupontia and Carex produce secondary roots with small diameter, which are found pri- marily in the surface aerobic soil horizons and are important in nutrient absorption because of their large surface-to-volume ratio. These thin roots probably meet most of their oxygen requirement with soil oxygen rather than by diffusion through the primary roots. The small diameter rooting strategy allows continued exploitation of the surface soil horizon with a minimal carbon investment (Chapin 1978). Graminoids at Barrow produce very few root hairs in the field, as is typical of aquatic and emergent plants (Sculthorpe 1967), Root hairs pre- sumably require an external oxygen supply for proliferation and main- tenance and for this reason are of minimal importance in Barrow tundra. ALLOCATION OF CARBON COMPOUNDS Seasonal Patterns Growth and production depend in part upon availability of photo- synthate in the form of total nonstructural carbohydrate (TNC), a pool that includes sugars and storage polymers. The TNC concentrations in leaves, stem bases and rhizomes of Dupontia are quite high, ranging Control of Tundra Plant Allocation Patterns and Growth 149 TABLE 5-1 Biomass and Carbohydrate Composition of V2 tillers of Dupontia fisheri Plant Biomass TNC Lipid Othertt part (mg part" ') (% dw) (mg part"'; ) {% dw, ') (mg part"') (mg part"') 19-30 June Blade 5* 26* 1.3 16 0.8 2.9 Stem base 9* 24* 2.2 6 0.5 6.3 Rhizome 33* 36 11.9 6 2.0 19.1 Root 40t 8t 3.2 10 4.0 32.8 Total (mg) 87 18.6 7.3 61.1 % ot total plant pool 1-15 Aug 21.4 8.4 70.2 Blade 35* 20 7.0 7 2.5 25.5 Stem base 10* 44* 4.4 4 0.4 5.2 Rhizome 20* 40* 8.0 5 1.0 11.0 Root 40t 2t 0.8 4 1.6 37.6 Total (mg) 105 20.2 5.5 79.3 % of total plant pool 19.2 5.2 75.5 23 Aug-20 Sep! Blade 5" 19* 1.0 5 0.3 3.7 Stem base 20* 40* 8.0 6 1.2 10.8 Rhizome 33* 42* 13.9 5 1.7 17.4 Root 40t 5t 2.0 6 2.4 35.6 Total (mg) 98 24.9 5.6 67.5 % of total plant pool 25.4 5.7 68.9 • Lewis and Tieszen (unpubl.). tEstimated as 30 to 40% of total biomass at peak season (from McCown 1978); seasonal changes assumed negligible with senescence equal to production. t McCown (1978). ** Estimated as equal to early season. tt By subtraction. between 20 and 40% of the total dry weight (Shaver and Billings 1976, McCown 1978, McKendrick et al. 1978) (Table 5-1, Figure 5-5). High TNC levels are typical of arctic and alpine species (Russell 1940, Mooney and Billings 1960, Warren Wilson 1966a, Fonda and Bliss 1966, but see Payton and Brasch 1978). This raises the question of whether tundra plants have a carbon/energy surplus or whether there is strong selection to maintain high TNC levels in spite of high growth demands for TNC. Russell (1940), Warren Wilson (1966a) and Haag (1974) concluded that low temperature somehow prevented carbohydrate use in growth and respiration, resulting in large sugar accumulations. We suggest that in- adequate nutrient supply is a major factor limiting use of carbohydrates in growth. Ahhough there is considerable interhabitat variation in pro- duction along a gradient of nutrient availability at the Biome research 150 F. S. Chapin III et al. FIGURE 5-5. Seasonal changes in concentrations of total nonstructural carbohydrates (a), polysac- charides (a), lipids (o) and sugars (•) in leaves of Dupontia fisheri and in stem bases and rhizomes of composite samples of all moist meadow graminoids. Vertical bars indicate standard error. (After McCown 1978.) area (Tieszen, unpubl.), populations show markedly similar TNC con- centrations (Table 5-2). Total nonstructural carbohydrate concentration tends to be higher in habitats with low tissue phosphorus and low pro- duction. These observations suggest that Dupontia 's growth is limited more strongly by nutrients than by carbohydrate availability. Computer Control of Tundra Plant Allocation Patterns and Growth 151 TABLE 5-2 Variability of Chemical Composition, Pro- duction, and Photosynthetic Potential in Leaves and Rhizomes of Dupontia fisheri Along a Moisture-Nutrient Gradient Coefficient of variation (%) Leaves Rhizomes Concentration Nitrogen 19.2 42.5 Phosphorus 37.0 64.9 Potassium 33.5 46.3 Calcium 33.1 51.0 Total nonstructural carbohydrate 16.4 3.6 Total community production 62.0 — Photosynthetic potential 17.0 — Source: Calculated from Tieszen (unpubl.). simulations suggest a similar conclusion, i.e. that an increased photosyn- thetic rate, and hence increased TNC availability, would increase pro- duction at Barrow but that this TNC limitation of growth is less marked than is limitation by nutrients (Miller et al. 1976, 1979). It appears that although arctic tundra plants grow in a low radiation environment, car- bohydrate availability does not unduly limit growth. In mature shoots o{ Dupontia storage polysaccharides constitute the bulk of the TNC pool (Figure 5-5) (McCown 1978), in contrast to the predominance of sugars observed in other arctic and alpine species (Mooney and Billings 1960, Fonda and Bliss 1966, Warren Wilson 1966a). These reserves may be important for periods of intensive grazing. Although soluble carbohydrates are generally less than 15% of dry weight in Barrow graminoids, these levels are nonetheless high in com- parison with temperate plants, and may contribute to the frost tolerance (Weiser 1970) that allows arctic plants to survive subzero temperatures at any time during the active growing season (Sdrenson 1941). In Dupontia the rhizome is the largest compartment for TNC storage. From 40 to 65% of the total TNC is located in the rhizome throughout the year (Table 5-1). Fructosans rather than starch are the main storage polysac- charide in arctic (McCown 1978) and cool temperate (White 1973) grasses. Tissue age strongly influences both the types and amount of carbo- hydrate present. In new roots and in the rhizomes and stem bases of young tillers, monosaccharides are the predominant carbohydrate, reaching concentrations as high as 40% dry weight (Shaver and Billings 152 F. S. Chapin III et al. 1976). In these tissues carbohydrates play an active metabolic role, pro- viding the energy necessary for rapid growth in a short growing season and protecting important tissues from freezing. It is mainly in the mature tillers that carbohydrate storage as polysaccharide becomes important. Hence, high TNC levels in plants at Barrow may reflect quite different processes, depending upon tissue age (Shaver and Billings 1976). There is a gradual accumulation of carbohydrate reserves in rhi- zomes and shoots as the growing season progresses (Figure 5-5). Overwintering green leaf sections enclosed within the sheath bases may be important reservoirs for respiratory energy and precursor molecules utilized during the rapid production of photosynthetic tissue in spring (McCown 1978) and are an important energy source for grazers during the winter and spring months. In early spring sugar levels are high, presumably the result of hydrolysis of the polysaccharides built up the previous autumn (Shaver and Billings 1976, McCown 1978). After the period of rapid leaf production in June, sugar levels drop and do not rise again until autumn. The seasonal pattern of lipids is unclear. In 1970 ear- ly season levels of leaf hpids were unusually high (17%), but rapidly dropped to a constant level of 697o (McCown 1978). In other years Hpid levels remained low throughout the growing season (Figure 5-5). High lipid levels in early spring could reflect energy storage (Bliss 1962b, Hadley and Bliss 1964) or could reflect small cell size and the abundant membrane lipid associated with meristematic tissue and high metabolic activity (Kedrowski and Chapin 1978). Early season lipids appear to have a lower melting point than those observed later and hence would func- tion effectively in membranes at lower temperatures (McCown 1978). The TNC concentration of Dupontia exhibits greater seasonal stabiHty than that of temperate plants (McKendrick et al. 1975, McCown 1978), a feature of arctic plants also noted by Warren Wilson (1966a). The seasonal constancy of TNC pool size reflects a seasonal stability of allocation patterns as demonstrated by '"C translocation studies in Dupontia. From snowmelt until time of maximum aboveground biomass (which was the duration of the study) new photosynthate is largely re- tained in the shoot where it is synthesized, presumably to support the continuing production of new leaves through the season (Figure 5-6) (Allessio and Tieszen 1975b). Except in VI tillers, which are in their first year of aboveground growth and which produce the majority of new roots and rhizomes, there is relatively little translocation of photosyn- thate to belowground structures during this first half of the season. Moreover, production of new leaves by a tiller is largely self-supported and is relatively independent of carbohydrate reserves stored below ground. Even reproductive tillers are largely self-sufficient and do not withdraw large quantities of photosynthate from their own or neighbor- ing rhizomes (Allessio and Tieszen 1975a). These observations indicate a Control of Tundra Plant Allocation Patterns and Growth 153 Source tiller VI (7%) Leaf Stem base Rhizome Root Inflorescence Leaf Stem base Rhizome Root VI (85%) V2(6%) V3 (2%) 0.4% 34% 0.2% t 1 . 2.8 26 0.3 '' 1.8 13 4.2 • ' 2.1 13 0.9 Source tiller V3 flowering (95%) 38% VI (1%) V2 (4%) ■ 0.4% trace 48 i ' 0.2% 0.6 4.8 i, 0.3 0.2 2.4 ■ ' 0.2 3.2 2.4 FIGURE 5-6. Distribution of '"C in tillers and plant parts of a Dupontia fisheri tiller system after the source tiller (VI or V3 flowering) was labeled. (After Allessio and Tieszen 1975a.) substantial degree of tiller independence and suggest that the large belowground carbohydrate reserves may serve primarily to support late- season root growth and maintenance and as a reserve for rapid leaf growth in the event of grazing rather than to support the normal course of leaf growth as had been previously assumed (Mooney and Billings 1960, Fonda and Bliss 1966). However, leaf growth in the spring may be 154 F. S. Chapin III et al. Dupontia fisheri 100 50 1 s , a ^ 2 2l a> o u « 100 50 i5 0 100 r 50 I L _L Andromeda polifolia I M_^ Rubus chamaemorus s^ L I L i May Jun Jul Aug Labeling Date Root DUl Stem Base Rtiizome Q Aboveground Ports Root and Rhizome FIGURE 5-7. Distribution of recovered '"C in different plant parts of Dupontia fish- eri in mesic meadow tundra at Barrow (Allessio and Ties- zen 1975a) and Andromeda polifolia and Rubus cham- aemorus in an arctic mire at Stordalen, Sweden (Johans- son 1974). Dupontia was har- vested two days after label- ing; the others were harvested three weeks after labeling. highly dependent upon belowground mineral reserves. Gas exchange and harvest measurements indicate that over the en- tire growing season only 34% of the annual carbon fixation is used in shoot growth and respiration. The remaining 66% is presumably translo- cated below ground, perhaps late in the season, to support growth and maintenance of roots and rhizomes. A downward translocation of this magnitude would be necessary to provide the energy source for the sub- stantial root and rhizome respiration measured by Peterson and Billings (1975). The retention of radiocarbon in shoots of labeled tillers and the sea- sonal stability of the allocation pattern in Dupontia differ strikingly from the allocation patterns of other growth forms (Figure 5-7) (Johans- son 1974). In Andromeda polifolia, an evergreen shrub growing in Swedish tundra, 75% of the fixed carbon was translocated below ground early in the season. As leaf growth began, a larger proportion of the '"C was retained in the shoot. Finally, at the end of the season, fixed carbon Control of Tundra Plant Allocation Patterns and Growth 155 was allocated above and below ground in approximately equal propor- tions (Figure 5-7). Rubus chamaemorus from Sweden had an allocation pattern consistent with its deciduous habitat. During the leaf production phase, virtually all the assimilated carbon was retained in the shoot, as in the actively growing Dupontia shoot. As the season progressed, an in- creasing proportion of the photosynthate was translocated below ground (Figure 5-7). Clearly, the allocation patterns of these three species are closely tied to their growth forms, phenological calendars, and locations of storage. Environmental Influence upon Allocation Pattern Environmental factors affect allocation pattern as well as the total quantity of production . High root-to-shoot ratios of plants observed in the field at the Biome research area (Dennis and Johnson 1970, Chapin 1974a, Dennis 1977, Dennis et al. 1978, Webber 1978) may indicate greater environmental limitation upon shoot than root growth (Dennis and Johnson 1970) or more likely reflects a genetically and environment- ally controlled allocation of biomass to nutrient absorptive tissue (Chapin 1974a), as observed in laboratory studies (Brouwer 1965, David- son 1969). Low root temperature may indirectly result in a high root-to- shoot ratio by decreasing rates of nutrient uptake, thus lowering the nu- trient status of the plant, i.e. low root temperature and low nutrient status may influence allocation through similar mechanisms (Patterson et al. 1972, McCown 1975). Laboratory studies indicate that the increase in root-to-shoot ratio resulting from an impoverished nutrient status serves to increase nutrient supply and decrease nutrient demand, thus compensating for the nutrient deficiency (e.g. Leonard 1962, Brouwer 1965). Direct field evidence for this comes from long-term fertilization studies at Barrow, where after 10 years of fertilization, the root-to-shoot ratio was reduced from 7:1 to 3:1 (Dennis 1977). McCown (1978) ob- served that the root-to-shoot ratio of Dupontia was less affected by root temperature than was that of the temperate grass Poa pratensis. A relatively inflexible root-to-shoot ratio is typical of slowly growing species (Grime 1977). When growth is strongly nutrient-limited, nonstructural carbohy- drates accumulate to high levels (Leonard 1962) as observed in Dupontia at the Biome research area (Shaver and Billings 1976, McCown 1978). TNC levels are reduced, and shoot growth is enhanced by fertilization, a further indication of the importance of nutrients in limiting production in the tundra (McKendrick et al. 1978). The number of daughter tillers (VO's) initiated is positively correlated 156 F. S. Chapin III et al. TABLE 5-3 Correlation of Production Parameters with Nutrient Composition of the Plant Part in Tillers Sampled along a Moisture-Nutrient Gradient Production parametei No. VO No. leaves Leaf (mature tiller- length tiller)-' Carex aquatilis Nitrogen -0.21 -0.29 — Phosphorus 0.49 0.52 — Dupontia fisher i Nitrogen 0.10 -0.51 0.80* Phosphorus -0.12 0.37 -0.69t Eriophorum angustifolium Nitrogen 0.47 -0.06 — Phosphorus 0.45 0.59 — Note: Numbers are correlation coefficients. *Significant at the 0.05 level of probability. tSignificant at the 0.01 level of probability. Source: Calculated from Tieszen (unpubl.). with nitrogen concentration and negatively correlated with phosphorus concentration (Table 5-3), whereas leaf length shows the reverse correla- tion. No clear trend is evident with leaf number. This suggests that the intercalary and rhizome meristems respond quite differently and perhaps are limited by different nutrient balances. Apparently, leaf intercalary meristems are stimulated by favorable phosphorus status and/or a low nitrogen-to-phosphorus ratio. A high phosphorus content might be an- ticipated in meristematic cells because of the high requirement for mem- brane phospholipid and phosphorylated sugars. The high leaf meristem- atic activity under such nutritional circumstances would lead to strong apical dominance and could partially explain the concomitant low rates of rhizome growth and vegetative reproduction. Translocation is inhibited by low temperature only temporarily, even in temperate plants (Swanson and Geiger 1967). Temperature does influence the rate of active loading and unloading of phloem cells, an ef- fect that is indistinguishable from the temperature effect upon source and sink strength (Crafts and Crisp 1971). In contrast to temperate plants, Dupontia is capable of translocating ""C to root and rhizome pri- mordia frozen in soil (Allessio and Tieszen 1975a), as discussed above. Similarly, low temperature inhibits translocation of sugars from leaves in C4 grasses growing at their upper elevational limit but not in alpine gram- to s: o bo s: a On S o I ■ PQ < nj D. 3 t« c« »j 3 U > ;/) ^ ,1- O ^ > 3 _U |.2 -I - I CTJ qj >^ S 2 O Z o U D. ^ E E g ao o. E s s O O o d o oo o d +1 d +1 SO so O C 1) so r- Os ? 13 o _ t s aa ju O ^ (U . E 2 E-§ D. C 3 O U E I o is c c 03 c 03 - w-i 03 r-- Os OC = I 03 fS so o B o _o OQ •o c c "5u o. S 'S 3 « Q. 00 = 1 ^ c — ■7; 1> 2 3 ;« -^ a .2i o c 3 H -■ c o .y c •a o es •;: ^ 3 00 ^ „ o c '^ (U oj 'i: c 3 — E 03 00 c _o 03 •- ^ < 2 o w 00 u a 157 158 F. S. Chapin III et al. inoids in Wyoming (Wallace and Harrison 1978). Thus, it appears that the effects of low temperature upon allocation in tundra plants are in- volved more with growth processes than with inhibition of carbohydrate and nutrient transport. Data in support of this hypothesis are largely in- ferential, and direct experimental evidence is needed. Roots are more strongly buffered against temperature variation from one growing season to another than is the plant canopy, and shoot growth increases more rapidly with increasing temperature than does nutrient uptake. For these two reasons, in a warm growing season pro- duction is greater and the demand for nutrients outstrips the slightly en- hanced supply. The situation is aggravated by the asynchrony of shoot production and nutrient uptake, the bulk of the nutrient uptake occur- ring after shoot production is largely complete (Chapin and Bloom 1976). The asynchrony is demonstrated by a comparison of the 1972 and 1973 growing seasons (Table 5-4) and suggests the following hypotheses: 1. In tundra communities, warm years will result in greater nutrient deficiency than will cold years. Nutrient deficiency will be particularly severe with respect to phosphorus because phosphorus absorption is more strongly affected by temperature than is absorption of other nutrients (Nielsen and Humphries 1966). The effect upon nutrient con- tent of yearly temperature differences will be particularly pronounced in tundra underlain by permafrost, where the temperature gradient is most pronounced from the canopy to the bottom of the rooting zone. 2. Changes as slight as 1 °C in air temperature above the plant can- opy are sufficient to profoundly alter the nutritional status of plants. Data from the Biome research area indicate the sensitivity of allocation pattern to small changes in environmental parameters. Similar conclu- sions were independently reached by Wielgolaski et al. (1975) in studies of the Norwegian alpine. In summary, studies of carbohydrate allocation suggest that plants of the coastal tundra at Barrow have compensated for the effects of low temperature by allocating a large proportion of their biomass to nutrient- absorptive and belowground storage tissues. The high TNC levels suggest that growth of ungrazed tundra plants is not unduly carbohydrate-limited. CARBON COST OF PLANT GROWTH Concepts The carbon fixed through photosynthesis can be converted to new biomass, but there are certain energy costs of producing and maintaining an increment of biomass that must be considered in order to understand patterns of carbohydrate allocation. The rate of change in live biomass Control of Tundra Plant Allocation Patterns and Growth 159 of a plant compartment is the balance between the rates of photosynthe- sis, respiration, and death and the net translocation flux to that compart- ment. Growth also involves the accumulation and incorporation of nutri- ents. The rate of change in nutrient content of a plant compartment is the balance among the rate of uptake from soil, the net translocation flux, and losses from the living system due to leaching, exudation, grazing and death. There is inadequate information currently available to allow growth to be analyzed in terms of nutrient costs. Respiration provides energy for a number of distinct metabolic pro- cesses and can be separated into at least three components: maintenance respiration, growth respiration and translocation respiration. Mainten- ance respiration provides the energy required to sustain the existing live biomass at its present level of metabolic activity. Maintenance respira- tion includes the energy requirement of basic metabolic processes and the replacement or turnover of structural and functional substances, particu- larly of proteins, and is assumed to be proportional to protein content and turnover (Penning de Vries 1975). The proportionality constants of maintenance respiration have been calculated (Miller 1979) for various plant parts according to the relationship derived for crop plants (McCree 1974, Penning de Vries 1975) and corrected for values measured in plants from the Biome research area (Chapin et al. 1975, Billings et al. 1978), The relatively high protein content of graminoids in the tundra at Barrow (Chapin et al. 1975) suggests a high maintenance respiration and is in agreement with the high shoot respiration rates measured (Tieszen 1975). Leaves have a higher proportionality constant of maintenance respira- tion (0.012) than do roots or rhizomes (0.006) because of higher protein content, which is consistent with the higher respiration rate measured in leaves than in roots. Growth respiration provides the energy required to synthesize vari- ous components of new tissues from glucose (Penning de Vries 1972b; Table 5-5). Although plants contain a variety of sugars that function as metabolic intermediates, these will be viewed as glucose equivalents for the sake of simpHcity. The investment in new tissue is equal to the quan- tity of glucose equivalents contained in the added biomass plus that glu- cose respired in the synthetic process, the growth respiration (Figure 5-8). The total growth respiration for production of new biomass is calculated from the increase in biomass and a constant c that apportions the cost of synthesis among the biochemical constituents of the new tissue (Table 5-5): c = 1.1 5(% lignin) + 0.17(% cellulose) + 2.03(07o lipids) + -l-0.59(<7o protein) + 0.17(% polysaccharide) + 0.09(% sucrose). s: I 5 5 I I c ^ o ■^ §■ k: to ^1 ^ l>3 5 ■ U CQ < ^_^ ■^^ *-> O o >» a: l-i « •a 15 « G O CJ ■« a> u^ E o N c o D. 1 Ic o E c o Oi QQ o c u (^- o # -J c/5 O E 3 13 o •)— C 5b t/3 u c 3 o e 1) t/) Vu c c *^ O i £ lU on ao ,^m^ O o c » 15 3 u« 0 c c o 'c5 lU >< 00 C/5 £ s k. o 3 T3 • to O Q. E 1 > c >> 1/5 60 lU c o "^ c u 3 "w -«-• o C 1> E 3 J= 4-' 1/3 o c o ffl u O VO O — ■^ O oo rn — O w^ r- r- o \0 vO O O O (J c/l !> 00 5 -^ c/5 OO °P c .E :5 3 (/5 "O o ts 3 C QO O (J t« Tr, o r- o r^ r- w-i ,^ — lo o — ■ — H — ■ — — d — «N u 3 "C c« J= u o rt lU u t/5 C/5 00 >. o "o O D. i! -O 3 •£ 2 .§ ^ .2f 00 c op T3 c Ov _3 . u = E E Ji 2 « ■a ?, c5 2 E E 1 S c "^ — C/5 '(U _0 D. -o ■a o 00 o. .E E c o c ^ cu 15 T3 <~ C O r- o — ^ ^ Si ^^ CO ^ O i> 00 T3 c 00 c 'c c a> O. c o «■?;•-- c ? o U S ^-' j= t/3 < U^ *^ OO X> 3 (N t/i c^ Z « .ii T3 nJ 00 J= c 'S « c t/) u Q^ a. c C/5 I/) T3 = (U 3 c/5 (/5 OQ < » -1- OO .2? r~- — ON .. c r- * 2: o U _: 5 (J * 160 Control of Tundra Plant Allocation Patterns and Growth 161 Glucose New Tissue or Compound B+Rq B FIGURE 5-8. Processes of growth and growth respiration. When new tis- sue is created, glucose is metabolized, some glucose appearing in the form of the new biochemical compounds (B), and some being lost (Re) as it supplies energy for the synthesis of these compounds. For example, to produce 1 g of lignin (BJ, 2. 15 g of glucose is used fIB + Kc) and Re involves 1.15 g of glucose. Other com- pounds use varying amounts of glucose per gram, as listed in Table 5-5. Re is calculated for each compound by subtracting the gram of product from the total gram of glucose used. The total Re for creating new tissue is found by summing the Re's for synthesizing each compound, weighting each by the percentage com- position. Tundra graminoids have a relatively low lignin content and a relatively high polysaccharide and sugar content and hence may be able to produce more biomass per unit carbon fixed than can temperate counterparts. More complete data on chemical composition of tundra graminoids are needed to verify this hypothesis. Translocation is a vital process requiring energy, particularly for the loading and unloading of phloem elements, and energy cost is propor- tional to the translocation rate; available estimates (Penning de Vries et al. 1974) suggest a value of 0.05 for the proportionality constant. Be- cause both loading and unloading costs are involved, 1 g of glucose will be required to translocate 10 g from source to sink. Because of the high belowground-to-aboveground biomass ratio of tundra graminoids, more of the maintenance and growth respiration occurs in nonphotosynthetic tissues than would be the case in their temperate counterparts. Hence, translocation respiration may be relatively more important in tundra than in temperate graminoids. The carbon cost involved in growth is partially regained during se- nescence when some of the biomass components are broken down and retranslocated to storage areas to support future growth. Remobilization is not complete, and different classes of compounds are remobilized to different extents (Table 5-5). Sucrose and the various storage polysaccharides require a minimal energy investment for synthesis and subsequent breakdown. In contrast, Hpid requires a major energy investment, half of which is lost during reconversion to glucose. Although complete conversion of lipid to glu- cose is unlikely in any system, it is clear that lipid is a costly compound to synthesize and would be utilized as an energy source primarily where space or weight is limited, as in a seed. Plants growing in permafrost soils have larger amounts of membrane lipid but smaller amounts of storage 162 F. S. Chapin III et al. lipids than do plants in warm soil (Kedrowski and Chapin 1978). The high lipid content reported for arctic and alpine plants (Bliss 1962b) is probably associated with membrane lipid for metabolism or for anti- herbivore defense, such as resins, and not with an energy storage func- tion, as previously assumed. Lignin and cellulose, two important structural compounds, also dif- fer strikingly in their carbon cost of synthesis. Neither is significantly broken down for retranslocation during senescence. Lignin should be an important component of a tissue only where strong support is essential or where selection has led to strategies to discourage herbivory. Neither selective force appears to have led to a high lignin content in tundra graminoids. The implications of biosynthetic cost for successful adaptation to the tundra environment are discussed in Chapter 6. It is of interest, especially when considering the relatively anaerobic soil conditions of the coastal tundra, that biosynthetic processes under anaerobic conditions are 18-fold less efficient because of the incomplete oxidations that can be performed. In addition, toxic end products must be excreted, requiring additional energy expenditures. Although the native graminoids appear to have efficient oxygen transport systems and thus circumvent such problems, introduced plants and some dicoty- ledons may be severely stressed by the saturated soils and anaerobic con- ditions characteristic of many arctic soils. Biomass Flux Analyses for Dupontia A detailed picture of the fluxes involved in growth and allocation of carbohydrates and nutrients through the season can be built from simu- lations of the seasonal courses of dry matter accumulation, percentage composition, and metabolic cost estimates in Dupontia (Figure 5-9), Estimates of photosynthesis and respiration derived from the flux analy- sis can be compared with direct measurements of net photosynthesis and respiration. Analyses of this type are essential to a thorough understand- ing of the way the plant uses available resources to grow and produce offspring in a particular environment. Biomass changes provide an indi- cation of biosynthetic activity but the costs of biosynthesis, transloca- tion, uptake, and maintenance must be known to truly understand the allocation pattern and overall physiological balance. The following des- cription is based upon simulations of translocation, respiration and growth. In a newly initiated tiller (VO), simulated rates of translocation and growth generally increase steadily through the season until about 20 August. Subsequently, the rate of growth falls as maintenance respiration Control of Tundra Plant Allocation Patterns and Growth 163 Vegetative Tiller Flowering Tiller Inflorescence o o o i ■o E -2- y?^>^ . -r^, - //"^ * liv^ / . ^> ^ ^N ' '-^ 1 J 10 20 Jul 10 20 Aug 10 20 Sep FIGURE 5-9. Simulated seasonal changes in rates of growth and main- tenance respiration, and in dry weight gain by gross (Pc) and net (Pn) photosynthesis, and translocation to rhizomes (Tr), leaves (Yl) and in- florescences (T,) for vegetative (V2) and flowering tillers o/ Dupontia during 1973. (Calculated from Lawrence et al. 1978.) increases in parallel with biomass. The translocation rate represents a net import from other members of the system, because there are no leaves and hence no photosynthesis in a VO. Simulations suggest that the rhizome of a vegetative tiller (Figure 5-9) has a significantly negative growth rate, i.e. decreases in weight, through July and recovers through August and early in September. The simulated seasonal course of growth in blade tissue is negatively cor- related with growth of the belowground organs, changing from positive values early in the season to negative ones in late August. Calculations predict accurately the observed seasonal changes in biomass. The simu- lated seasonal changes in maintenance respiration reflect changes in size 164 F. S. Chapin III et al. of the live compartments. Most of the photosynthate produced in the blade is retained within this compartment. Both the stem base and rhi- zome export carbohydrate in early July but become net importers as the season progresses. The calculated rate of net photosynthesis increases gradually through the season even though the blade biomass decreases from early August onward. Net photosynthesis rates calculated from the flux analyses are consistently lower than measured rates until the end of August but then persist at unexpectedly elevated levels until freeze-up. No convincing explanation can now be offered for these discrepancies. The simulated patterns of growth and allocation in flowering tillers (Figure 5-9) contrast strikingly with those in vegetative tillers. In flower- ing tillers the growth rate of the rhizome is strongly negative at the begin- ning of the season, and the rhizome continues to lose weight throughout the season, although it never assumes the role of a major exporter of photosynthate. The stem base grows rapidly and blades develop their ex- porting capacity early in the season. The phase of positive leaf growth is truncated compared with vegetative tillers and never attains the same maximum rate. The second, smaller peaic in leaf growth may be associ- ated with the production of the flag leaf. The inflorescence exhibits the highest growth rate, particularly through midseason, and represents the largest sink for photosynthate. In contrast to vegetative tillers, there is no late-season recovery of stem base and rhizome because the photosyn- thetic rate decreases and inflorescence maintenance respiration increases throughout the season. Implications of Biomass Flux Analysis Simulations suggest that although the rhizome rapidly decreases in weight early in the season, there is relatively little upward translocation of reserves at this time. Maintenance respiration accounts for the major change in rhizome weight. In general, these simulations suggest that the rapid early-season development of leaves is self-sustaining and that re- mobilization of reserve carbohydrates is of minor importance. This con- clusion is supported by evidence from radiocarbon and carbohydrate analyses and contrasts strongly with earlier conclusions concerning arctic and alpine forbs. Russell (1940), Mooney and Billings (1960) and Fonda and BUss (1966) concluded on the basis of correlations between growth and changes in carbohydrate levels that rhizome reserves were utilized to support rapid spring shoot growth. Further research is necessary to determine whether the difference in conclusions results from differences in methods and interpretation or from differences between graminoid and forb growth patterns and phenology. Simulations, seasonal measurements of TNC, and autoradiography Control of Tundra Plant Allocation Patterns and Growth 165 all suggest that the large pools of nonstructural carbohydrate contained in belowground structures of Dupontia are not redistributed to any great extent through the growing season and mainly provide substrate for maintenance respiration and, except in the reproductive tillers, are re- plenished during August and September in readiness for metabolism the following season. Simulations support '"C translocation studies (Allessio and Tieszen 1975b) showing that inflorescence development is self- sustaining and not dependent on reserve utilization, even though the rhi- zome rapidly loses weight early in the season. Later development of the inflorescence may well depend on photosynthesis produced by its own chlorophyllous tissue. NUTRIENT ABSORPTION Seasonal Patterns The high levels and relatively stable pool size of nonstructural car- bohydrate strongly suggest that nutrient availability, absorption rates and/or translocation rates are among the important factors limiting plant growth at Barrow. The nitrogen and phosphorus status of the soils is quite low (Chapter 7). More than 99% of the nitrogen and phosphorus present in the rooting zone is organically bound and available to plants only after being released by decomposition, a process that occurs slowly at low temperatures and low oxygen levels. Nitrogen is frequently cited as a key limiting nutrient at the Biome research area and in other tundra systems (Russell et al. 1940, Warren Wilson 1957, BHss 1966, McKend- rick et al. 1978). Exchangeable ammonium, the principal ionic form of nitrogen, remains seasonally constant in the rooting zone but increases through the growing season in lower soil horizons (Flint and Gersper 1974), suggesting that nutrient absorption by plants has a significant im- pact upon soil nutrient dynamics. In contrast to nitrogen, available soil phosphorus and potassium decrease in concentration through the grow- ing season (Bar^l and Barsdate 1978), suggesting that phosphorus and potassium will be absorbed by plants earlier in the season than will nitrogen. An important asynchrony is apparent in environmental favorability for aboveground and belowground plant processes. Light and air temp- erature regimes are most favorable for photosynthesis immediately fol- lowing snowmelt, whereas soil temperatures and thaw depth increase un- til late July or early August. Seasonal changes in depth of thaw, soil temperature, and the quantity and activity of root biomass influence the seasonal pattern of nutrient absorption. However, simulations suggest that these parameters are less important than seasonal changes in soluble 166 F. S. Chapin III et al. o q: o ■o o tn < « o a. M o E 8.0 o in 5- e.oh Q. -O in o o 0. 4.0- 9 E •I > 3 E 3 o 2.0- d r^-^"^"^ / 1 1 15 15 15 15 Jun Jul Aug Sep FIGURE 5-10. Predicted rate of phosphate absorption per gram of Dupontia fisheri root during the 1973 growing season, a) Measured phos- phate absorption capacity of roots corrected for percentage of root bio- mass present in thawed soil (calculated from Dennis 1977). b) Absorp- tion rate in a corrected for soil temperature (MacLean 1973). c) Absorp- tion rate in b corrected for soil solution phosphate concentration (Barel and Barsdate 1978). d) Predicted cumulative phosphate absorption per gram of root during the growing season. Vertical arrow indicates date when net downward translocation of phosphorus from shoots to below- ground storage organs begins. (After Chapin and Bloom 1976.) Control of Tundra Plant Allocation Patterns and Growth 167 soil phosphorus in determining the seasonal pattern of phosphorus up- take (Chapin and Bloom 1976). The seasonal pattern of phosphorus re- lease from decomposers is probably the single most important factor governing phosphorus availability and therefore phosphate uptake by vascular plants (Chapin et al. 1978). Absorption of phosphate continues actively until late September when the soil begins freezing from the sur- face downward (Figure 5-10). More than 40*^0 of the total phosphorus absorbed by a given root biomass is acquired after 25 July, the date when shoots begin a net downward translocation of phosphorus for below- ground winter storage. Clearly, aboveground phenological patterns are an inaccurate gauge for determining periods of plant activity. Because total root biomass increases through the growing season, end-of-season nutrient absorption is probably even more important than the above dis- cussion would indicate. Physiological Basis Tundra graminoids differ from their temperate counterparts in hav- ing higher phosphate absorption rates under standard measurement con- ditions (Figure 5-11). Furthermore, tundra plants maintain substantial rates of phosphate absorption at temperatures that would inhibit active uptake by most temperate plants (Sutton 1969, Chapin 1974a, Carey and Berry 1978). For example, Dupontia grown in the field still maintains 35% of its 20 °C phosphate absorption rate at 1 °C (Chapin and Bloom 1976), which suggests that tundra plants actively absorb phosphate from cold soils and do not depend upon daily or seasonal warming of the soil to fulfill their phosphate requirements. Phosphate absorption by tundra plants is relatively insensitive to temperature changes and has an opti- mum temperature of at least 40 °C. It would appear that the phosphate absorption process in graminoids has adapted to low temperature by a decrease in temperature sensitivity below optimum temperature, by an increased affinity of roots for phosphate at low temperatures (Chapin 1977), and by an increase in uptake rate at all temperatures, but not by any change in temperature optimum. Similar conclusions were reached for the photosynthetic process (Chapter 4). The ability to acclimate in compensation for temperature changes is not well developed in plants at Barrow, as might be anticipated in a thermally stable environment such as the tundra soil (Chapin 1974b). The overall effects of temperature upon rate of phosphate absorption by Dupontia are such that the rate at the bottom of the soil profile (0.2 °C) is approximately 75% of the rate at the top of the profile (5.0 °C) (calculated from Chapin and Bloom 1976). A plant's capacity to absorb nutrients depends upon its nutrient status and allocation pattern. Plants with low concentration of an essen- 168 F. S. Chapin III et al. o Q. ■ O O c o J3 < o c Q. 08 • 2 fresh wt O " •! 5»\ .^0 4 — \^ S^6 If) 1 r* 1 1972 o o 1 1 1 1 1 - 1 o 1 1 1 1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Phosphorus Concentration in Blades, % 0.9 FIGURE 5-14. Tiller weight of aboveground material o/Dupontia fish- eri in relation to the phosphorus concentration of the leaf blades (after Ulrich and Gersper 1978) and the range of concentration measured in the field in 1972 and 1973 (Chapin et al. 1975, Tieszen, unpubl. data). In the laboratory, experimental plants (%) were grown in solution culture with different phosphorus concentrations and the total leaf phosphate was estimated from measurements of acid-soluble phosphate. Field measurements of biomass and nutrient concentration (o) were made in the moisture-nutrient gradient. always higher than the critical level necessary for maximum growth in the laboratory (Figure 5-14). Moreover, weights of field plants were consis- tently below those of laboratory plants when either nitrogen or phos- phorus was tested in the laboratory as the only factor limiting growth (Ulrich and Gersper 1978). Laboratory studies on plant critical levels suggest two hypotheses: 1) nitrogen and phosphorus never act as the sole limiting factor in the field but are both among a complex of limiting fac- tors, and 2) tundra graminoids never produce nutrient-deficient tissues, but rather limit growth rate. Graminoids at Barrow produce new tissue only if an adequate quantity of nitrogen and phosphorus is available for maximal development and presumably for optimal function. Environ- mental factors such as nutrients restrict growth rather than compromise the effectiveness of new tissues that are produced. Agricultural crops and species from fertile habitats differ substantially in this regard, respond- ing to nutrient stress with deficiency symptoms and reduced respiration and photosynthesis (Chapin 1980a). Control of Tundra Plant Allocation Patterns and Growth 173 In spite of a 15-fold variation in soil solution phosphate concentra- tion and a 6-fold variation in leaf phosphorus between microtopographic units, leaf phosphorus concentration shows no correlation with photo- synthetic rate (Chapter 4), suggesting that the photosynthetic apparatus of a graminoid is relatively insensitive to changes in phosphate concen- tration. In contrast, shoot production is positively correlated with 1) availabihty of soil phosphorus, 2) capacity of the plant to absorb avail- able phosphorus (Chapin 1978), and consequently 3) concentration of phosphorus in leaves (Table 5-3). The correlations suggest an intimate as- sociation between the phosphorus nutrition of the plant and growth un- der natural conditions in the field. However, the nature of the relation- ship between phosphorus nutrition and growth requires further study. The maximum nitrogen and phosphorus contents of graminoids at Barrow are as high as or higher than those of native temperate zone graminoids (Chapin et al. 1975). This might be anticipated because of a low percentage of structural material. Nitrogen requirements may be high in tundra plants because of high enzyme complements, as discussed previously. A high phosphorus complement may result from 1) a high in- cidence of polyploidy (Johnson and Packer 1965) and consequently high DNA content, and 2) high concentrations of membrane phosphoHpid to support metabolism and convey cold tolerance (de la Roche et al. 1972, Thomson and Zalik 1973, Kedrowski and Chapin 1978). Dupontia prob- ably has a higher proportion of its phosphorus tied up in structural mate- rial, DNA and phospholipid than do temperate plants (Figure 5-15). The total phosphorus complement of leaves varies considerably through the season and between habitats of different phosphorus status. Plants with low phosphorus content may have as much as 85 Vo of their phosphorus complement structurally bound and have essentially the same phos- phorus composition as standing dead material (Figure 5-15). Along a nutrient gradient nitrogen concentration of mature leaves is not correlated with photosynthetic potential. The percentage change in carboxylation activity from early season to mid-season and then from mid-season to end-of-season is greater than the corresponding change in nitrogen content, indicating that early and late in the season a higher pro- portion of shoot nitrogen is bound as structural protein and as nonpho- tosynthetic enzymes than at mid-season (Figure 5-16). This may partially explain the low photosynthetic rates early in the growing season (Chapter 4). The parallel decrease in carboxylation activity and total nitrogen con- tent in the latter half of the growing season suggests that ribulose diphosphate carboxylase is broken down more rapidly than it is synthe- sized after mid- to late July and that the nitrogenous breakdown pro- ducts are translocated out of the shoot at this time. The decrease in total carboxylation activity and shoot nitrogen after mid- to late July indicates a strong selection for early senescence and downward translocation of 174 F. S. Chapin III et al. 0.6 Owl Mound o c a> u c o o o ■g. o 0.0 Temperate Plant Dupontia ftsheri FIGURE 5-15. Estimated compart- mentalization of phosphorus into various classes of compounds in shoots of a temperate plant (Bieleski 1973) and Dupontia fisheri from three microsites. Ester-P includes inorganic-P and was determined by a weak acid extraction (Ulrich and Gersper 1978). Lipid was measured (McCown 1978) and assumed to con- tain the same proportion of phos- pholipid as cold-acclimated winter wheat seedlings (de la Roche et al. 1972). Phosphorus in DNA is as- sumed to be constant in Dupontia and 40% greater than that observed in temperate grasses (Bieleski 1973). The phosphate contained in RNA was determined by subtraction. 0.036 FIGURE 5-16. Seasonal course of ribulose diphosphate carboxylase activity per unit nitrogen in leaves o/ Dupon- tia fisheri, Carex aquatilis and Eriophorum angustifoli- um. (Calculated from Tieszen 1975, Chapin et al. 1975.) Control of Tundra Plant Allocation Patterns and Growth 175 nitrogen, despite the importance of length of photosynthetic season in hmiting total carbon gain (Chapter 4, Miller et al. 1976). At Barrow, the probability of a killing frost increases after early August and may select for downward nutrient translocation at this time to minimize the proba- bility of large nutrient losses. Moreover, root and rhizome growth is quite active in late July, and the belowground demand for nitrogen may be met in part by nitrogen translocated from leaves (Chapin et al. 1980). REPRODUCTIVE ALLOCATION AND POPULATION STRUCTURE Tiller Interaction Growth and allocation patterns have been discussed in terms of a single tiller unit. However, the tiller system is the genetic unit in Dupon- tia, and it responds to selection and interacts to maximize short-term car- bon and nutrient gain and long-term reproductive success for the entire tiller system. '"C and '^P translocation studies (Allessio and Tieszen 1975b, Chapin and Bloom 1976) indicate that although mature tillers act largely as independent physiological units, there is still considerable transloca- tion from old V2 and V3 tillers to young VO and VI tillers that have little or no photosynthetic tissue. Experiments in which individual tillers are severed from the tiller system corroborate these conclusions (Mattheis et al. 1976). Only tillers in their first season of leaf production were highly dependent upon other members of the tiller system for normal seasonal growth. Mature tillers initiated and supported more daughter tillers when their reserves were not tapped by the tiller system (Mattheis et al. 1976). Tiller interdependence may be particularly important during regrowth following grazing. Similar intertiller relationships have been observed in temperate grasses (Evans et al. 1964, Marshall and Sagar 1968). Neighboring tillers compete for available resources. Nutrients are probably more critical than carbohydrates in governing dry weight in- creases of tillers. Even young tillers have high available carbohydrate concentrations. The number and the length of new VO rhizomes are strongly correlated with rhizome nutrient concentration (Table 5-3). Clearly, the nutritional status of the plant governs the balance between growth of an individual shoot and vegetative propagation and hence has profound effects upon population structure. 176 F. S. Chapin III et al. Population Dynamics New shoot production and the interaction between members of a single tiller system are important primarily because of implications for population structure. Each tiller may draw upon the resources of older tillers to colonize, grow and reproduce, particularly during its first two years, thus minimizing carbon costs of growth and respiration and the risks of mortality during the early stages of growth. Unlike Carex big- elowii (Callaghan 1976) Dupontia tillers show minimal mortality until after flowering occurs in V3 or V4 tillers. Consequently, VO, VI, V2 and V3 tillers have similar frequencies, whereas V4 tillers are relatively un- common (Figure 5-17). Tiller interdependence averages out the growth and reproductive performance over an environment that is variable in both time and space. The vegetative reproduction characteristics of Du- pontia and other graminoids are not an ahernative to sexual reproduc- tion but rather a strategy of expanding the number of loci at which flowering may eventually occur (Lawrence et al. 1978). Vegetative repro- duction is important under situations where 1) a given tiller has a low probability of survival, e.g. due to grazing, 2) the probability of success- ful seed set in any given year is low, or 3) the probability of seedling establishment from the seed crop produced in a given year is low. All of these conditions characterize the coastal tundra at Barrow and may be important in selecting for extensive vegetative reproduction. The selected strategy, of allocating most resources to competition and to growth rather than to extensive short-term reproductive output, typifies tundra Sexual V4 Vegetative V4 V3 V2 V3' V2' VI' VO Frequency, % 10 0 10 20 30 Density, no. m"^ 100 0 100 200 300 FIGURE 5-17. Relative frequency and den- sity of Dupontia fisheri tillers in various age classes and sexual conditions in moist meadow tundra in August 1973. (Lewis and Tieszen, unpubl.) Control of Tundra Plant Allocation Patterns and Growth 177 species (Bliss 1971). Rough calculations based on carbon cost of inflores- cence production, percentage seed set, percentage seedling survival, etc., suggest that the carbon cost of producing a new tiller in coastal tundra by sexual reproduction is 10,000 times greater than the cost of tiller produc- tion through vegetative reproduction. The fact that sexual reproduction receives substantial carbohydrate allocation in spite of the low frequency of seedHng establishment points out the necessity of a long-term evolu- tionary framework within which to view growth and allocation pro- cesses. On successional and evolutionary time scales there must be sub- stantial selection for sexual reproduction to maintain genetic variability and flexibihty and to permit dispersal to new areas. The selective advan- tage of dispersal capability results from the heterogeneous nature of the microtopography that limits the expansion of clones, and the occasional creation of new unvegetated areas such as frost scars and drained lake basins that are not effectively colonized by clonal expansion. Population processes and evolutionary strategies deserve attention in future tundra studies. EFFECT OF GRAZING ON ALLOCATION AND POPULATION STRUCTURE Lemmings periodically graze wet meadow communities to such an extent that maximum aboveground biomass may be decreased from 100 g m'^ to as little as 5 g m"^ (Dennis 1968, Tieszen 1972b, Chapter 8). Sim- ulations suggest that the long-term effect of lemming grazing is to reduce total foliage by an average of 33% (Lawrence et al. 1978). Grazing alters graminoid allocation patterns in the tundra at Barrow and affects plant survival and community structure. Certain gross morphological features of Dupontia and other graminoids appear to be adaptive under a grazing regime. The shoot meristem and virtually all the storage and perennial tissues are located below ground, protected from grazing. Tissues that are available to grazers are only weakly lignified and are potentially replaceable at a minimal biochemical cost. Grazing may be a critical fac- tor leading to the dominance of the graminoid growth form at Barrow. Clipping experiments on Dupontia indicate that the movement of carbohydrate from one tiller to another is important in allowing Dupon- tia to survive intensive grazing. Dupontia leaves regrow rapidly following grazing because of the abundance of belowground carbohydrate reserves (Mattheis et al. 1976). Light clipping to simulate grazing can even result in a slight increase in the available carbohydrate pool of the rhizome, provided the chpped tiller remains attached to the tiller system (Mattheis et al. 1976), perhaps due to decreased shading and increased photosyn- thesis. After six consecutive clippings in a single season, to simulate the 178 F. S. Chapin III et al. 20 10 h VO Leaf Blade VI V2 0 I 2 7 20h Sheath and Stem Base ^ lOh 9 E o* z 0 0 I 2 2 6 0 12 6 V3 0 I 2 26 0126 0126 0 Number of Simulated Grazings 0 12 6 20 Rhizome 10 r-A r\ -^ r—i-n 1 — 1 2 6 FIGURE 5-18. Total nonstructural carbohydrate (TNC) con- tent in various plant compartments and tiller age classes o/Du- pontia fisheri following clipping at various intensities. Plants were clipped at weekly intervals to simulate grazing and were sampled on 20 August in moist meadow tundra. (After Mat- theis et al. 1976.) maximum grazing intensity that might be sustained in a lemming high, total nonstructural carbohydrate concentrations still remained high in rhizomes (Figure 5-18), and shoot weight was not affected. However, when a tiller no longer had access to the reserves of the entire tiller system, clipping decreased shoot weight substantially (Table 5-7), as has also been noted by Babb and Bliss (1974). Results of clipping experiments suggest that even chronic grazing would have little effect upon survivorship of vegetative tillers, because the tiller meristem is normally not damaged. However, at times of high lemming densities and inadequate food supply, lemmings may grub in the moss layer and remove the shoot meristem, killing the shoot. The im- pact of grazing upon reproduction is most pronounced 1) through the grazing of flowering shoots, since the inflorescence is lost, and 2) through the general lowering of the reserve status of the entire tiller sys- tem so that the chances of successful seed set are diminished. The number of new rhizomes produced is curtailed by grazing much less than is shoot production (Table 5-7). Grazing causes not only a de- crease in reserves but also a shift in the allocation pattern from shoot production toward greater rhizome production and vegetative reproduc- tion. Grazing causes a change in age class structure only because above- Control of Tundra Plant Allocation Patterns and Growth 179 TABLE 5-7 Effect of Simulated Grazing upon Dry Weight of Shoot Regrowth and Number of Newly Initiated (VO) Rhizomes of Dupontia fisher i Growing in Moist Meadow Tundra Treatment One clipping Two c lippings Clipped Ringed and Clipped Ringed and Tiller age Control only clipped only clipped Dry weight above ground (mg) First year tiller (VI) 7.7±3.4 8.8±2.4 2.5±0.7 5.8±1.8 0.3 ±0.1 Established tiller (V2 or V3) 6.8 ±2.8 1 1.4±2.1 4.0±3.0 0.0 2.0±0.6 Number daughter tillers (no. VOs tiller ') First year tiller (VI) 0.8 1.1 0.7 0.6 0.4 Established tiller (V2 or V3) 0.2 0.7 0.2 0.0 0.3 Note: Treatments involved clipping all leaves of the tiller at the moss surface (clip- ping) and/or severing all rhizome connections between the treated tiller and the rest of the tiller system (ringing). Tillers were clipped or ringed and clipped on 25 July, leaves were reclipped on 4 August, and shoots and rhizomes were harvested on 16 August 1973 (n = 10). Source: Mattheis et al. (1976). average numbers of newly initiated (VO) tillers are recruited into the pop- ulation in the year of a lemming high. Computer simulations suggest that the differential mortality and reduced competition due to grazing are not responsible for change in age structure (Lawrence et al. 1978). Hence, grazing by moderate lemming populations influences population struc- ture more by increasing recruitment of new tillers than by increasing mortality. The maximum stress that lemmings are likely to exert upon the nu- trient reserves of the wet meadow vegetation is simulated and shown in Figure 5-19. Simulations suggest that grazing depletes nitrogen and phos- phorus reserves more rapidly than carbohydrate reserves which, in turn, are depleted more rapidly than are calcium reserves. The strain on below- ground nitrogen or phosphorus reserves may not differ significantly in grazed or ungrazed situations until after four or five defoliations, be- cause the early part of the growing season is characterized by rapid up- ward nutrient translocation regardless of whether grazing occurs or not. Grazing would, however, prevent downward translocation of nutrients lost to herbivores and would likely affect growth primarily in subsequent years. In fact, detailed studies (Tieszen and Archer 1979) of various growth forms at Atkasook, Alaska, have not only shown that seasonal carbon balance is seriously affected (Chapter 3) but also that a single grazing event can reduce reserves in a manner which influences growth 180 F. S. Chapin III et al. Ungrazed CVI C o c o 'E E 3 D o D 2.0- 1,0- Grazed CM 600 en I/) ^ 400 E o CD 200 I 0 MM 5 15 5 3 24 June July Aug ^ Prior Years' Roots ^ Current Year's Roots 5 12 19 26 3 June H Rhizome m Stem Base 10 17 24 31 July n stioot 7 14 Aug FIGURE 5-19. Standing crops of plant nitrogen, calcium and biomass in moist meadow tundra with and without maximum lemming grazing, cal- culated assuming: 1) the total plant nutrient and biomass contents were altered through the season only by grazing (i.e. that no nutrient gain or loss occurred through uptake, senescence, or leaching, and that mainten- ance and growth respiration equaled photosynthesis); 2) translocation to aboveground parts after grazing occurred at the maximum rate observed during the 1970 growing season; 3) grazing occurred weekly and totally removed the aboveground compartment; 4) roots did not normally serve a storage function (nutrients and carbohydrates contained in roots were translocated upward only after all other reserves were exhausted); and 5) no new roots were produced when plants were intensively grazed. (After Chapin 1975.) I Control of Tundra Plant Allocation Patterns and Growth 181 and production the next year. Since this longer-term effect is most harm- ful to evergreen shrubs and then deciduous shrubs, and least harmful to graminoids, it is clear that vegetation units will change under different kinds of grazing regimes. VARIABILITY IN GROWTH AND ALLOCATION PATTERNS Substantial interspecific differences in the growth and allocation patterns of three dominant graminoids (Dupontia fisheri, Eriophorum angustifolium and Carex aquatilis) have resulted in niche differentiation. Carex is generally dominant in phosphorus-poor sites such as pond mar- gins and basins and rims of low-centered polygons, which receive low grazing pressure. In contrast, Dupontia predominates on phosphorus- rich sites such as polygon troughs. Eriophorum also tends to occur in more phosphorus-rich sites, particularly where vegetative cover has been broken by frost or human disturbance. Differences in allocation pattern between the three principal gram- inoids partially explain distribution patterns (Table 5-8). Leaf produc- tion and elongation occur earlier in the growing season and are more syn- TABLE 5-8 Characteristics of the Growth and Allocation Patterns of Dupontia fisheri, Carex aquatilis and Eriophorum angustifolium Dupontia Carex Eriophorum fisheri aquatilis angustifolium Leaf production* Asynchronous Somewhat syn- Somewhat syn- chronous chronous Shoot longevity* 3 to 4 (5) yr 4 to 7 yr 5 to 7 (8) yr Root longevity t 4 to 6 yr 5 to 8 (10) yr 1 yr Root elongation ability! 1 yr 2 to 3 yr 1 yr Seasonality of root elongation! Mid-late season Continuous Continuous Lateral root production! 90% 1st yr 2nd to 4th yr None Seasonality of lateral production! Mid-late season Continuous — Root origins! Predominantly rhizome nodes Stem base Stem base ♦Shaver (1976). !Shaver and Billings (1975). 182 F. S. Chapin III et al. O Temperoture, °C 0 4 8 12 4 8 12 16 20 24 28 Permafrost FIGURE 5-20. Patterns of root distribution of Dupontia fisheri (D.f), Carex aquatilis (C.a.) and Eriophorum angustifolium (E.a.) at peak season. chronous in Carex and Eriophorum than in Dupontia (Mattheis et al. 1976), suggesting that shoots of Dupontia might be able to regrow more rapidly following grazing than would shoots of the other two gram- inoids. However, in the ungrazed situation earlier canopy development by Carex and Eriophorum may give these two species a more favorable carbon balance that in turn allows them to invest in new structures, par- ticularly roots and rhizomes, to a greater extent than can Dupontia. The interspecific differences in growth and allocation patterns for the three species are more pronounced below ground than above ground (Shaver and Billings 1975), suggesting that at the Biome research area competition is more intense and niche differentiation more clearly deline- ated in soil than in air. Nutrient and oxygen concentrations, pH and temperature all vary substantially with depth. The three principal gram- inoid species have strikingly different rooting patterns and exploit differ- ent depths (Figure 5-20). Roots of Dupontia are concentrated in the top 5 cm of the soil where phosphate and potassium are most abundant and where temperature and aerobic conditions are most favorable for absorption (Shaver and Bill- ings 1975). Eriophorum has thin annual roots that grow vertically downward following the seasonal thaw (Bliss 1956, Shaver and Billings 1975). At the freeze/thaw interface, phosphorus may be highly available (SaebcT 1969). The disadvantages of the deep-rooting habit are that soils are colder and less aerobic and that deep roots are locked in frozen soil Control of Tundra Plant Allocation Patterns and Growth 183 the following season, which is perhaps why the entire root system of Eriophorum is replaced annually. Annual replacement avoids the cost of maintenance respiration during the second year when the roots may be less functional (Shaver and Billings 1975). Carex produces long-lived, thick primary roots that exploit intermediate soil horizons and thin se- condary roots that are quite abundant in surface horizons. Carex invests proportionately more tissue in roots than do the other two species and is most successful in nutrient-poor situations. Interspecific differences in growth and allocation, tiller interdepend- ence, and age class distribution lead to distinct population structures in the three graminoid species. Dupontia from the nutrient-rich habitat shows considerable tiller interdependence, low tiller mortality, and a uni- form age class structure (Shaver and Billings 1975, Allessio and Tieszen 1975a, Lawrence et al. 1978). Tiller interdependence may be important in allowing Dupontia to survive acute and chronic grazing. In contrast, the longer-lived Eriophorum tillers lose rhizome con- nections and become physiologically independent within two or three years (Shaver 1976). Heavily grazed Eriophorum tillers cannot rely upon the reserves of an interconnected tiller system and tend to occur less fre- quently in heavily grazed polygon troughs. Lack of tiller interdependence may be important in explaining Eriophorum 's apparent success in sexual reproduction, since its reserves may accumulate to support the inflores- cence rather than being continuously siphoned away into the rest of the tiller system. Allocation to sexual reproduction and the windblown seed dispersal pattern in Eriophorum are in part responsible for its abundance in disturbed habitats. An annual rooting pattern is adaptive in disturbed sites, where roots are subject to breaking by frost heaving. Little is cur- rently known about mortality of Eriophorum tillers, but the greater tiller independence in Eriophorum than in Dupontia may well lead to greater variability in recruitment and death in the former and hence more varia- tion in age class structure. Carex has a dual tillering pattern. Some tillers (clumping tillers) have very short rhizome internodes so that new shoots are produced ad- jacent to the parent tiller (Shaver and Billings 1975). Other (spreading) tillers have elongated rhizome internodes so that daughter tillers occupy space quite far from the parent tiller. The spreading pattern is most com- mon in phosphorus-deficient sites like polygon basins. The low phos- phorus-to-nitrogen ratio may stimulate rhizome elongation, minimizing competition between parent and daughter tillers. Such an effect has been proposed in temperate grasses (Evans et al. 1964, Langer 1966). These consistent differences in allocation patterns among the vari- ous graminoids are consistent and appear to explain their distribution patterns. Differences in life forms are more pronounced and are dis- cussed in Chapter 6. 184 F. S. Chapin III et al. SUMMARY The component processes of growth and allocation in graminoids of the coastal tundra at Barrow have compensated for low temperature to such an extent that these processes occur at nearly the same rates in situ as those observed in temperate grasslands. Temperature compensation may be achieved at substantial carbon and protein cost and thus be ac- companied by lowered reproductive output. Latitudinal temperature compensation has been demonstrated for shoot growth, photosynthesis, respiration and phosphorus uptake, but requires documentation for growth of belowground organs. Because the air warms faster than the soil, shoot growth and photosynthesis predominate early in the season, whereas root growth and uptake continue well after shoot senescence. Nutrients absorbed in one year have their most pronounced effect upon growth and reproduction in subsequent years. The lower levels of annual plant production in tundra than in com- parable temperate communities are more a consequence of shortness of the growing season than of the difference in ambient summer tempera- ture. This hypothesis is supported by computer simulations but has not been tested by field experiments or long-term observations. Allocation patterns are altered genetically and environmentally in a way that minimizes limitation to growth by any one environmental re- source but maximizes long-term survival through vegetative and sexual reproduction. Growth is limited simultaneously by several factors. Ex- perimental manipulatipns under field conditions indicate that phosphor- us and nitrogen strongly limit plant growth. Simulations predict that light, carbon dioxide concentration, and water availability also limit growth to a lesser extent. Low temperature limits production in a com- plex fashion involving all of the above environmental variables. The high root-to-shoot ratio of the graminoids compensates for low nutrient avail- ability. The environmental and genetic determinants of typical allocation patterns remain to be determined. Plant growth at Barrow is generally more strongly limited by an in- adequate supply of certain nutrients, especially nitrogen and phosphor- us, than by inadequate carbohydrate. Available carbohydrate levels are high in the graminoids, particularly in habitats of low nitrogen and phos- phorus availability. Carbohydrate levels are reduced by fertilization. We suggest that the low radiation environment of the Arctic limits plant growth more strongly by the indirect effects of low temperature upon nu- trient availability than by a direct effect upon photosynthesis. Shoots also become photosynthetically self-sufficient quite early in the season and depend upon rhizome reserves for nutrients more than for carbohy- drate. The decrease in rhizome weight that coincides with early season Control of Tundra Plant Allocation Patterns and Growth 185 shoot growth is primarily a consequence of high rates of maintenance respiration and translocation to roots. The evolutionary response of plants to the nutrient-limited environ- ment is to limit production to the formation of a small amount of tissue that is well supplied with nutrients and highly effective metabolically. This would explain the apparent paradox of relatively high nitrogen and phosphorus concentrations in leaves of plants that respond dramatically to nitrogen and phosphorus fertilizers, in contrast to graminoid crop plants where nutrient deficiency is evident in foliar nutrient analysis. Rapid upward translocation early in the season supports rapid shoot growth when radiation is most favorable for photosynthesis. Net down- ward translocation begins six weeks after growth commences, a full month before onset of obvious leaf senescence. The belowground carbo- hydrate and nutrient reserves appear to exceed levels required for growth in any given season and may allow the graminoids to successfully survive intensive lemming grazing and to regrow even after successive clippings. The Vegetation: Pattern and Succession p. J. Webber, P. C. Miller, F. S. Chapin III, and B. H. McCown INTRODUCTION This chapter analyzes paths by which the environment acts upon tundra plants and interdependence between environmental factors and plant growth form characteristics (Figure 6-1). The analysis emphasizes the distinguishing characteristics and the environmental distributions of some of the principal growth forms recognized for tundra regions by Webber (1978) (Figure 6-2). Although selection acts on whole individuals, growth form charac- teristics follow environmental gradients (Mooney et al. 1974) somewhat independently of each other. The objectives of this chapter are to iden- tify the patterns of vegetation and plant growth forms in the coastal tun- dra at Barrow, the principal factors controlling these patterns, and the paths of influence between the major environmental controls and the plant growth forms which lead to the patterns of vegetation observed in the field. The major pathways of plant succession and the effect of natural and other perturbations are also examined. To approach these objectives, the distribution of species, growth forms, and plant characteristics was determined along environmental gradients in the field, and the distribution of plant characteristics along environmental gradients was predicted from the physical and physiologi- cal information given in preceding chapters. The field distribution can be regarded as the realized niche, the predicted distribution as an indication of the fundamental niche (Hutchinson 1959). 186 The Vegetation: Pattern and Succession 187 Environmental Factors Irradiance Length of Tempera- Snow Soil Soil solar growing Wind Nutrients ture cover water oxygen infrared season Crazing Growth Form Characteristics Photosynthesis and Growth and Structure water relations allocation Population canopy height leaf resistance growth rates density leaf area index internal resistance leaf turnover patterns dispersal leaf clustering root resistance temperature optima vegetation/sexual leaf inclination leaf water potential temperature reproduction leaf absorptance water loss rates sensitivity age structure leaf width photosynthesis rates process rate at a age specific biomass leaf duration temperature optima given temperature age specific fecundity above belowground temperature chemical composition age specific mortality ratio sensitivity of plant parts (e g , grazing susceptibility root biomass process rate at a sugar, lignin, lipid. leaf biomass given temperature protein, nitrogen. woodiness enzyme storage in phosphorus, calcium) winter Growth Forms Deciduous shrub Evergreen shrub Single graminoid Caespitose graminoid Rosette forb Mat forb Cushion forb Bryophyte Lichen FIGURE 6-1. The principal environmental factors, plant growth form characteristics, and plant growth forms in the coastal tundra at Barrow. TOPOGRAPHIC VARIATION AND VEGETATION PATTERNS IN THE COASTAL TUNDRA Topographic and Environmental Variations Topographic variation in the coastal tundra at Barrow causes var- iation in the environmental factors which control the growth and dis- tribution of different plant species and growth forms (Table 6-1). The controlling factors can be indicated by the techniques of factor analysis or ordination (Whittaker 1967), The phytosociological gradients formed by indirect ordinations are interpreted as complex environmental gradi- ents which represent the probable major controlling factors. Using these techniques, 9 of a total of 17 measured environmental factors were sig- 188 P.J. Webber et al. DECIDUOUS SHRUB Salix arclica SINGLE GRAMINOID Dupontia fisheri EVERGREEN SHRUB Cassiope leiragona ROSETTE FORB Saxifraga nivalis MAT FORB Cerastium beeringianum ERECT FORB Polygonum vivipariim CAESPITOSE GRAMINOID Luzula confusa CUSHION FORB Saxifraga caespilosa BRYOPHYTE Pogonalum alpinum 'If^^f f r LICHEN Cladonia pyxidala FIGURE 6-2. Examples of the ten plant growth forms. (After Webber 1978.) nificantly correlated with the growth and distribution of different plant species and growth forms. The principal controlling factors were soil moisture, soil anaerobicity (indicated by soil odor of hydrogen sulfide), soluble phosphate in the soil and, to a lesser extent, snow cover. Sampling sites were classified in terms of topography, hydrology and soil morphology. Samples were taken from six microtopographic units: ponds, meadows, polygon troughs, tops of high-centered polygons o CO o s: o o CO QQ < -S; -a _ O E > 1 = ! a s; ■« ^ '^ 5 R c OO —' 00 3 a ZZ rs| -n E > rr, a. >> c o OX) > s: o S" S S > > i: o vo ^ "^ "^ "^ ^ i: E ~ 6 5^ = :< € C^ '^ ^ = J1 J= — OO 3 o o d q r4 — 1/^ d d OO -^ — r- q /-i ■^ R :;; -^ d - <=^ r- :2 ^. -^ r^ t o SO 3 ON C/^ C/5 On 00 u 3 tn '5 E '5 00 3 < ■a 00 c D. u-1 On ON 00 3 < „ -a 03 4> -- E Q. •a CO 00 00 -a c 03 1/3 03 £ "E 03 OO o 'J CO (*- L. 3 y3 03 E 'E 03 00 u. o 03 (*- 3 on = ±: II 00 3 5 ii 00 03 O '-> -n 03 E u o w 00 j: a >, t/5 o '[/5 x: c a 1) fU T3 X3 ^ 3 3 O oa m O CO — ■^ ^ -Q O OO a E a. CO C/5 t/5 dJ c 2 c o t/5 "o o E 3 C c 1) ^ " H o 4J u— D. ^.^ >. Q. "^ 1) ^ CJ o X 03 OJ N IE -^ t. i D. c/; VI O u a ■q. 0^ E x: CO E o O ;— u- c C CO E o •a D 03 _3 00 ON a> m 1- ~^ Um lU L. CO t/5 03 S X5 _3 03 > X ^ 1} u. Ponds'' Dnn.4.1 Moist UU.t P°"<1* Meadows )S5l,dow. u> c Dry Troughs 01 o Meadows Tops' T3 Rims Tops^ -J LO moisture W LO soluble phosphote H/ ^ Wet Meadows a> o Troughs n Dry 2 Meadows Ponds u> Tops o Moist Meadows O. Rims Tops^ (U ^ 3 Basins O v> 1 Ci Ponds' <«J LO moisture H/ FIGURE 6-3. The distribution of microtopographic units within the three principal axes of the ordination. The axes are soil moisture, soil odor of hydrogen sulfide, and soluble soil phosphate. Low basins are the centers of low-centered polygons. Tops' are the centers of high-centered polygons with little or no peat, Tops^ are the centers of high-centered polygons with thick peat, Ponds' are those with no significant flow of water, and Ponds^ are those with flowing water. (After Webber 1978.) and rims and basins of low-centered polygons. Two categories of ponds — those with and without flowing water — and three types of meadows — dry, moist and wet — were distinguished. Two categories of polygon tops — those with little or no peat at the surface and those with thick sur- face peat — were distinguished. The different microtopographic units can be plotted within the axes of the indirect ordination (Figure 6-3). The sequence of units along the soil moisture gradient from low moi':ture to high is: tops of high-centered polygons with shallow soil, polygon rims, dry meadows, moist meadows, basins of low-centered polygons, tops of high-centered polygons with thick peat, polygon troughs, and wet meadows. The sequence along a gradient of soil hydrogen sulfide from low to high is: tops with thick peat, rims, tops with shallow soil, dry meadows, troughs, wet meadows, moist meadows and basins. The sequence along a gradient of soil soluble phosphate from low to high is: ponds with no flow, basins, dry mea- dows, moist meadows, rims, tops, troughs, and wet meadows. Basins of low-centered polygons and polygon troughs have the greatest snow ac- cumulation and are the last microtopographic units to be free of snow, while polygon rims and tops of high-centered polygons have only a thin snow cover and are first to be snow-free. The duration of snow cover af- fects the length of the growing season, which varied from 30 to 42 days in 1973 (Table 6-1). These observations coincide with the conclusions of Wiggins (1951) and Britton (1957) who emphasized the control by the microrelief of the substrate conditions, which in turn control the distribution of plants. In contrast to the results of most tundra ordinations (e.g. Webber 1971, The Vegetation: Pattern and Succession 191 Webber et al. 1976) and other tundra studies (Gjaerevoll 1956, Bliss 1963, Scott and Billings 1964), snow cover does not emerge as a major controlling factor from these analyses. The variations in microrelief among most of the sampled locations are insufficient to produce a pro- nounced variation in snow cover. However, where ravines, creek banks, beach ridges and snow fences occur, snow cover becomes an important factor influencing vegetation distribution. Vegetation Turnover Patterns As discussed in Chapter 1 (Table 1-4), eight major vegetation types or noda were distinguished in the vegetation of the coastal tundra at Bar- row. Most species occur in more than one vegetation type, but the types are distinguished by the presence of certain indicator species and the con- sistent importance of other species. Each vegetation type has a reason- ably distinct standing crop composition at the period of peak above- ground biomass (Table 3-2). The Cochlearia meadow is a rudimentary vegetation restricted to recent alluvium and disturbed sites; it is therefore not included in the discussion. Each vegetation type has a unique distribution within the three axes of the ordination (Figure 6-4). The vegetation types are numbered from Luzula heaths to Arctophila pond margins, following the primary con- trolling gradient of increasing soil moisture which is associated with in- creasing snow cover. Along a gradient of increasing hydrogen sulfide, the sequence of vegetation types is from Luzula heath, Salix heath, and Carex-Poa meadow, all found on well-drained sites, to Carex and Du- pontia meadows, Arctophila pond margin, and Carex-Oncophorus mea- dow, all found on poorly drained sites with still or stagnant water. Along moisture H/ LO soluble phosphate hi lo moisture HI FIGURE 6-4. The distribution of seven mature vegetation types within the ordination. I) Luzula heath, II) Salix heath. III) Carex-Poa meadow, /K> Carex-Oncophorus mead/ow, V)Y)\ix>oni\3. meadow, K/y Carex-Erio- phorum meadow, VII) Arctophila pond margin. (From Webber 1978.) 13 s: o o s: o o I U fiQ < fN s: ?3 '^ a c o It 00 > ^ i^ i ^ n ^ •0 -^ nj <«; 0 On ON ON § 0 r-1 r»1 !/5 0 0 0 0 "" 0 0 0 0 0 c 0 vO 1^ [^ r^ 0 NO m 0 rn 0 Ov ON ca ON /-i NO t^ O NO rn O > JH T3 3 O cj ■^ CO 2i x> CO CO 2 -o ^ § CO 00 CO CO £ O X) 00 T3 1> C a> > i CO ao — '^ * 00 -1 < < ^ CO T3 C CO 3 O T3 CO 00 c T3 3 CJ ID ID > o c OO 00 c 3 C a> > 0 r 3 X) u < 0 1 Z CO 192 The Vegetation: Pattern and Succession 193 the gradient of increasing soluble phosphate, the sequence of vegetation types runs from Carex-Eriophorum meadow, Carex-Oncophorus mea- dow and Carex-Poa meadow to the heaths, Arctophila pond margin and Dupontia meadow. The three vegetation types associated with the lowest phosphate concentrations occur on sites that receive no drainage water or influx of minerals, while the types associated with higher concentrations are found on mineral soils on sites receiving an influx of drainage water. Turnover rates (grams incorporated annually per gram standing crop) in a particular vegetation type depend on the species or growth forms present (Table 6-2). Turnover rates for aboveground biomass are about 1.0 yr"' for those vegetation types that are mainly composed of graminoids and forbs. In the Salix heath and Carex-Poa meadow, woody dicotyledons are common and biomass turnover rates are about 0.6 yr"'. The woody dicotyledons in these vegetation types have low turn- over rates; several individuals of Salix pulchra were at least 20 years old, as determined by counts of terminal bud scars on the branches. Turnover rates of standing dead range from 0.3 yr"' in the Luzula heath to 3.2 yr"' in the Arctophila pond margin. The turnover rates for standing dead in the Carex-Oncophorus, Dupontia and Carex-Erio- phorum meadows are 1.0 yr"' to 1.3 yr"'. The turnover rate of standing dead is affected by the rates at which standing dead is blown or washed away from the site or pressed prostrate by snow and rain. Turnover rates of litter and prostrate dead are from about 0.2 yr"' in the Salix heaths to 1.8 yr"' in the Arctophila pond margin. Turnover rates are 0.6 to 0.9 yr"' in the Carex-Oncophorus, Dupontia and Carex- Eriophorum meadows. These turnover rates are affected by the rates of removal of material from the site, by rates of decomposition and by rates of incorporation of the vascular material into belowground material, which depends partly on the rate of vertical growth by moss. The turnover rates for all aboveground dead material are about 0.2 yr"' in the heaths, 0.4 yr"' in the Dupontia and Carex-Eriophorum mea- dows, and 1.2 yr"' in the Arctophila pond margin. These turnover rates imply residence times of dead material of less than a year in the Arcto- phila pond margin, to 2 to 2.5 years in the meadows, to 6.3 years in the heaths. Residence times of aboveground material in all vegetation types are between 1.9 and 7.7 years, and are about 3.4 years in the Dupontia and Carex-Eriophorum meadows. Belowground turnover rates are estimated from root longevity data when available (Chapter 5). Root longevities are about 5 years in Dupon- tia and 8 years in Carex aquatilis, giving turnover rates of 0.2 to 0. 12 yr"' for the Dupontia and Carex-Eriophorum meadow types. Lower turn- over rates are expected for the belowground parts of the woody dicoty- ledons. 194 P.J. Webber et al. GROWTH FORMS AND ENVIRONMENTAL CONTROL Definition of Growth Forms The factors controlling the distribution of plant populations could not be analyzed at the scale of resolution of individual species, however desirable such analyses might be, because of the lack of complete data on all species. Thus the plant species were grouped into growth forms based on aboveground characteristics (Figure 6-2), Since the early work of Raunkiaer (1934), the importance of plant growth forms as strategic adaptations to the tundra environment has been emphasized (Bliss 1962a, Tikhomirov 1963, Chabot and Billings 1972). The evidence is that growth forms are selected in different habitats, and therefore are a mean- ingful basis on which to analyze different plant-environment interac- tions. The growth form categories are based primarily on the nature of the shoot habit, although some categories, such as bryophytes and lichens, are systematic or phylogenetic in character. Nevertheless, bryo- phytes and lichens seem to be valid growth forms in the tundra. Woody shrubs in the coastal tundra at Barrow are all of low stature and many are prostrate. Shrubs are subdivided on the basis of being ever- green or deciduous. The genus Salix is the principal representative of the deciduous shrub while Cassiope tetragona and Vaccinium vitis-idaea represent the evergreen shrub. Herbaceous plants with elongated, narrow leaves are represented by the graminoids, i.e. grasses, sedges and rushes. The graminoids are subdivided into those with crowded, bunched shoots, here called caespitose graminoids, and those with well spaced individual shoots, here called single graminoids. Luzula confusa and Eriophorum vaginatum are examples of caespitose graminoids and Dupontia fisheri and Carex aquatilis represent the single graminoids. Four growth forms are recognized for the broad-leaved herbs or forbs. Acaulescent, or essentially stemless, plants with a rosette of radi- cle leaves, such as Saxifraga nivalis and Pedicularis lanata, are called rosette forbs. Broad-leaved herbs with erect leaves or leaves supported into the canopy on long petioles or on an erect stem are called erect forbs and are represented by Polygonum viviparum and Petasitesfrigidus. Mat forbs such as Cerastium jenisejense and Stellaria humifusa have tightly matted, often long, prostrate stems with leaves along the length of the stems. Cushion forbs have short, crowded stems, often coming from a single tap root, which give the plant a hemispherical shape illustrated by Saxifraga caespitosa and Silene acaulis. The Vegetation: Pattern and Succession 195 Distribution of Growth Forms An analysis of the distribution of plant growth forms along the en- vironmental gradients shows that some growth forms have high fidelity for a vegetation type, e.g. the evergreen shrub in the heaths, while other growth forms are distributed through several vegetation types, e.g. the single-stemmed graminoids and the mosses. The above- and belowground standing crops and foliage area indices of the vascular plants form clear patterns along the complex environ- mental gradients (Figure 6-5). A comparison of the patterns of foliage area index and productivity shows a general correlation between these two variables similar to that described by Miller et al. (1976). The principal factor controlling the distribution of bryophytes ap- pears to be slight differences in microrelief which influence soil moisture regimes. Bryophyte biomass is low on sites with high hydrogen sulfide and highest in the presence of moderately low values for soil moisture and soluble phosphate. The standing crop of forbs and woody dicoty- ledons is highest in dry, well-aerated sites with moderate levels of phos- phorus. Aboveground biomass of graminoids increases along the moisture gradient from the Salix heath to the Arctophila pond margin. The Luzula heath has a higher graminoid biomass than the Salix heath because of the caespitose graminoids, such as Luzula confusa, which are abundant on dry sites. In all types except the two heaths, which are relatively dry, the aboveground standing dead is less than the aboveground live for both graminoids and forbs. The decay index for litter and prostrate dead vas- cular plant material, which is the ratio of net aboveground productivity to litter and prostrate dead, is lower on dry sites than on wet sites, and standing dead material is incorporated into prostrate dead and litter frac- tions more easily on the wet sites. The distribution of belowground biomass in relation to environ- mental factors differs from that of aboveground material. Belowground biomass is greatest in the most anaerobic soils, and in soils with moderate moisture and high concentrations of soluble phosphate (Figure 6-5). The ratio of aboveground to belowground biomass presents yet another pic- ture, with the lowest ratios (1:40) in moist, partly anaerobic soils with low soluble phosphate concentrations. The ratio increases along the moisture gradient and from anaerobic to aerobic soils. These patterns of belowground biomass support the conclusion (Chapin 1974a, Wielgo- laski 1975c, Chapter 5) that vascular plants develop greater amounts of absorptive root tissue on anaerobic, phosphate-poor soils. But they are also partly an artifact of the changing growth form spectrum along these gradients, i.e. a shift from woody-stemmed species to rhizomatous spe- 196 P.J. Webber et al. Aboveground Biomoss of Graminoids.g nrT Z.0 moisture content/// ^Osoluble photpt^ote/// Aboveground Biomoss of Forbs.g m~ LOmoit^urt content/// insoluble photphote/// Aboveground Biomoss of^ Woody Dicotyledons, g m~ -2 ^£7mol(ture content/// LOto\ub\» phosphate/// Stonding Crop of Green Bryophyte Biomoss, g m' ,-2 ^C7moisture content/// Z.<7$oluble phosphate/// Biomoss of Lichen, g m"^ ^Cmoisture content/// LOto\\ib\t phosphate/// Belowground Biomoss, g m"^ i. , c . $r Net Aboveground Vascular -^2 -I Productivity, g m yr 100 •--50-- LO moisture /// iOsoluble phosphote/// Live Vascular Foliage Area Index iOmoisture content/// /.Osoluble phosphate/// Vosculor Litter and Prostrate Dead,g m $, , . ^ -2 Z.0maisture content/// ^£7soluble phosphate/// Litter Decoy Index Z.Omoisture content/// /.^soluble phosphate/// Above:Belowground Biomoss Ratio /.^moisture content/// Z.Osoiuble phosphate/// Belowground Dead Material, g m"^ /.^moisture content/// insoluble phosphote/// -2000- -4000 i(?moisture content/// i. ? s -= 5 a q r- dec rn r*1 J? :^ ■a ***M 2f <3 Cj G ^ g 3 rn — .1 zz ^^ c^ " "•** >«. 13 a. 0 0 QJ S 0 ^ s (N f*^ "S - ^ r^ V. •s; si 1 o a E to s - 1 i O ^ E _> -•>« ^ .2 ^ 1 2 00 ^ 5 ■S 0 ^ 5. 1 d ft; o o ^ ■o :2 .^ i Jj 0 H :S •S E f== 3 0 00 V. o =5 t^ 00 « S ^ ^ = ■V- 0 0 5 S ll ^ E 3 c ^ 2 -^ 3 >o ■* •* !J 00 »/ Major s at Barr 3 0 X) > 3 q ibution c ome Site c 5 ? > >" 3 .5 1 rn h ^ .^ ^ ■^^ h^ <=> u .ti ,— V *? C/D ™, ^ 1 j: ^ « j: U « NJ 5 0 0Q 3 3 .i< «< "5 H -J to 1- ■* r^ •* — 1 — d d i d > _ 0 0 > „,.^ -0 ■a to ra c ^ u AS u *5b "-^ E > E L« 0 CO S i 5 E 1) E CI 1 0 E 1 1 •0 c 0 a G c .3 c a. 1 0 1 c 1 I •-; ■< 0 i^ 0 a; it 5. a 0 13 5 3 13 c c U Q c '^ O •D C o Lf) I 00 5-20 years on wet sites Little recovery on dry sites New terrestrial communities in equilibrium FIGURE 6-12. Impact of crude oil upon the coastal tun- dra at Barrow. Recovery times are estimated from obser- vations by Wein and Bliss (1973), Deneke et al. (1975), Hutchinson et al. (1976), and Walker et al. (1978). lively retard nutrient cycling and decrease nutrient availability in an eco- system that is already nutrient-poor and highly dependent upon internal recycling of nutrients. Oil is hydrophobic and once within the soil greatly reduces water movement both into the soil and from the bulk soil to plant roots (Raisbeck and Mohtadi 1974, Everett 1978). Because nutrient The Vegetation: Pattern and Succession 217 movement within the soil is dependent upon water movement, the hydro- phobic nature of oil-contaminated soil may decrease nutrient as well as water availability to plants. SUMMARY Within the prevailing coastal tundra macroclimate, the topographic position of a site causes the variation in environmental factors which in turn control the growth and distribution over the tundra surface of dif- ferent plant species and growth forms. The complexes of controlling en- vironmental factors were identified by indirect ordination. They are, in order of importance in explaining the overall variation of the vegetation, the complexes of soil moisture, soil anaerobicity, soil phosphate, and snow cover. The distribution of vegetation types, plant growth forms, and various vegetation and growth form attributes such as standing crop of above- and belowground material, productivity, foliage area, turn- over rates, and diversity are described in terms of the controlling en- vironmental complexes. The principal growth forms recognized at Barrow are single gramin- oids, erect forbs, deciduous shrubs, and bryophytes. Others that may be locally abundant are caespitose graminoids; cushion, mat and rosette forbs; evergreen shrubs; and lichens. Dry, exposed sites with little snow cover usually contain a pre- ponderance of caespitose graminoids and lichens. They may also have evergreen and deciduous shrubs and cushion forbs. Mesic sites tend to have the most forbs but have an abundance of bryophytes and deciduous shrubs. Single graminoids are abundant over much of the tundra and are dominant in moist and wet sites along with bryophytes. The distribution of plant growth forms within the tundra is related to structural and functional characteristics, such as leaf longevity, timing of leaf growth, location of perennating organs, location of stored nutri- ents, leaf resistance to water loss, photosynthetic rates, carbon and nutri- ent costs of making new leaf material, and the location of absorbing roots; to the availability or abundance of wind, water, light energy, heat, and inorganic nutrients; and to the influence of grazing animals. A hypothetical scheme is presented for the major paths of plant suc- cession in the Biome research area. The scheme is based on the thaw lake cycle, and the major vegetation types are seen as phases in the cycle. Plant colonization of stable surfaces may take only a few years, but the major vegetation types are generally long-lasting and change primarily in response to physical changes, such as microrelief and drainage, rather than in response to biological changes, as a result of competition. Most human impacts upon tundra are within the range of natural 218 P.J. Webber et al. perturbations, and their severity can be predicted from an understanding of the controls over ecosystem processes in natural tundra. Vehicle tracks that decrease canopy cover but do not destroy the vegetation or depress the soil surface create an impact comparable to intensive lemming graz- ing and the tundra may recover within a few years, or it may not. Dis- turbances that depress or destroy the vegetation mat initiate a series of changes in physical environment comparable to those occurring in the thaw lake cycle, and result in a corresponding recovery time, probably thousands of years. The Soils and Their Nutrients P. L. Gersper, V. Alexander, S. A. Barkley, R. J. Barsdate, and P. S. Flint INTRODUCTION Soils of the coastal tundra are formed under conditions of low tem- perature and high moisture. Mean annual precipitation is low, but rela- tive humidity is high and drainage is impeded by permafrost; conse- quently, soil moisture content is high. Low temperatures and high mois- ture contents lead to the accumulation of organic matter. Because of the cold, impervious permafrost there are strong gradients of temperature and oxygen saturation within the thawed soil, but soil profile differentia- tion is retarded by the restriction of downward leaching and associated chemical transformations. Visibly distinct horizons are largely associated with organic matter, a product of organic input from primary produc- tion and physical redistribution via frost churning processes. Overall, the soils of the coastal tundra at Barrow are similar to those of other tundra areas. In the data gathered by French (1974) on 27 soils from nine tundra sites studied during the International Tundra Biome ef- fort, soil from a wet meadow of the Biome research area at Barrow falls near the middle of the range of values observed for most soil parameters, although this soil is somewhat wetter than those of most other circum- polar sites. In an analysis based on climate and soil factors (Rosswall and Heal 1975), five microtopographic units from the coastal tundra at Bar- row (meadows, polygon troughs, rims and basins of low-centered poly- gons, and centers of high-centered polygons) were found to be similar to each other, but were also very similar to the meadow site on Devon Island and the moss turf and moss carpet on Signy Island, Antarctica. The soils near Barrow are different from those at Prudhoe Bay, 320 km to the east, which are calcareous and lower in organic matter (Douglas and Bilgin 1975, Everett and Parkinson 1977). The properties of the soils of the coastal tundra at Barrow are therefore the products of both the climatic factors common to arctic tundra regions and the specific geologic history of this area. 219 220 P. L. Gersper et al. SOIL PHYSICAL PROPERTIES AND NUTRIENTS Biological processes in the soils of the coastal tundra at Barrow oc- cur in an organic-rich layer less than 50 cm thick that is thawed for less than four months of the year. This layer contains over 70% of the living biomass of the tundra ecosystem. In it, roots grow and take up nutrients and water, organic matter decomposes, invertebrates graze and prey upon one another, and lemmings burrow for summer protection from predators. This shallow layer of thawed soil is the reservoir from which inor- ganic nutrients are initially supplied to the living organisms. Calcium, magnesium, potassium and sodium are all retained by the cation ex- change complex, which is made up largely of humified soil organic mat- ter. The organic matter itself contains the major pools of nitrogen and phosphorus. However, most of the available nitrogen is in the form of ammonium and is retained on the soil exchange complex, while most of the available phosphorus is bound to iron or aluminum ions. The pools of available nutrients are highly variable, both spatially, because of the different kinds of soils associated with the different microtopographic landforms, and temporally, in response to fluctuations in environmental conditions. The underlying permafrost affects the nutrient supply through its effects on temperature gradients in the thawed soil and by isolating large quantities of nutrients contained in the frozen soil. Organic Matter The predominant characteristic of the soils is their high proportion of organic matter. More than 95% of the total organic matter in the ter- restrial tundra ecosystem is below the ground surface, and one-third is in the upper 10 cm of soil, where biological activity is concentrated. The large amounts of organic matter impose a particular structure on the soil, influence the flux of moisture, oxygen and heat, and modify the chemical properties, particularly in the cation exchange complex. The large pro- portion of organic matter in these soils has a strong effect on the nutrient supply, as is typical for arctic tundra soils (Babb and Whitfield 1977, Chapin and Van Cleve 1978). Carbon and nitrogen, in an average ratio of 20:1, make up from 10 to 40% of the total soil weight (Figure 7-1). Total carbon contents in the upper 15 cm of soil typically range from about 12,000 to 16,000 g m"^ but may be less than 10,000 g m"^ in comparatively warm, nutrient-rich, wet Pergelic Cryaquepts of polygon troughs, where decomposition rates are high. The Soils and Their Nutrients 221 Meadow Avg C:N = 20.4 Rim 20.2 E 1 17^ N| o 1 ^a^ . 10 J W SI ^ Q. r ^a ^ 20- L i 30 ] 1 40 20 0 2 40 20 0 2 40 20 0 2 40 20 0 2 Percent Carbon and Nitrogen in the Soil 10- I 20 30- 5000 0 5000 0 g Carbon m (5cm depth) r-i ■ 1- 5000 0 -I 5000 FIGURE 7-1. The percentage of carbon and nitrogen in soils from different tundra microtopographic units, including meadows, basins of low-centered polygons, polygon troughs, and rims of low-centered polygons. In nearly all the soils, organic matter in the surface horizons is most- ly fibric; the degree of decomposition increases with depth. Fibric organ- ic matter includes slightly decomposed, reddish- to yellowish-brown fib- rous materials whose generic characteristics can be recognized and which are usually interlaced by an abundance of living roots and rhizomes. Live belowground plant parts averaged 660 g m"^ in 1972, most of it in the up- per 10 cm (Dennis et al. 1978). Fibric materials are commonly associated with wet meadow and polygon trough soils, but such materials rarely dominate the entire soil profile. Sapric inclusions of black to dark reddish-brown, generically uni- dentifiable, fibrous to granular organic materials which disintegrate completely under the mildest mechanical manipulation are commonly found below the surface horizon. In the better drained, more highly oxi- dized soils such as are found on the tops of high-centered polygons and the rims of low-centered polygons, highly decomposed sapric organic matter may predominate throughout the entire active layer and continue down into the permafrost. Hemic materials, those of intermediate decomposition, represent the most common form of organic matter in the soils. These generally range in color from dark grayish-brown to dark brown, and in nearly all cases 222 P. L. Gersper et al. identifiable plant components can be recognized which disintegrate only after considerable mechanical manipulation. Such materials commonly dominate the entire soil profile in basins of low-centered polygons and in some of the large, very low relief orthogonal polygons of mesic and wet meadows. Differences in the amount of organic matter over lateral distances of only 1 m can be more marked than differences with depth. In an appar- ently uniform area of wet non-polygonal terrain, the organic carbon in two soil profiles sampled only a meter apart differed by a factor of 2. However, differences between microtopographic units are generally 1.5 to 2 times greater than variations within these units. Bulk Density, Porosity, and Texture The bulk density of a soil has a strong effect on heat conduction and temperature, depth of thaw, soil water content and movement, soil por- osity and air content, and the penetration of roots. Bulk densities in the soils of the coastal tundra at Barrow range from about 0.05 to about 1 .50 g cm'\ Differences in bulk density (D,,) along a microtopographic gradi- ent are strongly associated with the percentage of organic carbon (C.„«) present: C.,(% by weight) = 1.1-57.15 log(A,) r = 0.93, n > 200 The lowest bulk densities are found in surface horizons, which have high contents of fibric organic matter, live and dead roots and moss (Fig- ure 7-2). Within the thawed layer, soils of wet and mesic meadows have bulk densities that range from moderately high to low, with a tendency toward increasing bulk density with increasing wetness. Rims of low- centered polygons and centers of high-centered polygons have soils that range in bulk density from low to intermediate, depending on the nature Meadow Basin Trough Rim BULK DENSITY, g cm FIGURE 7-2. Bulk density of soils from four dif- ferent microtopographic units. The Soils and Their Nutrients 223 of the soil formed before elevation of the rims and centers. Soils of the basins of low-centered polygons have low bulk densities. The general in- crease in bulk density between soils of basins and rims of these microtop- ographic units is partly due to the more highly decomposed organic mat- ter in the rims and partly because the rims generally contain a higher con- tent of fine-textured mineral matter. The relatively low content of organ- ic matter in the Cryaquept soils found in most polygon troughs results in high bulk densities, although some soils in the troughs are highly organic, with correspondingly low bulk densities. The highest bulk den- sities of the coastal tundra at Barrow are generally encountered in the sandy soils along stream margins. The bulk density of a tundra soil, as it reflects mineral content and water-holding capacity, is an important determinant of the depth of thaw (Gersper and Challinor 1975). Under similar moisture regimes, mineral soils generally thaw more deeply than organic soils because of their higher bulk density and consequent increased heat conduction. However, soils with low bulk density also tend to have high moisture content, which is also associated with deeper thawing. Although the variables are strongly covariant, multiple regression analysis suggests that bulk density alone accounts for 35% of the variation in thaw depth, while moisture content accounts for an additional 31%. The soils are highly porous, with a range of porosity from 50 to 65% for mineral layers, increasing to more than 90% for organic layers (Fig- ure 7-3). Thus, these soils are more porous than mineral soils of temper- ate regions, which range from 30 to 60% porosity (Hausenbuiller 1974). Pore Volume, % 25 50 75 GO 0.25 0.50 Bulk Density, g cm 0.75 3 FIGURE 7-3. Bulk density, per- centage total pore volume, and 1 00 percentage air-filled pore volume in soil of a moist meadow. 224 P. L. Gersper et al. Percent Sand FIGURE 7-4. Particle size distribution and tex- ture of mineral horizons in soils from A) Aerie Pergelic Cryaquepts on sloping areas marginal to Footprint Creek (10 sam- ples), and B) Histic Per- gelic Cryaquepts and Pergelic Cryohemists in a moist meadow (86 sam- ples). Textural diagram after Soil Survey Staff (1975). The very small particles and particle aggregates of sapric soils form a rather dense and relatively impermeable mass that is slow to transmit moisture or oxygen (Figure 7-3). Because of their highly aggregated con- dition, sapric soils have many small pores and retain large quantities of moisture, even in topographic positions that would normally be con- sidered xeric. Further, the porosity of sapric soils may be high, especially in surface horizons, because of repeated ice segregation, which produces lenticular openings. However, vertical permeability is usually very low. In comparison, fibric materials are of low density and have a very large proportion of interconnected, free-draining macropores that permit rapid movement of air and water in all directions. Soils on low topo- graphic positions have fibric surface horizons that remain filled with cir- culating water throughout much of the summer. Horizons containing hemic material have intermediate soil moisture properties. Soils with high contents of clay- and silt-sized mineral materials gen- erally tend to have very fine pores and consequently very low permeabil- ity to air and water. Mineral layers in soils of the coastal tundra at Bar- row are generally of this type, although they are often admixed with por- ous organic matter and thus are more permeable. The dominant textures of the mineral horizons (Figure 7-4) are silty clay and silty clay loam, with some clays and a few soils of coarser texture. Mineral fractions of Histosols tend to be finer textured (silty clays) than the Inceptisols, which are silty clay loams, although Cryaquepts of polygon troughs and frost boils commonly have the finest textures. Sandy-textured and loam- textured soils are not common at the Biome research area, and appear to be restricted to alluvial positions and the stream banks along Footprint Creek. The Soils and Their Nutrients 225 TABLE 7-1 Seasonal A verages of the Percentage of Moisture, Water Potential, and Redox Potential in the Upper 15 cm of Soil, and Soil Temperature During Summer 1972. Micro- Dominant Percent Water Redox Temperature, topographic vegetation moisture potential potential 10 cm unit type (g water gdw') (bars) (mV) (°C) Slough VII 292 ND 306 8.1 Meadow IV 331 -0.009 405 6.3 Rim II or III 118 -0.044 511 7.3 Basin VI 208 -0.033 423 7.7 Trough V 265 -0.008 430 8.0 Top I or II 158 -0.041 487 6.5 Soil Moisture and Aeration Because of their very high content of organic matter and related high porosity, soils of the Biome research area have a high capacity to hold water. Field moisture contents range to over 1000% dry weight. However, the extreme variability in bulk density and porosity makes ex- pression of soil moisture content on the basis of volume, rather than weight, more useful. In terms of soil volume, maximum moisture content in the upper 15 cm of meadow soils averaged 47% during the summers of 1970 and 1971. The underlying mineral layers ranged between 55 and 60% by volume. In these soils the minimum moisture content in the up- per 15 cm averaged 67% in 1970 and 65% by volume in 1971. During the summers of 1970-73 the soils in much of the nonpolygonal terrain and in the lower-lying areas of polygonized terrain remained almost completely water-saturated within approximately the 5-to 15-cm depth interval. On polygonized terrain, in the relatively warm and dry summer of 1972, only the centers of high-centered polygons had moisture contents of less than 65% by volume in the upper 10 cm of soil on 31 July. In 1972, moisture tension was measured using tensiometers in soils along a moisture gradient in the polygonal terrain (Table 7-1). In the dri- est soils, on the top of a high-centered polygon and on a polygon rim, water potentials averaged over the 0- to 5-cm depth interval were never lower than -0.070 bar during the summer (Figure 7-5). Potentials at 0- to 5-, 5- to 10- and 10- to 15-cm depth intervals at a given location were similar, but with a tendency toward progressively higher potentials with depth. The low water potentials in the soil indicate that most of the water present is available for plant uptake. Variations with time were similar at all three depth intervals and within each microtopographic unit across the entire moisture gradient (Figure 7-5). Despite the general wetness, 226 P. L. Gersper et al. 0.02 0.01 - 0- -0.01 ■S -0.02 -0.03 o a. 5-0.04 I -0.05 - -0.06 - -0.07 1 I T-- 1 1 V \ >s ^ / ^ N y -^-^__\ \\Meadow / _ /> ^ - ^^^s. ^p:^ Trough /^ - ^"^^^^ 1 ^'^^ 1 \ \ ^ v\ X 1 -i \\ "^^-^ / // \ A v:/ / / \ ^\ / \Basin / v \ ^ ^ ^ / // \ \ ^-^ rs\ r> // \ \ ^^^ // \ \Topof high ^ center V hi \\ \ / ^ \ V // \\Vy/ '-;/ / \ ^ / / \ / \ — '' / " \y/^ V RirK / - 1 1 1 1 1 1 5 10 15 Jul 20 25 5 Aug FIGURE 7-5. Seasonal courses of soil water potential in the upper 5 cm of soils of different microtopographic units in 1972. 0 Slough Oxygen, % 0 20 ^ " 10 S" 15 20 25 1- Site 2 Basin Trough Oxygen, % 0 20 40 60 0 20 Site 4 Basin Trough Top Oxygen, % 0 20 40 0 20 40 0 20 40 I I I I I I I I I I I I I I I I 1 FIGURE 7-6. Average oxygen saturation in soils of different microtopographic units in 1972. (Benoit, unpubl.) The Soils and Their Nutrients 227 moisture regimes are sufficiently influenced by microtopography to re- sult in differences in species composition and physiognomy of the above- ground plant community, and in the soil microflora and fauna. Oxygen concentrations at 10 cm ranged from 40 Vo saturation to 0%, with the highest values found in the tops of high-centered polygons and the lowest in the wet sloughs (Figure 7-6). As the soil thaws, the depth to fully anaerobic conditions follows the thaw front downward, and by mid-season oxygen saturation is zero at about 25 cm depth. Although the soil continues to thaw, the aerated layer in the wet meadow soil seldom exceeds 25 cm because oxygen flux is impeded by high bulk density and water saturation in the mineral horizons (Benoit, unpubl.). Alternating organic and mineral layers in the soil can produce a very complicated pattern of air and water movement through the active layer (Figure 7-3). Histosols generally lack a continuous mineral layer and con- tinue to drain freely as thaw progresses. These soils are unsaturated in the upper part, permitting air movement within the soil, except in wet summers. In contrast, the mineral layers of Inceptisols restrict drainage, and these soils often remain at or near saturation throughout the sum- mer. Air movement is restricted, and reducing conditions prevail in and below the mineral layers. Cation Exchange Capacity and Acidity The cation exchange capacity (CEC) of the soils is dominated by the organic fraction. Thus there is a strong correlation between CEC and or- ganic carbon content. For example, within the upper 30 cm of meadow soils this relationship was: CEC[meq(100 g)"'] = 2A5C^^^{%) + 15.54 r = 0.97, n = 86. This equation indicates that the mineral clay fraction, which makes up an average of 21% by weight within the upper 30 cm of these soils, contrib- utes 15 milliequivalents (meq) per 100 grams of soil to the CEC, while the organic fraction, which averages approximately 2097o carbon, contributes 40 meq (100 g)"'. These combine to give the average soil in the meadows a total CEC of 55 meq (1(X) g)'', which is well above that of most mineral soils. In general, poorly decomposed fibric organic matter contains rela- tively few phenolic hydroxyl and carboxyl groups, and thus contributes comparatively little CEC to soil horizons in which it occurs. On the other hand, well-humified sapric organic matter generally contains many such groups, and in many of the soils may be the main source of the CEC. Cation exchange capacities ranged widely among soils of the differ- 228 P. L. Gersper et al. o X E u O E C7> 16 12 0 4 - Na V\ A (A Wet Moist Meadow Basin Trough Rim Top Polygon FIGURE 7-7. Average quantities of exchange- able cations and cation exchange capacity in the upper 10 cm of the soils in 1972. ent microtopographic units. For example, average CEC within the upper 10 cm of soil in summer 1972 was approximately 50 meq (100 g)"' in wet meadows and 69 meq (100 g)"' in mesic meadows. In polygonized terrain the averages were 44, 70, 89 and 91 meq (100 g)"' in the troughs, rims, basins and tops of high-centered polygons, respectively. Variations in CEC among the microtopographic units were different when measured on the basis of volume rather than weight of soil, be- cause of variations in bulk density. For example, soils of polygon troughs had a CEC of approximately 27 meq (100 cm)"^ in the upper 10 cm compared to 25 meq (100 cm)"^ in the upper 10 cm of polygon basins (Figure 7-7), even though the CEC on a weight basis in the soil of the basins was more than double that of the troughs [89 versus 44 meq (100 g)"']. Thus, actual concentrations of nutrients in the soils of troughs were The Soils and Their Nutrients 229 higher than those of basins (Figure 7-7) because of the much higher bulk density of the trough soils. Soil pH values from the different microtopographic units generally range from 5.1 to 5.7; thus these soils are moderately to strongly acid by agricultural standards. They are, however, less acid than those of peat bogs, which have pH values between 3.0 and 4.0 (Moore and Bellamy 1974). The high concentrations of H* ions in the soil tend to favor their adsorption by the cation exchange complex, and decrease the adsorption of metallic cations. Soil acidity varies both spatially and temporally, is generally con- stant with depth, and shows some association between the more basic values and high plant production. Polygon troughs and the rims of low- centered polygons have relatively high soil pH values of 5.6 to 5.7, while the basins of low-centered polygons and the centers of high-centered polygons with peaty soils are the most acid sites, with soil pH values of 5.1 to 5.3. In studies of vehicle track disturbance the soil pH in the de- pressions where vegetative growth was abundant was 5.8, while in the control area it was 5.5 (Challinor and Gersper 1975). A drop in the mean soil pH in the mesic meadow from 5.4 to 5.1 between 1970 and 1971 was associated with a 20'^o drop in primary production (Dennis et al. 1978). Major Cations Over the range of microtopographic units from wet meadows to tops of high-centered polygons, the quantity of exchangeable cations per square meter in the upper 10 cm of the soil ranges from 3.7 to 7.9 g Na, 4.4 to 16.2 g K, 19.6 to 76.9 g Mg and 71.3 to 384.5 g Ca. The rims of low-centered polygons have the largest pools of all cations; however, potassium is equally abundant in the centers of high-centered polygons. Wet meadows are lowest in all cations except potassium, which is lowest in the basins of low-centered polygons. Mesic meadows are generally richer than wet meadows, and troughs are richer than basins in ex- changeable cations. Patterns of cation concentration in soil solution (Figure 7-8) differ sharply from those of exchangeable pools. Soil solutions used for the analysis of metallic cations were obtained using porous ceramic cups in situ and a mild suction of -0.75 bar. High cation concentrations in the soil solution of polygon basins and the tops of high-centered polygons occur with low plant production while low concentrations in the troughs occur with high plant production (Webber 1978). The properties of the soil solution fluctuate during the summer and range widely between years in response to thaw, precipitation, evapotranspiration, surface and sub- surface flow, nutrient uptake by roots, and microbial activity. 230 P. L. Gersper et al. TABLE 7-2 Mean (x) and Coefficient of Variation (CJ for Cations (meg m'^) in Soil Solution Extracted from the Upper 15 cm of Soil in Moist Meadow, 1970 (n = 60) and 1971 (n = 90) 1970 1971 c. c„ Cation X (%) X (%) Calcium 48.2 17.0 59.4 13.6 Magnesium 40.9 12.2 58.4 16.7 Potassium 2.7 66.1 3.7 30.6 Sodium 62.8 11.1 79.7 6.3 The concentrations of soluble cations change markedly between years as well as throughout the season. Sampling in a mesic meadow site in 1970 and 1971 revealed changes in yearly averages up to 40% (Table 7-2), Averages of every nutrient were higher in 1971 than in 1970. The summer of 1971 was warmer and wetter than 1970. This may have pro- duced an increase of mineral nutrients in solution due to increased min- erahzation of organic materials, or increased leaching of canopy and Utter Wet Moist Meadow Basin Trough Rim Polygon Top FIGURE 7-8. Average concentrations of cations in solution extracted from the upper 10 cm of soils in 1972. The Soils and Their Nutrients 231 as a result of more precipitation. Whenever soil solution was sampled im- mediately following precipitation, large increases in nutrient concentra- tions were observed, suggesting that leaching of aboveground plant ma- terials may be a major factor in nutrient transport. Of the major cations, only potassium occurs to a significant extent in a mineral form in soils. The clay mineral illite, which contains fixed potassium, is the dominant mineral in the clay fraction of the soils of the Barrow tundra (Douglas and Tedrow 1960). However, the bulk of the available potassium and almost all of the other metallic cations are bound on the exchange complex and are supplied from it to the soil solution. Nitrogen and Phosphorus The distribution of nitrogen and phosphorus is similar in that these elements are found mainly in the organic form in soil. The pools of nitro- gen and phosphorus of the moist meadows were calculated for the upper 10 cm of the soil (Table 7-3), since this portion is relatively homogeneous and includes more than 15% of the Hve root biomass (Dennis et al. 1978) and microbial biomass (Chapter 8). A total of 432 g N m"^ was found in TABLE 7-3 Pools of Nitrogen and Phosphorus (g m'^ 10 cm'^) in the Upper 10 cm of Soil in Moist Meadow Nitrogen Phosphorus Living 9 1.3 Belowground plant parts 7 0.6 Microbial organisms 0.70 Bacteria (20 gdw) 2.3 0.64 Fungi (5.5 gdw) o.r O.r Organic matter 420 15.5 Hydrolyzable N (6N HCl 15 hr) 336 Readily hydrolyzable N 1.4 (0.5N HCl 0.5 hr) Dissolved organic 0.2 0.0126 Inorganic matter 3.0 7.8. Resin-extractable P 0.0161 Dissolved inorganic 0.014 0.0006 nh: 0.013 NOi 0.0006 Total 432 24.6 'Nitrogen and phosphorus in fungi (Laursen 1975). 232 P. L. Gersper et al. TABLE 7-4 Exchangeable Ammonium Nitrogen of a Typical Per- gelic Cryohemist in Moist Meadow, 1973 Bulk Depth density (g cm') Exchangeable ammonium (cm) (meq 100 g"') (g m"' cm"') 0-5 0.191 1.006 0.269 5-10 0.656 0.350 0.321 10-15 1.043 0.424 0.620 15-20 0.479 0.480 0.322 20-25 0.600 0.753 0.634 Source: Flint and Gersper (1974). the upper 10-cm section, with more than 9597o bound in organic matter (FHnt and Gersper 1974). The organic nitrogen can be divided into hy- drolyzable and nonhydrolyzable fractions. The hydrolyzable fraction makes up 80% of the nitrogen in the soil organic matter, while the non- hydrolyzable fraction, which probably represents the most resistant core of the humus, makes up approximately 19.5%. Most of the remaining 0.5% is in the form of readily hydrolyzable nitrogen. This latter fraction is seasonally variable, indicating that it may be an integral part of the labile nitrogen in the system. The nitrogen content of the living soil microorganisms is uncertain since separation of the organisms from the soil material is difficult. Fungal biomass and nitrogen content of the fungi were both determined (Laursen 1975), but bacterial biomass may exceed fungal biomass by an order of magnitude (Chapter 8), and no measurements exist of the nitro- gen concentrations in the natural bacterial population. The inorganic nitrogen in the soil is almost entirely in the form of ammonium ions bound on the cation exchange complex, and in equi- librium with the ammonium and other cations in the soil solution. The vertical distribution of exchangeable nitrogen affects its availability to plants and soil organisms. Although the concentration of ammonium on a weight basis is highest in the surface 5 cm (Table 7-4), the amount in the 10-cm rooting zone is only a little more than 25% of the total ex- changeable pool in the active layer (Flint and Gersper 1974). Thus, a large fraction of the nitrogen present in exchangeable form is not physi- cally accessible to most of the plants or microorganisms. Soil solutions for nitrogen and phosphorus determinations were ob- tained from sample cores using pressure up to 9.4 bars (Barel and Bars- date 1978). The soil solution contains dissolved and colloidal organic The Soils and Their Nutrients 233 nitrogen, ammonium, and nitrate, in approximately 10:1.0:0.1 ratios. Most of the organic nitrogen in solution is readily decomposed, but plant uptake is from the inorganic nitrogen in the soil solution, and diffu- sion processes act primarily within this pool. The average concentrations of ammonium and nitrate in the soil solution in 1973 were 145 and 6 ppb, respectively (Bare! and Barsdate, unpubl.). The total amount of nitrogen in the soils of the drier microtopo- graphic units is commonly greater than 500 g m'^ (10 cm)'", slightly more than in the moist meadow soils, but the amounts of exchangeable nitro- gen are similar. The average nitrate concentration in the soil solution of the polygon rim was 5.9 ppm NO3-N in 1973, almost three orders of mag- nitude higher than the nitrate concentration in the wet meadow. The am- monium concentration on the rim was 750 ppb, also higher than in the meadow. The ratios of ammonium to nitrate in the soil solution change from 10: 1 in the moist meadow to 0. 1 : 1 on the rims of low-centered poly- gons. Nitrate is also found in greater concentrations than ammonium in the centers of high-centered polygons with mineral soil, but ratios drop below 1 in the other, slightly moister high-centered polygons with peaty soil and in mesic meadows. The total soil phosphorus in the upper 10 cm of the moist meadows is approximately 25 g m'^ of which two-thirds is in organic form (Table 7-3). Dissolved organic phosphorus is not believed to be available to plant roots but it is apparently susceptible, like dissolved organic nitro- gen, to rapid hydrolysis. The ratio of dissolved to total organic phos- phorus is very low, 0.0008:1 (Barel and Barsdate 1978), even when com- pared to that for organic nitrogen (0.002:1). The organic phosphorus contributed by soil microorganisms has not been determined, but calcu- lations based on decomposer biomass and species composition indicate that the standing crop of decomposers ties up a far larger fraction of soil phosphorus than nitrogen, 3% vs 0.4%. Thus, fluctuations in microor- ganism populations may have a significant effect on the overall distribu- tion of phosphorus. The fraction of the inorganic phosphorus that is in equilibrium with the soil solution appears very small when measured by extraction onto an anion-exchange resin (Barel and Barsdate 1978), and the concentration of inorganic phosphorus in the soil solution is correspondingly low, averag- ing 10 ppb in 1973. However, chemical fractionation of the inorganic phosphorus from the moist meadow soils indicated that a large fraction is extractable under reducing conditions (Chang and Jackson 1957). This fraction may contribute considerably more to the exchangeable and dis- solved pools of phosphorus under anaerobic, reducing conditions such as exist in the soils of wet meadows and polygon troughs than is apparent in laboratory analyses performed under aerobic conditions (Khalid et al. 1977). Most of the available phosphorus is bound to iron or aluminum ions (Prentki 1976). 234 P. L. Gersper et al. On the rims of low-centered polygons and the tops of high-centered polygons with mineral soil, total phosphorus in the surface horizon is somewhat more abundant than in the wet meadow soils, but a greater fraction of the total phosphorus is in an organic form, and inorganic phosphorus is less abundant. The inorganic phosphorus of the drier site is mainly NH4-F soluble, and considered to be readily available to plants (Barel and Barsdate 1978). Increased availability of phosphorus is also indicated by a higher average value of resin-extractable phosphorus (22.6 mg m"^) in the drier soils. However, the amount of inorganic phosphorus in solution is lower than in the wet meadow. On the tops of high-centered polygons the phosphate in the soil solution drops from 8 to 4 ppb at the boundary between organic and mineral soil, 4 cm below the soil surface (Barel and Barsdate 1978). INPUTS AND OUTFLOWS OF NITROGEN AND PHOSPHORUS The major input of nitrogen to the soils of the coastal tundra at Bar- row is through the fixation of atmospheric nitrogen by blue-green algae, either alone or in symbiotic relationships. The inorganic ions in precipi- tation are the major sources of phosphorus and also add to the pool of inorganic nitrogen. Inorganic forms of nitrogen move in the soil by dif- fusion, but phosphorus ions are relatively immobile. Losses of nitrogen and phosphorus, in both inorganic and dissolved or suspended organic forms, occur through surface and subsurface flow. Nitrogen can also be lost through reduction to gaseous forms, nitrogen oxides and nitrogen gas. Even though a large portion of the tundra surface is covered with lakes and small ponds, movement of nutrients between the terrestrial and aquatic subsystems seems to be restricted to the period of snowmelt, when there is a small net loss of nitrogen and phosphorus to the ponds (Prentki et al. 1980). Nitrogen Fixation Fixation supplies the bulk of the nitrogen input to the terrestrial sys- tem, although amounts entering by this pathway vary markedly between microtopographic units. Measured rates increase from 8 to 180 mg N m"^ yr"' along a moisture gradient from dry polygon rims to wet meadows. Blue-green algae are the most important agents of nitrogen fixation. These algae, primarily Nostoc commune, occur as free-living or moss- associated filaments, or symbiotically in lichens of several genera. Al- though Peltigera aphthosa is the most abundant nitrogen-fixing lichen. The Soils and Their Nutrients 235 Peltigera canina, Lobaria linita. Nephroma spp., Solorina spp., Stereo- caulon spp. and other Peltigera spp. also occur in significant amounts. With ambient temperatures of about 15°C nitrogenase activity ranges from 4.5 fig N gdw"' hr"' in Stereocaulon tomentosum to 41 .5 /ig N gdw"' hr"' in Nostoc commune. Nitrogen fixation per unit of biomass in the Peltigera species is high (8.8 to 25.8 /^g N gdw"' hr"' at 15°C), considering the large proportion of its biomass contributed by fungus and thus not directly involved in nitrogen fixation. Heterotrophic bacteria also may contribute to nitrogen fixation (Chapter 8). Azotobacter was isolated from a mesic meadow, but the numbers were low, 10^ to 10^ cells (gdw soil)"', and nitrogen fixation within the soil was consistently less than 1 \xg N m"^ hr"'. If heterotrophic bacteria are indeed active fixers of nitrogen in the soils, their activity is very low compared with the free-living and symbiotic blue-green algae of lichens. No significant nitrogen fixation was found to be associated with any higher plants (Alexander and Schell 1973). Alpine tundra soils in central Alaska, however, have a substantial input of nitrogen from vascular spe- cies, including Dryas spp., Lupinus, Astragalus and Oxytropis spp., which are abundant in the alpine sites and in the Prudhoe Bay region but absent or rare in the coastal tundra at Barrow. Similar constellations of organisms have been found dominant in the nitrogen-fixation regimes of other tundra sites (Alexander 1974, 1975, Jordan et al. 1978). Nostoc commune is a cosmopolitan species that is important in nitrogen fixation in a variety of natural ecosystems (Fogg et al. 1973). In particular, the Nostoc-moss association, which has drawn much attention in circumpolar studies, also appears to be an im- portant feature of the grassland ecosystem (Vlassak et al. 1973). Biomass of Nitrogen-Fixing Organisms Nitrogen fixation rates for any location on the tundra depend pri- marily on the distribution and biomass of the nitrogen-fixing organisms and secondarily on the various abiotic variables that influence nitrogen- ase activity within organisms. The distribution of nitrogen-fixing organ- isms is correlated with moisture regime and vegetation. Nostoc commune occurs in wet environments, and is especially abundant in wet, low-lying meadows, where it may occur as extensive mats floating over the moss layer. One low-lying area contained 19.5 g Nostoc m'\ 10% of the stand- ing crop (Williams et al. 1978). Additionally, Nostoc forms an epiphytic or intercellular association with various genera of mosses. The nitrogen-fixing lichens, primarily Peltigera aphthosa, tend to occur at intermediate moisture levels such as the slopes between troughs 236 P. L. Gersper et al. and rims, although P. aphthosa, P. canina and Lobaria linita are com- mon in wet meadows, as well as in depressions between clumps of Erio- phorum vaginatum in better-drained meadows (Williams et al. 1978). In the microtopographic units that are more favorable for lichens, such as the rims of low-centered polygons, a total lichen biomass as high as 180 g m"^ has been observed; however, only about 2 g m"^ of this is capable of nitrogen fixation. Thus the biomass of nitrogen-fixing organisms in- creases from dry to wet areas, with the major fraction made up of free- living or moss-associated Nostoc, which is confined to wet areas, and the remainder composed of nitrogen-fixing lichens, which are relatively more abundant in the mesic areas. Environmental Controls on Nitrogen Fixation In laboratory studies the principal environmental factors modifying rates of nitrogenase activity in nitrogen-fixing organisms are tempera- ture, moisture, light and oxygen tension. Response of Nostoc and Pelti- gera to climatic factors, and to some inorganic nutrients, is described by Alexander et al. (1978). Diurnal temperature fluctuations of 10 °C for the rim of a low-centered polygon and 15°C for a polygon trough were re- corded in July 1972. Thus rather high temperatures can be attained in the immediate vicinity of the maximum algal biomass. Fluctuations in both light and temperature appeared to exert a strong influence on field rates of nitrogen fixation (Alexander et al. 1974). The most critical environmental factor in determining the rate of ni- trogen fixation is moisture. The response of Peltigera to moisture (Figure 7-9) shows that saturation of nitrogenase activity does not occur until the moisture content exceeds 250% of dry weight (Alexander et al. 1978), a response similar to that shown by other nitrogen-fixing lichens (Kallio 1973). Although no similar data exist for Nostoc, it shows no activity at all when dry, but rapidly resumes activity when moistened above 100% dry weight. The nitrogen-fixing organisms appear to be well adapted to handle periodic desiccation, and are able to make effective use of mois- ture whenever it is available. On a season-long basis, highest inputs from nitrogen fixation occur on wet, mossy areas. In drier areas, seasonal nitrogenase activity is limit- ed by available moisture (Alexander et al. 1974), and overall rates are somewhat lower during summers with low rainfall. Extremely wet areas devoid of moss cover also have very low rates of nitrogen fixation. In dry summers, such as 1972, there was a decline in fixation in moderately moist areas as the season progressed and soil moisture declined. The total seasonal input from nitrogen fixation, integrated over an area compris- ing a variety of microtopographic units, was lower in 1972 (85 mg N m"^ The Soils and Their Nutrients 237 800 S. 600- u> J> o E c « c — UJ 200 400 Moisture, % dw 40 60 Percent Oxygen 600 400 1 1 1 1 1 1 1 B - 300 -\ - 200 ^- \ - 100 - \. 0 1 1 o lM 1 -1 lo- - FIGURE 7-9. Response of nitrogen fixation rates of Peltigera aphthosa to mois- ture (A) and oxygen concen- tration (B), and of Nostoc commune to oxygen concen- tration (C). (After Alexan- der et al. 1974 and Alexan- der 1978.) than in the wet summer of 1973 (1 19 mg N m"'), the difference being due primarily to a higher rate of fixation in the relatively dry areas during the wetter year. These differences between years are considerably less than the differences between specific microtopographic units. For example, in the wetter summer seasonal input on the dry rim of a low-centered poly- gon was only 6.7 mg N m"% whereas in a nearby low, mossy area it was 150.4 mg N m'^ The response of Nostoc commune to oxygen tension is of special ecological interest, since the greatest nitrogenase activity of this organism occurs in wet, mossy areas, where these algae exist in extremely close as- sociation with mosses. In the water associated with the mosses, oxygen saturation may range from 5 to 24% over 24 hours (Alexander et al. 1974). 238 P. L. Gersper et al. The strong inverse relationship between oxygen tension and nitrogenase activity (Figure 7-9) indicates the variation could be a very significant factor influencing rates of nitrogen fixation by algae associated with moss. Similar relationships have been described for a mire site in Sweden (Granhall and Selander 1973). The relationship between nitrogen fixa- tion in lichens and oxygen is complex, particularly because there appears to be a strong interaction between light and oxygen requirements and a conflicting influence between the inhibitory effects of oxygen on the ni- trogenase enzyme and the need for photosynthetically produced sources of energy. A simple model, nfixr, was developed that integrates the available laboratory measures and permits evaluation of their general applicability against field observations (Bunnell and Alexander, unpubl.). The model assumes that the influences of temperature, moisture and oxygen interact in a multiplicative fashion. Thus fixation is reduced as any single envir- onmental control departs from the optimal range, even though other conditions may not be Umiting. Seasonal courses of nitrogen fixation for specific genera can be predicted from measured environmental variables and compared with observed fixation rates (Figure 7-10). Although actual magnitudes differ, the observed seasonal courses of nitrogen fixation in polygonal terrain at both the Biome research area and in a birch site at Kevo, Finland, are similar to those predicted by the model. Apparently, the measured relationships are broadly applicable to lichens and algae inhabiting a variety of sites. The inaccurate prediction of the magnitude of rates of fixation apparently is largely due to the dif- ficulties in estimating biomass of the fixing organisms, particularly algae. Blue-green algae are relatively more important at Barrow than at Kevo, and the predictions for Barrow are therefore less accurate. In light of the recent observation that non-heterocystous blue-green algae also contain the enzyme nitrogenase and are capable of nitrogen fixation under conditions of low oxygen (Kenyon et al. 1972, Stewart 1973), special interest centers on the ecology of these moss-associated algae. Present findings suggest that the majority of blue-green algal forms found in the moss layer may contribute to nitrogen fixation, and that estimates of nitrogen-fixing biomass based only on heterocystous algae may be greatly in error both in the wet, mossy layer and in soils. There is no marked adaptation by the major nitrogen fixer, blue- green algae, to the arctic environment. The predominant nitrogen-fixing form, Nostoc commune, is found in the Antarctic and in all circumpolar tundra regions. Its temperature optimum is not greatly different from temperature optima for blue-green algae from temperate and tropical regions. Arctic lichens, however, appear to be rather well adapted. Nitro- genase activity of lichens recovers after freezing, with the rate of recovery depending on the length of time the lichens were kept frozen and the tem- The Soils and Their Nutrients 239 1500 o c 1000 4. X o Barrow, Alaska — Predicted Values ■— Field Measurement FIGURE 7-10. Comparison of simulated and measured rates of nitrogen fixation at Barrow and at Kevo, Finland. (Bunnell and Alexander, unpubl.) perature at which recovery is taking place (Kallio and Alexander, un- publ., Kallio 1973). Rates of nitrogen fixation per unit of ground surface measured for arctic lichens are somewhat higher than those measured in other Biomes (Stewart 1969). Inputs of Nitrogen and Phosphorus by Precipitation Snowfall includes approximately 30% of the total nitrogen supplied by precipitation. Ammonium is the predominant form of nitrogen found in snowfall (Dugdale and Toetz 1961), aUhough organic nitrogen has not been measured. The fraction of the inorganic nitrogen present as nitrate 240 P. L. Gersper et al. in snowfall declined from almost 30% to less than 10% between early September and late October of 1960 (Dugdale and Toetz 1961). Nitrite concentrations in fresh snow are exfemely low, with an upper limit of about 1 /ig N liter"'. Comparisons of these values with nitrogen distribu- tion in snow columns in May indicate that ammonium may be converted to nitrate in the snowpack. The concentration of inorganic nitrogen in both samples was similar, approximately 80 ^g liter"', but the concentra- tion of nitrate in the spring sample was higher by 15 /:4g N liter"'. Concen- trations of all three inorganic forms of nitrogen are higher in rain than in snow. The total concentration of nitrogen in summer precipitation is 340 /ig hter"', of which ammonium contributes 15% and organic nitrogen less than 20 u \^ \ ^ \ V \ ^ \ X ^v N. \ V \ >. \ X X >» \ V \_ ■x — P > / • / •' / • / • / ^ / • / • / • / • / • / y / • / • / ^ / • /• — y> A f */ • / - ^d 1 1 1 It 20 40 20 40 20 40 60 80 100 120 NO3, /i.g liter' FIGURE 7-12. Profiles of NH\ and NOl concentrations in the soil solu- tion from the moist meadow. ported from depths of 20 cm or more to the surface of a wet meadow soil during the growing season. In 1973 detectable transport began on about 20 July and continued into September. During this period the maximum net rate of flux from the well-decomposed organic layer into the mineral layer above it was about 0.049 g m"^ day'. At this rate at least 2 g N m"^ could be transferred from the subsoil to the rooting zone in a period of about 60 days. The mechanism of transport has not yet been verified. However, concentration profiles of exchangeable ammonium in the soil through the summer period indicate that a diffusion mechanism is oper- ating along the soil exchange complex. This may be the primary mechan- ism of nitrogen transport, far exceeding the amounts that diffuse through the soil solution. Results also indicate that the amount of nitro- gen transported by diffusion is strongly affected by soil temperature, thaw depth, and length of the thaw season. No experimental studies have been conducted on phosphate diffu- sion in the soils of the coastal tundra at Barrow, but diffusion rates can be assumed to be generally low (Olsen et al. 1962) and added phosphorus fertilizer is strikingly immobile. Ten years after the last treatment, plots fertilized with phosphorus by Schultz (1964) still showed levels of labile and dissolved organic phosphorus that were 50 times as high as those of 252 P. L. Gersper et al. adjacent control plots, suggesting lack of movement of phosphorus (Barel and Barsdate 1978). Other Effects Considerable quantities of both nitrogen and phosphorus can be transferred directly to available pools in the soil during a lemming high. During these population peaks lemmings consume up to 40 g m'^ yr' of graminoid plant material, nearly 50% of the annual aboveground pro- duction, and most of the minerals in this are excreted. However, this ef- fect on available pools of nitrogen and phosphorus may be relatively in- significant during population lows, when consumption may fall below 1 g m"^ yr"' (Chapter 10). Nitrogen is mainly excreted in the urine and is immediately available to plants and microorganisms. Phosphorus is dis- tributed between urine and feces (Barkley 1976). Leaching experiments using an analogue of the surface runoff showed over 90<^o removal of phosphorus from feces in 24 hours (Chapin et al. 1978). During a high year, lemming feces would release about 90 mg P m"^ The freeze-thaw effect, described by Saebcf (1968) for Sphagnum peat, is another way nutrients may be transferred from unavailable to available pools. After freezing and thawing, peat samples showed con- centrations of dissolved and dilute acid-soluble phosphorus several times higher than did the control samples. The solution concentration returned to control values after remaining thawed for 48 hours, but values for acid-soluble phosphorus remained somewhat above controls for the same time period (Saebd 1968). Patterns of dissolved and resin- exchangeable inorganic phosphorus in the soils of the coastal tundra at Barrow indicate that the same effect is occurring (Barel and Barsdate 1978). A similar effect was observed in solution concentrations of am- monium and nitrate (Barel and Barsdate, unpubl.) and in soluble carbo- hydrates in soils of other areas (Gupta 1967). These similarities, and the lack of any effect on calcium levels, indicate that the freeze-thaw mech- anism may involve a physical disruption of the organic matrix. The me- chanics of the effect, and its magnitude, are still unclear. The mineral fraction of the soil contains a significant fraction of the total phosphorus pool in non-exchangeable form (Chapin et al. 1978). Chemical transformation of the mineral matrix in which the phosphorus is bound would allow the transfer of some phosphorus to the exchange- able pool. Although weathering rates are low in arctic conditions (Doug- las and Tedrow 1960), this source of inorganic phosphorus may not be negligible under the low-phosphorus regime of the wet meadow soils. The Soils and Their Nutrients 253 SUMMARY Organic matter, generally in a partially decomposed (hemic) state, dominates the soil profiles of the coastal tundra at Barrow, and consti- tutes the major pool of fixed carbon in the ecosystem. The bulk density of the highly organic soil is low, but increases with advancing decomposi- tion. The soils remain very moist throughout most summers, have high cation exchange capacities, and are moderately acid, with the lower pH levels correlated with lower primary productivity. Almost all the nitrogen in the soil is present in organic form, and a large fraction of this is associated with poorly decomposed material. A small and variable amount of labile organic N is also present. In wet meadows inorganic nitrogen is mainly in the form of ammonium, and more than half is found below the primary rooting zone. In the wet areas, nitrate concentrations in the soil solution are very low, but in the driest units, nitrate concentrations exceed those of ammonium. Most of the soil phosphorus is also in organic forms, and the concentrations of inorganic phosphorus in the soil solution are extremely low. Nitrogen fixation by blue-green algae is the major input mechanism for nitrogen. These algae may be free-living forms, but in many cases are associated with mosses or occur symbiotically in lichens. The predomi- nant algal and lichen forms involved in nitrogen fixation are Nostoc commune and Peltigera aphthosa, respectively, although several other li- chen species are also active. The biomass of nitrogen-fixing organisms is highest in wet, mossy areas, and is extremely low in dry areas. In mesic sites, moisture is usually the major factor controlling the input of nitro- gen, but oxygen concentration and temperature are also important. The low oxygen concentrations that occur in wet, mossy areas may enhance the rates of nitrogen fixation. A simulation model indicates that a simple multiplicative interaction between these factors may be involved, and that the control mechanisms for nitrogen fixation may be similar at other tundra sites. Inorganic nitrogen and phosphorus enter the system through precip- itation. The amounts added are small in comparison to the total pools of these elements, but substantial with respect to available inorganic pools. The major losses of nitrogen and phosphorus occur in runoff during snowmelt and are mainly of organic forms. The combination of precipi- tation and runoff yields a net loss of nitrogen and phosphorus. Some ni- trogen is also lost by denitrification, but the rate is low compared to the potential for nitrate reduction that exists in the wetter microtopographic units. Nitrogen fixation is sufficient to lead to a net accumulation of soil nitrogen. The ratios of carbon to nitrogen and organic phosphorus are suffi- 254 P, L. Gersper et al. ciently high to suggest that weak nitrogen immobiHzation and strong phosphorus immobiHzation should be associated with decomposition. However, nitrogen mineraUzation has been shown to occur under cold, anaerobic conditions, perhaps because of low tissue nitrogen concentra- tions and low efficiency in the decomposer population. Phosphorus min- eralization may respond to these same factors and be further facilitated by selective degradation of phosphorus-rich substrates. Nitrifying bacteria are not common in the soil, and their activities are inhibited by low temperatures. Changes in the amount of nitrate present indicate low rates of nitrification in the wet meadows, and higher rates in drier microtopographic units. Several internal pathways may aid in replenishing inorganic nutri- ents in the rooting zone. Studies with "N indicate a substantial flux of N from the subsoil to the surface. Freezing and rethawing of the soil liber- ate some available nitrogen and phosphorus. Weathering of minerals containing non-exchangeable phosphorus may also occur. In a high lem- ming year, lemming excreta contribute substantial amounts of available nitrogen and phosphorus. 8 The Microflora: Composition 9 Biomass, and Environmental Relations F. L. Bunnell, O. K. Miller, P. W. Flanagan, and R. E. Benoit INTRODUCTION Tundra microflora show the same low species diversity evident among the plants and herbivores. Two groups with broadly different dy- namics dominate the microflora: rapidly changing bacterial populations that typically degrade smaller molecular compounds and more slowly changing fungal populations that are more capable of degrading larger molecules. The accumulation of organic matter in tundra systems reflects historical imbalances between production and decomposition. This chap- ter presents an overview of the composition, abundance and diversity of decomposer populations in the coastal tundra at Barrow as related to soil characteristics previously described (Chapter 7). Activities of the micro- flora are treated in Chapter 9. The overview presented is unavoidably influenced by the isolation methods employed (Table 8-1). The usual cautions appropriate to inter- pretations of microfloral isolations apply. Isolation techniques select only a portion of the total viable flora. Microorganisms that grow readily on artificial media and appear to be dominant may not be dominant in their natural habitat. Similarly, fungi that fruit prolifically in the field may not contribute the greatest biomass of active vegetative cells. There- fore, the orientation of this discussion is broadly functional rather than strictly taxonomic, particularly when bacterial groups are being dis- cussed. The bacteria are examined as they contribute to transformations of carbon, nitrogen and sulfur, whereas fungi are treated in more taxo- nomic detail. 255 -5: o s: I o O « to s: <^ o ?> <4j o ?3 CO QQ 90 QQ < 00 3 o 1- 1— (N OO O vO 00 oo >o r~- — r- 00 <£ d> -^ r»1 O O Tt -■ '^ -~ C- 2- Tt — oo (N >o m d o d (Nl vo \0 Tt rsi t^ ■a 00 E o tS o o o — — On oo O ,,^ lo O oo rsi — . (N — ; ^^ d> /-> O — r- d d d '^ o o S NO rrj CnI NO CD '5b NO o r~- d • . ^ o o e a rs| r- NO o "^ "H T3 r^ ■ — o . X) CO — >/-, o o X X V~\ NO d •~~. — On X X d -* Q> ^5 •5 CO il i: -0 ? ^ i: « 3 S E > ■z. ^ c/i a> w j=. yi tft *-• (U « >. c/5 i2 S X) .— c '^ 0 5^ E eS . ^ 0 C fs^ cO u- j= E c^ E — 0 TJ u. 00 •- i! .OJ c XI cO a> ~" 00 E ■0 T3 03 g S C/5 1) X3 -' radius r sonal av E 3 0 > E 3 c Xi NO NO d S3 03 1^ = l> C/3 XI ece ss(s c 0 0 ki D. J= ca -0 c c E C/5 0 E C 'S' iZ Xi s u. D. T3 > t/5 c •0 c (u 'r; 3 ■n ca d imat rela C E (U u« 3 03 1) XJ .5 •5 X 1/5 i_ E E ON Kwas e heses a 3 c a> X3 03 C/5 3 03 II V-i E £i E X) > k. d — a 0 0 ■5 ^^ "" v-\ = 5 'E D. ^-^ II II ri I.E 0 m 0 E ■5 t/i t/5 0 E T3 C 03 t/5 c 3 t/) c 3 S ■^ s ■^ II Organisrr Numbers E 3 E 0 D. 0 a E 0 E 0 II ^ T3 # ^ u. 3 LO •0 03 D a E 00 00 E 0 0 '03 c 0 k« 1^ m "■ 0 00 1) 'E 0 1) c 0 II Ol II # II 0 ^ •a 03 c/: 0 T3 0. -a t/i lU k- 00 > (J > II E 0 Q/ 2 >N •0 0 E c 0 E (U k« 0 X! 0 X) 00 e' 3. _L T3 00 C/J 0 >> 03 03 0 ] ts for the ercentage _c — ' c C > 00 1) > 'ob E 00 II e ^*< c a 'S 0 "E s 0 e' E ,-, 3. e' asureme y e, and ■a E 3. a. 0 XI CO c > '5) II k> d II Soils me e, densit 3 0 (/5 k- e T3 00 1^ 1^ Note: : volume E 0 CJ s E 3 E 3 II II II II ? ^ U Z Z :i. :i. i:^ l:^ :i^ f-t '» ■^ •D '^ O "Z o E 0 4000- 2000- i^ 6000 (•)Stonding Dead (o)Litter Jun Jul Aug [Sep Jun Jul Aug Sep Troughs Rims 4000- 2000 Jun I Jul Aug Sep Jun [ Jul [ Aug [Sep FIGURE 8-6. The seasonal progression of fungal density in standing dead vegetation and plant litter, 1973. (Flanagan, unpubl.) 1973 was unusually wet (Figure 2-2). High amounts of moisture in the more concave microtopographic units, such as basins of low-centered polygons and polygon troughs, apparently depressed fungal growth. The more productive troughs experience highest moisture levels early and late in the season, producing a mid-season peak in biomass (Figure 8-6). Rims of low-centered polygons are elevated, exposed, and subject to periodic wetting by precipitation and fog and drying by wind, and conse- quently show the most erratic seasonal trend in fungal density. Fungal density in the standing dead vegetation of wet meadows gradually de- clines through the season as drying progresses. Although the seasonal pattern was the same, during the moister summer of 1973 average fungal density within the standing dead vegetation of wet meadows was 12% higher than in 1972. The Microflora 275 Meadows Basins o £ o CD I - Depth (0)6 fo7cm AAa^ Troughs Jun Jul I Aug [ Sep Jun[ Jul | Aug [ Sep Jun I Jul I Aug I Sep Rims 3- h N ?- \ / i 'e \ / \ o> \> \ « 1 - N^ in 0 E 0 m 0-1 1 1 1 Tops Aug Sep Jun FIGURE 8-7. The seasonal progression of fungal biomass at two soil depths, 1973. The values are the means of three to five replicate plots. (After Laursen 1975.) Seasonal courses of fungi in the litter layer show somewhat more consistency in pattern than is apparent in standing dead vegetation (Fig- ure 8-6), but the variation among microtopographic units remains large. Generally, the pattern is more akin to that observed for fungi in the soil, with a peak in biomass shortly after snowmelt followed by a gradual decline (cf. Figure 8-7). The peak in fungal biomass is most evident in wet meadows and rims of low-centered polygons, which accumulate less moisture than do the concave microtopographic units. In the basins of low-centered polygons, fungal biomass in litter shows a gradual decline 276 F. L. Bunnell et al. during the growing season, and never does attain the levels measured in other habitats. In wet meadows, seasonal mean estimates of fungal densities in litter (x ±SE) were 2442 ±654 m mycelia (gdw litter)"' in 1972, and 2127 ±664 m mycelia (gdw litter)"' in 1973. The microclimatic regime of the litter layer is typically moister than that for standing dead and the increased precipitation of 1973 apparently was insufficient to alter measures of average litter biomass. Fungi in the Soil The seasonal patterns of fungal biomass in the surface layer of soil (1 to 2 cm) show a peak immediately after snowmelt followed by a se- cond peak in early August, which may be as pronounced as the first, moderate, or weak (Figure 8-7) (Laursen 1975). The magnitude of the first peak for troughs and basins may be underestimated because the soils were frozen at the time of the first sample and the highest levels may have been reached before the next measurement. More pronounced early peaks in fungal biomass were observed in the surface soil of polygon troughs in 1972 and 1974 (Laursen and Miller 1977). Except for the se- cond, variably expressed increase, a general decline in biomass similar to that observed among litter fungi is evident over the season (Figures 8-6 and 8-7). The pattern of a moderate peak followed by a gradual decline that was observed in the wet meadow is also evident in data from Norwegian sites (Hanssen and Goks(^yr 1975). The rapid early growth of fungal biomass in the surface layer of soil, as well as in the litter, may be a response to the release of nutrients during snowmelt. The second peak occurs just before the basidiomycetes fruit and is often more pronounced in the more productive microtopographic units (Figure 8-7). The same factors that restrict biomass of soil fungi in the polygon troughs and basins of low-centered polygons may also restrict the ability of these fungi to respond quickly to favorable conditions. Thus the seasonal courses of fungal biomass are not only lower, but less variable in troughs and basins than in the other microtopographic units. Seasonal patterns at a depth of 6 to 7 cm do not mimic patterns at 1 to 2 cm depth (Figure 8-7). The only apparent consistency in the seasonal pattern among microtopographic units in the deeper soil is the peak in fungal biomass that occurs in all units around 22 August. The decline in fungal biomass in the surface soil that begins in late August is probably associated with freezing or near-freezing temperatures in that stratum. Fungal biomass in the deeper soil typically equals or exceeds the biomass in the surface soil by mid-September (Laursen and Miller 1977). The taxonomic structure of both soil and litter fungi appears to shift The Microflora 277 during the season and provides further circumstantial evidence concern- ing the nature of the rapid early-season growth. In both litter and surface soil, some members of the Mucorales and common molds are more evi- dent early in the season and become less evident as the season progresses. These species are cold-tolerant (Latter and Heal 1971, Flanagan and Veum 1974), fast-growing and can be termed "soft" decomposers since they are better able to exploit readily leachable compounds than cellulose or lignin (Chapter 9). In deeper soil layers the fungal population includes "hard" decom- posers, fungi with a relatively greater capacity to utilize cellulose and lig- nin than have the mucors and common molds. Many of these are basidi- omycetes and ascomycetes, although zygomycetes are present at inter- mediate depths. The "hard" decomposers generally fluctuate less rapidly than the "soft" decomposers of the Utter and surface soil layers, and tend to increase slowly in biomass over the season as soil temperatures in- crease and buried organic soils become drier (Laursen and Miller 1977), A third broad component of the soil fungi are the mycorrhizal form- ers. Profiles of these fungi with depth approximate the profiles of roots of dicotyledonous plants, particularly Salix, Cassiope and Vaccinium species. On the rims of low-centered polygons, fungal biomass increases at depths of 2 to 4 cm and the average proportion of basidiomycete cells increases from well below 20% in the surface layers to 29% at 1 to 2 cm depth and 42% at 2 to 3 cm depth, declines to 37% at 3 to 4 cm, and sub- sequently drops sharply below 4 cm. The temporal pattern of mycorrhi- zal formers is marked by rather stable biomass levels which increase steadily from early season levels until fruiting, during July and early August (Miller and Laursen 1978). In general, yeasts are most abundant in early, frozen samples. Num- bers of yeasts decline as the soil thaws and bacterial biomass increases early in the season. The seasonal decrease in abundance is greater in soils with deeper thaw — about ten times greater from 26 June to 5 August in 1973. Yeast numbers thus appear negatively correlated with bacterial numbers, but the data are sparse. Differences in average population levels between years suggest that the differential response of yeast and bacteria is associated with amounts of moisture. In the moist summer of 1973, yeast counts in the upper 2 cm of wet meadow soil were less than 20% of the levels observed in 1971. At depths of 5 cm, counts in 1973 were at least 100-fold lower than in 1971, less than 10^ (gdw soil)"'. Bacteria In the surface soil, bacteria appear to follow the same seasonal pat- tern as fungi and may be responding similarly to changes in substrate. 278 F. L. Bunnell et al. Both groups have early- and late-season peaks. While the peak early in the summer may be associated with the release of organic material from both aboveground and belowground material, the second peak is more probably associated with the senescence and death of roots. Because the second peak appears to coincide with the first frost, this increase may also be associated with subsequent leaching of aboveground vegetation. Further evidence for the concept of substrate abundance influencing bacterial abundance is apparent in the differences with depth. Before the Recovery of bacterial numbers in late August or early September, plate counts from meadow soils show a gradual decline from 10' to lO*" in the upper 2 cm and a more marked decline from 10* to 10'' at depths of 2 to 7 cm (Benoit, unpubl.). Direct counts from the surface soil are also depressed in mid-season, but show no mid-season decrease in numbers in the deeper soil. The lower soil layers depend more upon root growth and exudation for replenishment of potential substrate, and less upon leaching from aboveground organic matter. Root activity at mid-season is apparently sufficient to maintain the bacterial population in the deeper soil. Although bacterial biomass and ratios of fungi to bacteria differ among microtopographic units (Table 8-1), seasonal courses of bacterial plate counts do not. Composition of the biomass does show some taxo- nomic shift and chromogenic bacteria appear relatively more abundant in late season. The consequences of such a shift to decomposition are unknown. Estimates of Minimal Production The estimates of production discussed below consist simply of the sum of the positive changes in measured biomass sensu Ivanov (1955) and are thus minimal and frequently confused by the variability of the microflora. Although biomass varied considerably among sample points on a specific microtopographic unit, both fungal productivity and fungal den- sity showed broad differences among these units (Figure 8-8). Fungal density was lower in the wetter summer of 1973. The difference in density between years was most consistent in the wetter microtopographic units, suggesting that the decrease was due to higher moisture levels. Although minimal fungal productivity is broadly correlated with average mycelial denshy, the nature of the relationship apparently shifts between years. Despite lower mean density, the total fungal growth in 1973 was general- ly higher than in 1972 for drier microtopographic units, in particular the rims of low-centered polygons, and for basins of low-centered polygons. Much of this growth occurred in late July to early August (Figure 8-7) The Microflora 279 Meadows Basins Troughs Rims Tops S ° 2000 o ° 6. - 1000 _.ll]l i 1 II ni 10000 — in 5000- o lDL J ..MiM. ^ I ol Jo Q 1972 ■ 1973 FIGURE 8-8. The seasonal average density and estimates of minimal productivity (sum of positive changes in density) of soil fungi at the I- to 2-cm depth in 1972 and 1973. The pairs of bars represent replicates from similar microtopographic units. (Laursen and Miller, unpubl.) TABLE 8-3 Mean Biomass and Minimal Estimates of Productivity (Sum of Positive Changes in Biomass) of Soil Fungi in Different Micro- topographic Units, 1973 Depth (cm) Meadows Basins Troughs Rims Tops Biomass (g m-^) 1-2 0.86 0.75 0.38 1.87 1.66 6-7 0.29 0.25 0.46 Productivity (g m-' yr') 1.47 0.70 1-2 1.32 1.47 1.36 5.59 3.95 6-7 0.96 0.80 2.84 5.33 2.05 280 F. L. Bunnell et al. Meadows Basins Troughs Rims 10x10°- >, o ^ in ^ »5xlO^- d 1 t^ J 2x10 xlO' ■3 o 1 o DQ o> 2x10^- □ Density I iBiomass -2x10 FIGURE 8-9. 77?^ seasonal average density and bio- mass and estimates of minimal productivity (sum of positive changes in counts) of plateable bacteria in the I- to 2-cm soil depth, 1973. (Benoit, unpubl.) and appears associated with fruiting of the basidiomycetes. A comparison of estimates of minimal fungal productivity with mean biomass estimates (Table 8-3) indicates that minimal productivity ranges from about 2 to 3.6 times the average biomass in the surface soil, giving broad turnover rates of about 2 to 3.6 times per season, about 2 times the rates estimated by Hanssen and Goksdyr (1975) for Norwegian sites. Turnover rates in the deeper soil are generally about 20 to 70% greater. Although fungal biomass is low in the deeper soil layers, the ex- isting biomass appears more productive than near the soil surface. This latter observation may be a result of grazing microbivores consuming more of the fungal growth near the surface (Chapter 11), a phenomenon that may confuse comparisons between years as well. Estimated bacterial production is only broadly correlated with mea- sured abundance among microtopographic units (Figure 8-9). Although basins of low-centered polygons show low counts of plateable bacteria in the upper soil, the population is relatively more productive than the more abundant populations in wet meadows. Widden (1977) reported a similar situation from Truelove Lowland: the more abundant bacteria of the sedge meadow were less productive than the sparser bacteria on the beach ridge. When dilution counts are converted to biomass, polygon troughs and rims of low-centered polygons are the most productive mi- crotopographic units. The Microflora 281 In the wet meadows, average seasonal abundance and productivity of plateable bacteria in the surface soil were higher in 1971 than in 1973. Because most of the increase in bacterial biomass occurs at the end of the season, these observations suggest that the bacteria in the surface of the soils were not encouraged by either the higher late season temperatures or increased precipitation during 1973. The low temperatures en- countered in deeper soil apparently did not suppress bacterial growth se- verely. In 1971 direct counts gave an estimate of the minimal seasonal bacterial productivity in wet meadow surface soils of about 1.2 g m"^ (1 cm)"' yr"' in the surface soil while the minimum productivity in the deeper soil was 3.7 g m'^ (I cm)'' yr"'. Again it is possible that microbi- vores, more active in surface soils, confuse these comparisons. ENVIRONMENTAL CONTROLS ON MICROFLORAL BIOMASS Empirical spatial and temporal patterns are presented above. In the following section, an attempt is made to account for the observed differ- ences in microbial biomass among microtopographic units solely on the basis of measured variables: organic substrate, inorganic nutrients, temperature, moisture, and oxygen. The analysis is hmited to field data for measured values from a number of different but natural environ- ments. Soil fungi are the only group for which the data base is suffi- ciently large to address specific relationships. Substrate If no other factors are acting, the soil microflora should increase as the amount of organic substrate increases. The total amount of organic material available to microorganisms was seldom measured directly. However, the percentage organic carbon is strongly correlated with bulk density (Chapter 7). Soils of higher bulk density contain more inorganic matter and less organic substrate per gram of soil. Thus fungal density should decline with increasing bulk density. For samples of all but the lowest bulk densities, fungal density does decline with increasing bulk density (Figure 8-10). Regression of fungal biomass in the surface soil (g mycelia cm"') on amount of carbon per soil volume (g C cm"') suggests that the response is associated with available substrate: fungal biomass = 0.633+0.014 (g C) {F = 12.47, n = 406). Thus, large differences in amounts of carbon on a volume basis are 282 F. L. Bunnell et al. T looa — o o -^ o* a> c o 3 >. l^ E 500- ^12 cm depth De 7 cm depth 1 I a ill O-m .1.19 .2.29 .3.39 4.49 .5.59 .6.69 .7.79 .8.89 in o l\ m <^ o E o> C CP 3 .2. .1. I an. IDJ 0.09 .1.19 .2.29 .3.39 .4.49 .5.59 .6.69 .7.79 Bulk Density, g cm' .8-89 FIGURE 8-10. The density and biomass of soil fungi arranged by soil bulk density and soil depth. (After Laursen 1975.) associated with differences in fungal biomass per volume of soil. Total carbon in the top 15 cm of polygon troughs (Figure 7-1) is about half that found in the wet meadows and the basins and rims of low-centered poly- gons; fungal biomass in soils of troughs is also lower than in basins, mea- dows or rims (Figure 8-4). However, fungal biomass per volume of soil shows no consistent relationship with bulk density alone (Figure 8-10). Apparently at the highest bulk densities some factor, such as poor aera- tion, reduces the "hospitality" of a volume of substrate regardless of the amount of organic substrate available. Inorganic Nutrients Observed relationships between soil fungi and amounts of phos- phorus are equivocal but suggest that phosphorus is limiting to fungal growth in some microtopographic units. No clear relationship appeared in the surface soil between daily measurements of resin-extractable phosphorus expressed on a volume basis and fungal density (values of the correlation coefficient ranged from 0. 10 in polygon troughs to 0.42 in basins of low-centered polygons). However, fungal density and biomass The Microflora 283 BASIN TROUGH RIM 5 2000 ■o c 01 Q o c 3 '6 o E o CD 800 600 400 200 r=0.92 2.0 1.6 1.2 0.8 ^ 0.4- r=0.77 J I L J I L_ 0 4 8 2000 > ,r=0.67 • 800 - 600 - 400 • • • 200 • -111 1 1 • 1 1 1 2000- ,. r=0.73 800 - • 600 • • 400 - 200 • • - • — 1_. i. 1 1 0 4 8 12 16 2 2.0 _r=0.66 1.6 - • 1.2 - 0.8 - 0.4 • • 1 1 1 1 • • 1 1 i» 1 1 0 4 8 2.0 _r=0.77 • 1.6 • 1.2 - 0.8 • • 0.4 - • 1 1 ] 1 1 12 16 20 Labile P, mg m ^ cm ' FIGURE 8-11. 77?^ density and bio mass of soil fungi in relation to available phosphorus 10 days earlier. An r value of 0.81 is statistically significant at the 0.05 level. (Barel, Laursen and Miller, unpubl.) do appear to be related to amounts of resin-extractable phosphorus pres- ent ten days earlier (Figure 8-11). Correlation coefflcients for phos- phorus with density and biomass were 0.77 and 0.92 for the basins of low-centered polygons where concentrations of phosphorus are generally low. The correlations were poorest for polygon troughs, where concen- trations of phosphorus are much higher than in basins. Low concentra- tions of phosphorus may reduce fungal biomass below the levels that can be supported by the amounts of organic substrate present. In the phosphorus-rich troughs fungal density and biomass appear limited by either the reduced amount of organic matter or other conditions associated with high bulk densities. Regressions of resin-extractable phosphorus against levels of fungal biomass or density 5 or 10 days earlier reveal no significant relationships. Thus, low concentrations of phosphorus do not appear to be a function of low densities of fungi. It is impossible to evaluate the role bacterial 284 F. L. Bunnell et al. biomass or activity may play in modifying levels of soil nutrients, a prob- lem that is particularly disconcerting when considering nitrogen. No sig- nificant relationships were found between concentrations of ammonium and abundance of either blue-green algae or fungi. Temperature, Moisture and Oxygen Relationships between microbial activity and temperature, moisture and oxygen have been demonstrated in a variety of natural substrates from tundra, including soil (Bunnell et al. 1977a, Bunnell and ScouUar 1981). Relationships between microbial biomass and these abiotic variables are more equivocal. No relationship was found between fungal density or biomass in the soil and soil temperature. However, rates of fungal growth do show relationships with temperature. Relative growth rate (rgr) is defined as: rgr = ln(D, /A-,) 57 where In is the natural logarithm, D, and D,-i represent fungal densities [m mycelia (gdw soil)"'] at times t and t-l, and At represents the time in days between / and f-l. When values for relative growth rate are stratified by temperature regardless of microtopographic unit, statistical- ly different distributions of growth rates are observed within different temperature classes based on the mean temperature during that 10-day interval (Figure 8-12). Using the Smirnov one-sided test the probability of temperatures 4 to 5°C and 5 to 6°C having the same distribution of growth rates is < 0.1. A similar value is obtained when temperature classes 5 to 6°C and 6+ °C are compared. The probability that classes 4 to 5 °C and 6+ °C have the same distribution of growth rates is < 0.(XX)1. At higher temperatures a greater proportion of the observed changes in 0.6 01 > 0.4 :; 0.2 •I 0*^ AtoSOQ -0.4 -0T2 0 0.2 Relative Growth Rote 0.4 FIGURE 8-12. The fre- quency of the relative growth rates of soil fungi at different tempera- tures. The temperatures used were the means of the lO-day interval. (Laursen, Miller and MacLean, unpubl.) The Microflora 285 biomass are positive and the mean growth rate is more rapid. Mean growth rates were also rapid when the 10-day mean temperatures were less than 4°C. This temperature range includes the early part of the grow- ing season when fungal densities increase rapidly (Figure 8-7). When or- ganic and inorganic substrates are readily available, rates of fungal growth may be high despite low temperatures. Within different microtopographic units, relative growth rates are normally distributed with a mean of near zero (-0.024) and a variance of 0.13. The distributions are most different between polygon troughs and rims of low-centered polygons, but cannot be distinguished statistic- ally using the Smirnov test. Although temperature regimes differ be- tween troughs and rims the difference is not sufficient to generate differ- ent patterns of fungal growth, and the relationship between temperature and fungal growth rates appears similar in all microtopographic units. Under natural conditions temperature appears limiting to fungal growth in aboveground substrates only early and late in the season. In 1972 and 1973 growth appeared to respond rapidly to early-season in- creases in the temperature (Figure 8-13). Despite continued warming. 2000 FIGURE S-13. Seasonal courses of fungal density and mois- ture content in Eriophorum litter from a wet meadow, 1972 and 1973. (Flanagan, unpublj 286 F. L. Bunnell et al. U 2000|-June >> o ■^ J 1500 - o 1000 o .= a< « c u = ^ 5001- - July 10 0 ^ - August I 10 0 5 Moisture Content, g water (gdw soil) in o CD 75 «65 65-75 »75 <65 6575 »75 Moisture Content, g water cm" soil FIGURE 8-14. The relationship of the density and biomass of soil fungi to soil expressed on a dry weight and on a volume basis for each summer month. (After Laursen 1975.) growth declined in early July, apparently in response to declining mois- ture. With increasing precipitation and moisture in August, rapid growth resumed. Thus, in plant Utter, fungal density appears to be correlated with moisture. In the highly organic soils the measured percentage of water by weight is largely a function of the amount of inorganic material present in the organic matrix. Estimates of fungal density per gram of soil show a clear but somewhat misleading relationship with amounts of water per gram of soil (Figure 8-14). As bulk densities and the amounts of in- organic material increase, amounts of water per gram of soil decrease and so does the relative amount of organic material available to support microbial biomass and growth. Expressed on a volume basis, moisture content shows little variation seasonally or with depth. Over the range of measures available, estimates of fungal biomass per volume of substrate are generally highest for the moisture class of 0.65 to 0.75 g water cm"' (Figure 8-14). Within this class the percentage moisture on a weight basis ranges from less than 120% to over 8(X)% moisture. Thus, when moisture contents are expressed on a The Microflora 287 weight basis their potential influence on fungal biomass and bulk density may be confused by strong covariance between both factors. It is note- worthy that the form of the relationship between fungal biomass and moisture is similar to that between microbial respiration and moisture (Bunnell et al. 1977a). Relative growth rates of soil fungi show no rela- tionship with soil moisture. These observations are consistent with the fact that soil moisture is only rarely limiting in the horizons where fungi are abundant. Although there are few data on amounts of oxygen, those available (Figure 7-6), together with the documented relationships among fungal biomass, organic matter, soil moisture and bulk density, suggest that the low fungal biomasses of some microtopographic units are a function of excessive moisture and oxygen depletion or excessive carbon dioxide. In- creases in bulk density resulting from additional fine mineral particles decrease total pore volume and the percentage of air-filled pores (Figure 7-3). With relatively high bulk densities, concentrations of oxygen de- cline rapidly with depth, carbon dioxide presumably increases, and amounts of fungal biomass are low despite the relatively large amounts of organic substrate. The decline in fungal biomass is due not simply to aeration, but also to available moisture (Bunnell et al. 1977a). In the more mineral layers, field moisture contents decline to 0.55 to 0.60 g cm"\ below the optimal value for fungi (Figure 8-14). Total bacterial biomass is largely unaffected by the declining pore volume or oxygen saturation and appears to respond more directly to available substrate (Figure 8-4, Table 8-2). Dominant Controls The broad relations documented above plus the seasonal courses of biomass indicate which environmental factors exert the greatest control on microfloral biomass and productivity. Within the standing dead vege- tation the relatively high quality of the substrate in June and, to a lesser extent, in September permit high levels of microfloral biomass and pro- ductivity despite relatively low temperatures. At these times, moisture is not limiting. As temperatures increase during the growing season, mois- ture becomes limiting, substrate quality declines, and microfloral bio- mass declines. The limitation by moisture is more pronounced for micro- topographic units more exposed to wind and drying, such as rims of low- centered polygons and tops of high-centered polygons. The pattern is similar in litter with only slight modiflcations. The early-season period of positive influence by substrate quality lasts a few days longer in litter than in standing dead, and the mid-season period of control by low amounts of moisture is shorter, markedly so for concave microtopo- 288 F. L. Bunnell et al. graphic units. Temperature again exerts its greatest influence early and late in the season, but may be a controUing influence throughout the growing season in moist, concave microtopographic units. Marked differences in bulk densities among soils of different micro- topographic units modify the pattern of control for soil organisms. Again, enhanced substrate quality at snowmelt and late season leaching of aboveground vegetation encourage the microflora, especially in the surface horizons. Death of roots in the deeper soil provides a similar pos- itive influence late in the season. At all but the highest bulk densities mi- crofloral biomass is correlated with amounts of carbon present and amounts of resin-extractable phosphorus. These factors establish some upper value for potential biomass, while productivity is strongly gov- erned by temperature. Temperatures do not differ sufficiently among mi- crotopographic units to bring about the observed differences, which are largely a product of moisture and aeration. Throughout the top 10 cm of soil, moisture itself is seldom limiting to soil fungi, but insufficient oxy- gen or excessive carbon dioxide is. High bulk densities and low oxygen levels encourage bacteria over fungi, especially in deeper soil layers. The anaerobic flora is capable of rapid decomposition within heated soils, but under natural conditions the influences of impeded drainage and low amounts of oxygen appear more profound than those of temperature. While relationships among mineral content, porosity, aeration and moisture appear to modify the apparent "hospitality" of the organic substrate available in some microtopographic units, the basins of low- centered polygons remain an enigma. Physical measures suggest that ba- sins should be favorable to microbes, but populations in standing dead vegetation, litter and soil are consistently low. The high correlations be- tween fungal abundance and resin-extractable phosphorus for basins, to- gether with low amounts of phosphorus in basins, suggest a possible chemical limitation. SUMMARY All components of the microflora of the coastal tundra at Barrow are characterized by lower species diversity than is observed in other Biomes, with the possible exception of some deserts. As in other Biomes, algae (photosynthetic microorganisms) are best represented by green and blue-green forms, which in the Biome research area contribute 42 of the 59 species identified. Several algal species are characteristic of fecally contaminated environments, while others are planktonic, reflecting the high amounts of moisture. Bacteria present in the system similarly reflect the high amounts of moisture. Anaerobes, particularly facultative anaer- obes, constitute 50 to 70% of the bacterial population and are important The Microflora 289 in decomposition. Sulfate-reducers are present, while sulfide-oxidizers are rare. The wet, acidic nature of the soils similarly discourages hetero- trophic nitrogen-fixing and nitrifying bacteria, while denitrifiers are common, accentuating the importance of nitrogen fixation by algae. Although all major classes of fungi are present, their diversity also is low. Most fungi exhibit specificity for particular microtopographic units. Generally, endemics are few and sterile forms make up more than 50*^0 of the isolates. In comparison to other Biomes, wood-rotting fungi are rare, reflecting the low availabihty of suitable substrate, while mycor- rhizal fungi are common in some areas. Aquatic fungi are present pri- marily as algal parasites. The apparent dominance of basidiomycetes among the fungi is associated with their role as mycorrhizal formers and with their more versatile, dikaryotic hyphal system, which may facilitate adaptation to an extreme environment. Average amounts of total microfloral biomass also differ signifi- cantly among microtopographic units. Total biomass ranges from 12 to 20 g m"^ in the upper 7 cm. Biomass is generally highest in wet meadows and polygon troughs, lower on rims of low-centered polygons, and low- est in basins of low-centered polygons. The inhospitality exhibited by basins is evident in fungal densities in standing dead vegetation and litter as well as in the soil. Generally the eukaryotic decomposers, fungi and yeasts, are much better represented in well aerated soils found on rims of low-centered polygons and tops of high-centered polygons, while bacteria become dominant in the wetter soils of polygon troughs and wet meadows. Plate counts of bacteria are commonly a factor of ten lower than values reported for soils of the temperate zone, while direct counts are higher. Because of the smaller size of the bacteria found in the coastal tundra at Barrow, total bacterial biomass is similar to that in temperate regions. Fungal biomass is frequently much lower than values reported for other Biomes and fungi-to-bacteria ratios approximate values for temperate regions only in the well aerated soils. In wet soils decomposi- tion is governed by bacteria to a greater degree than in most soils from other Biomes. The broad relationships with moisture and aeration that govern rela- tive distribution across microtopographic units also influence patterns with depth. Densities of eukaryotic organisms decline rapidly with depth, producing a shift from fungal to bacterial dominance. This shift reduces the potential to degrade compounds of large molecular weight and total carbon declines much less rapidly with depth than does root production. Intra- and interseasonal differences in abundance also demonstrate the influence of moisture in the tundra system, with moister periods or moister microtopographic units generally showing depressed fungal bio- mass or productivity. Both bacteria and fungi exhibit a similar seasonal pattern of biomass with early- and late-season peaks. While the peak in 290 F. L. Bunnell et al. early summer appears to be associated with release of organic material at thaw, the late peak may be associated with senescence and death of plant parts. This underlying pattern generated by availability of substrate is subsequently modified by the amounts of moisture present. Depending upon the microtopography and climate, fungi in the soil show minimal turnover rates for biomass of two to six times per season. Although low temperatures may Hmit amounts of microbial bio- mass, temperature-related differences between years or among microtop- ographic units were not observed. Rates of activity (Chapter 9) and growth, however, are positively related to temperature. The organic sub- strate seldom governs the biomass of the microflora, but low concentra- tions of inorganic nutrients, particularly phosphorus, are limiting to growth rates in some microtopographic units. The major control on composition and biomass of the microflora appears to be moisture and associated aeration. These factors are indirectly controlled by tempera- ture through the impeded drainage associated with permafrost, and are much modified by physical characteristics of the soils such as bulk density. Microflora Activities and Decomposition P. W. Flanagan and F. L. Bunnell INTRODUCTION Decomposition results in the disintegration and mineralization of organic residues. Most physical and chemical changes that occur in and around the decomposing substrate cannot be separated from the effects of microbial activity. However, degradation of plant and animal remains may be well advanced before significant ingress of microbes occurs (Dowding 1974, Flanagan and Veum 1974). Losses of weight and specific chemical constituents, which are often considered as measures of decom- position, may be initiated by plant-soil environment interactions that in- duce senescence and autolysis in moribund tissue. Important ecological phenomena that cause loss of weight from organic residues but occur somewhat independently of microbial activity include leaching, micro- faunal activities, and chemical reactions that influence mineralization. Soil invertebrates and protozoans are considered to influence de- composition rates indirectly by modifying the activities of the decom- poser organisms or microbes. They modify the environment through comminution of organic matter, and the microbial populations by graz- ing upon them (Chapter 11). Plant components Hke soil algae and the roots of vascular plants alter the environment through the provision of particular substrates and physical structure, and by modification of pH and supply rates for oxygen and other chemical compounds. Decomposition of organic matter is accompanied by synthesis of mi- crobial tissues which themselves decompose, contributing to further mi- crobial production. If the amount of substrate is limited, the potential for production of microbial biomass is dependent upon the efficiency of the microbial population in solubilizing, assimilating and incorporating organic remains, i.e. the efficiency of conversion of grams substrate to grams microbe, or the yield coefficient. The yield coefficient is influ- enced by climatic and substrate variables and is decreased significantly by the maintenance demands of preformed and forming tissues (Gray 291 292 P. W. Flanagan and F. L. Bunnell and Williams 1971). These aspects of microbial activities are best related to inputs from primary production. Under aerobic conditions, microbes decomposing a substrate break down complex organic molecules to end-products that are primarily inor- ganic (carbon dioxide, water and minerals); under anaerobic conditions the end-products assume a variety of organic and inorganic forms. The breaking down is accompanied by a loss in weight and energy content of the substrate as well as disintegration of its physical structure. An ob- server of the decomposition phenomenon thus witnesses physical and chemical as well as biological changes. Measures of decomposition incor- porate varying features of these changes, and no one measure quantifies decomposition perfectly. Belowground events are especially difficult to decipher because of the simultaneous respiration of heterotrophic and autotrophic organisms and the complex geometry of hundreds of sub- strates and microorganisms showing vastly different responses. Soils of the coastal tundra at Barrow are frozen for a large part of the year and this further complicates examination of biological processes within them. Given the complex of physical, chemical and biological processes comprised in decomposition, diverse methods have been employed to re- late the findings of individual, specialized techniques to general con- cepts. The concepts of decomposition used recognize that different mi- crobes have different enzymatic potentials or capacities to utilize various chemical constituents of naturally occurring substrates. Not only do microbes have different capacities to exploit substrates, but their enzym- atic potential, growth and respiration rates also respond differently to the temperature, moisture, oxygen and pH in their environment. Thus the capacity to decompose inherent in a particular microbial population is at any time modified by the environment. Acknowledging this conceptual framework, our studies of decom- position have examined the physiological potentials of different micro- floral constituents to utilize particular chemical compounds as sub- strates; the response of these potentials to environmental conditions; the biomass, biomass yield per gram of substrate, and maintenance demands of major microbial species; the response of rates of respiration and growth to changes in important environmental variables such as temper- ature, moisture, oxygen and dissolved nutrients; and the resultant loss rates of particular chemical constituents, carbon, calories and net weight. Measures of ability to utilize particular chemical compounds are in vitro assessments of an organism's ability to exploit selected natural substrates. Responses of respiration and growth to selected environ- mental variables and measures of biomass yields and maintenance demands are also evaluated, primarily by laboratory techniques, par- ticularly Gilson respirometry. Laboratory measures are related to field observations by simulation models (Flanagan and Bunnell 1976, Bunnell Microflora Activities and Decomposition 293 et al. 1977a, b). Field measures of decomposition and decomposition- related phenomena are primarily measures of rates of loss of specific substrate components. Four field measures have been employed. Weight losses from litter bags measure the rate at which litter be- comes sufficiently disintegrated that it disappears from the litter bag. A portion of this loss is due to microbial activity but some weight is lost by leaching and physical comminution by invertebrates and by the freeze- thaw cycle. Ingress of microbial, plant, animal and mineral matter can confound estimates of weight loss. Chemical analyses of substrate composition coupled with measured weight loss estimate the rate of disappearance of major chemical com- pounds such as cellulose or phosphorus. These measures are also an inac- curate estimate of microbial activity. Not only are other processes also acting (e.g. leaching) but the microbial populations and their chemical composition are inseparable from the substrates. Measurements of rates of evolution of carbon dioxide represent the rate of mineralization of complex organic compounds to carbon dioxide, water and residual constituents such as minerals, and are perhaps the best measure of aerobic microbial activity. Depending on the substrate measured and the method used, various inaccuracies are introduced, either by the effects of methodology, as with the physical disturbance in Gilson respirometry, or by inclusion of carbon dioxide evolved from plant roots and soil invertebrates. Anaerobic decomposition processes are incompletely measured by carbon dioxide evolution and may present an important omission in some habitats. Measurements of microbial biomass during decomposition of above- and belowground substrates permit one to relate the abundance of major decomposer agents to substrate availability and quality. These measures, coupled with laboratory data on microbial growth and yield from varying substrates and information indicating microbial mainten- ance demands, allow approximation of the microbial production in the field. Microbial biomass and production may then be compared with similar measures from leaves, roots, microfauna etc. Additionally, stud- ies of microbial biomass permit compensation for underestimates of weight loss caused by growth of microbial tissue in litter bags. By know- ing microbial mineral content per gram we can approximate values for mineral immobilization and cycling in microbial biomass and produc- tion, respectively. In the subsequent discussion, both field and laboratory measures are employed to help define patterns of decomposition. Since decomposition is equated with the microbial mineralization of carbon, these measures are to varying degrees inaccurate. No one measure discretely encom- passes decomposition as defined, but integrated with a knowledge of mi- crobial biomass, dynamics, and physiology the measures contribute to a 294 P. W. Flanagan and F. L. Bunnell synthetic view of decomposition. Tools of integration include correlative analyses and computer simulation models. Although these tools are in some instances novel and sophisticated, the conceptual framework employed owes much to the seminal work of Douglas and Tedrow (1959) (see Bunnell and Tait 1974). The approach in the following discussion moves through the con- ceptual framework as it is presented above, first examining the potential to exploit particular substrates that different microfloral constituents possess. Then we examine the manner in which this potential and other critical activities such as respiration are influenced by environmental var- iables such as temperature, moisture and oxygen. The discussion in the section on Decomposer activities and decomposition utilizes simulation models to combine the responses of microbial respiration to individual environmental variables and compares the predictions with integrative measures such as weight loss. Measures of microbial biomass, yield and efficiency are related to substrate availability and potentials for mineral cycling and immobilization. POTENTIAL OF THE MICROFLORA TO UTILIZE SUBSTRATES The decomposition of organic remains proceeds mainly through the action of microorganisms that can use them as a source of energy and nutrients. Most soil microbial populations are heterotrophic and the or- ganisms compete for available substrates. In the coastal tundra at Bar- row, microbial saprophytes are more competitive and abundant than are parasites and the present discussion ignores the latter. Not all sapro- phytic microorganisms utilize and compete for the same substrates. Although an individual microorganism may be encouraged by the pres- ence of a specific substrate that it can use, its potential to exploit that substrate is further modified by environmental conditions and "competi- tive saprophytic ability" (Garrett 1963). The composition of the micro- bial population inhabiting a particular substrate and thus the decomposi- tion rate of the substrate are therefore dependent upon the environ- mental conditions and competitive ability of the microorganisms as well as the measured potential to utilize specific substrates. The data pre- sented here are based almost entirely on in vitro measurements. The dis- cussion is biased towards treatment of the mycoflora rather than of the total microbial population. Major substrates for decomposer organisms at Barrow can be divid- ed into two main categories on the basis of their chemical composition, pattern of dissolution and utilization by microorganisms. These categor- ies are 1) low molecular weight, water- and/or 80%-ethanol-soluble frac- Microflora Activities and Decomposition 295 tions that are readily leachable, and 2) the more recalcitrant compounds, such as lignin, cellulose, hemicellulose, pectin and starch. The first, more soluble group contains approximately 25<^o of aboveground plant pro- ducts, exits as leachate from moribund tissues and contains the bulk (> 80<^o) of plant leaf nitrogen, phosphorus and potassium. The second group represents the bulk of the available organic substrate and is rela- tively poor in nitrogen, phosphorus and potassium (Flanagan and Veum 1974, Van Cleve 1974). Substrates of the more soluble group are fre- quently utilized by organisms decomposing the more resistant group. Microflora in all Biomes display a broad diversity of enzymatic po- tential to decompose the various organic substrates. Tundra microflora share this capacity, with the restriction that the cold-dominated environ- ment has selected for taxa or strains that utilize these substrates under cooler conditions. The restricted number of taxa may or may not reduce the potential to complete a given phase of decomposition such as conver- sion of cellulose to carbon dioxide, but it does reduce the number of modes that such a reaction can follow in the ecosystem. As in many temperate zone habitats, bacteria in tundra are often weak competitors with the fungi for those substrates that both groups have the enzymatic potential to metabolize. The competitive difference is especially obvious in such habitats as standing dead and litter, and in drier surface soils. Conversely, in wet habitats bacteria play a propor- tionately greater role in decomposition. In extremely wet or anaerobic habitats, such as sediments of tundra ponds or soils at depth, bacteria are the dominant group of decomposers. The potentials to utilize specific forms of nitrogen and phosphorus are addressed in Chapters 7 and 12; here we consider only carbon. Bacteria The sources of carbon most commonly exploited by soil bacteria are of intermediate molecular size (Table 9-1). Large molecules such as pec- tin and cellulose, which form major structural entities of plant cells, can be decomposed by relatively few of the plateable bacteria. Cytophaga ap- pears important among the cellulose-decomposing bacteria since it fre- quently occurs on plate isolations. Enrichment studies of "most prob- able number" show Cytophaga populations varying from 10' (gdw soil)"' after thaw to 10* (gdw soil)"' at mid-season. No cultures of indigenous aerobic bacteria that could decompose humic substances were obtained. The relative ability to utilize humic substances is one of the major enzymatic differences that separate bacteria from fungi (cf. Table 9-1 and Figure 9-1). Widden (1977) has documented similar differences be- tween bacteria and fungi at Devon Island. Enzymes that cleave the aro- 296 P. W. Flanagan and F. L. Bunnell TABLE 9-1 Percentage of Bacterial Types at Barrow Capable of Utilizing Specific Carbon Sources Carbon Bacteria source (%) Succinic acid 92 Citric acid 84 Glucose 78 Maltose 66 Starch 42 Pectin 28 Lactic acid 25 Cellulose 5 Lactose 0 Lignin 0 Tannic acid 0 Note: Based on 200 randomly selected aer- obic types isolated from the 0- to 2-cm soil depth in wet meadows and tested at 15 °C. Source: Benoit (unpubl.)- matic ring of humic materials require the presence of oxygen. Thus, as humic materials move from the surface into the less aerobic subsurface layers of the soils, the probability of their decomposition is markedly reduced. The differential capacities of bacteria and fungi to survive low amounts of oxygen and to exploit humic material are instrumental in the accumulation of organic matter in the soil horizon. The strictly anaerobic portion of the microflora remains the least known in terms of physiology and in situ activity. Rapid development and intense activity of anaerobes on the soil plots heated to 15 to 20 °C il- lustrates the potential of this group. Enrichment cultures of strictly an- aerobic cellulose-decomposers and methane-producers were obtained from anaerobic soils on these plots. There is no evidence that anaerobic bacteria can degrade humic materials; therefore the activity of the anaer- obes becomes limited when the pool of rapidly decomposable material originating from the death of belowground parts is exhausted. Contin- ued decomposition requires the action of other microorganisms and a change of abiotic conditions. Bacteria Microflora Activities and Decomposition 297 Fungi substrate: Humic Acid — ^SD —(Tannic Acid) 'g Gallic Acid — ^ 11 — (Lignin) ,. _ 1*^ Cellulose 1 SD (Standing Dead Leoves) 1 ^( d 100 o 50- I — 1 — 1 — 1 — 1 — 1 — r -T ■!-■ T T - - A - M.S./ — / 1 1 . / M.s^ K/ . ■tl \ P^ A-c^- " - M.S./ \»/ • ../ \ _ ,' X a-/ \ // o..'^ay 0 \ o/^.. ° / A "~o jyj 500% dw) may depress microbial respiration. In a Gilson respirometer the depressant effect can be eliminated by increasing available oxygen (Flanagan and Veum 1974) and is likely associated with reduced rates of supply of oxygen. As tem- peratures increase, respiration in some substrates shows saturation of metabolic moisture demand at amounts of moisture less than 400% dw. Temperature-moisture interactions, as demonstrated by studies of microbial respiration from plant remains, may reflect differences in oxy- gen diffusivity in water or changes in microbial oxygen demand as tem- perature varies. Oxygen and Microflora! Metabolism Oxygen is critical in microbial metabolism because it serves as an electron acceptor in the breakdown of organic matter. In the soils at Bar- row the potential alternate electron acceptors such as nitrate and sulfate are present in such low concentrations (< 1 ppm) that they can support little activity. Diffusion of oxygen into the soil is often below levels re- quired for optimal microbial activity, and in very wet soils pore spaces become filled with moisture. The effect is decreased microbial activity among the aerobic component; the cause is not too much moisture, but apparently too little oxygen. It is equally plausible that carbon dioxide 308 P. W. Flanagan and F. L. Bunnell operating separately from or in concert with a deficiency of oxygen reduces microfloral metabolism (Burges 1958, Griffin 1972). The highest oxygen levels measured within the upper 20 cm of soil were observed in the soils of the basins and rims of low-centered poly- gons and the tops of high-centered polygons (Figure 7-6). It is surprising to find that on the dates they were sampled the basins of low-centered polygons had high oxygen values throughout the soil profile. Soil thin sections from basins, unlike rims, indicate a structure that should impede drainage and aeration (Everett, pers. comm.). The data for both in vitro respirometry and measured evolution of carbon dioxide indicate that basins of low-centered polygons have low rates of decomposer activity (Figure 9-9). For example, in 1972 the mean seasonal rate of respiration, as measured by Gilson respirometry, for 0- to 2-cm basin soils was 10.06 lA O2 (gdw soil)"' hr"' whereas the values from similar depths in the trough and very wet meadow were 31.37 and 43.15 m1 O2 (gdw soil)"' hr"' respectively. The consistently low decomposer activity and low primary productivity probably act to maintain relatively high levels of oxygen in the basin soils. The shift of bacteria-to-fungi ratios along the oxygen gradient (Chapter 8) suggests that anaerobiosis does not eliminate decomposition of the soil organic matter but changes the quality of that decomposition. In the wet meadow, depressed oxygen concentrations in the deeper soils were associated with high levels of decomposer activity measured by Gil- son respirometry (Figure 9-6). Thus the high values for respiration from subsurface samples represent activity of facultative bacteria after full in- duction of potential for oxidative phosphorylation in the Gilson respir- ometer. The rapid response to oxygen indicates the active enzymatic state of the cells. The zone of anaerobiosis at the front of the thaw zone ap- pears to be a result of the rapid decomposition of readily available substrates. Experiments with heated soil further demonstrate that anaerobic conditions per se do not prevent decomposer activity in tundra soils. Oxygen saturation in the soil solution in heated soils declined to 0% by the 3-cm depth. On these plots methane routinely composed 60% (with a range of 42 to 65% of the gas released), with the balance primarily of carbon dioxide. Methane production provides good evidence for anaero- bic activity because methane producers represent one of the most strictly anaerobic groups, and can be killed easily by transitory exposure to oxy- gen. Despite anaerobic conditions, rates of decomposition were high in heated soils. Evolution of carbon dioxide over a 33-day period in late summer was 39% higher in heated soils, and after 12 months the energy content in the heated soil was less than in control soils by 21 % in the up- per 2 cm and by 16% at depths of 7 to 12 cm. The higher rates of micro- bial respiration in the heated soils are a result of the increased tempera- A- Microflora Activities and Decomposition 309 tures and approximate the rates expected, given the respiratory Q,o of 1.78 measured from unheated soils and a 4° to 6°C increase in mean daily soil temperature. The heat treatment documents the in situ poten- tial for anaerobic decomposition present in tundra soils suggested by the high number of facultative bacteria relative to strictly aerobic species (Chapter 8). DECOMPOSER ACTIVITIES AND DECOMPOSITION Two general approaches are used in analyzing the response of de- composer organisms to temperature, moisture and other environmental phenomena. The first approach employs integrative measures of all mi- crobial groups. The evolution of carbon dioxide, the rates of weight loss from selected substrates, and the patterns of nutrient concentration are each considered direct or indirect functions of the activities of all micro- bial groups. The second, more direct, approach examines the individual activities of specific microbial groups. The specific activities suggest the contribution particular groups make to the general integrative measures such as carbon dioxide evolution. We discussed some specific responses earlier in this chapter without relating them to patterns of weight loss or carbon dioxide evolution measured in the field. The present discussion il- lustrates how these activities are enacted in the changing environment to produce the observed patterns of weight loss or decomposition. Decomposition of aboveground parts of graminoids begins at the time when necrotic patches appear on the leaves and stem bases. The ne- croses are most apparent from mid-August onwards. Before or concur- rent with visual signs of senescence the leaves lose up to \2% of their weight as green healthy tissue (Figure 9-7). The initial loss, which occurs prior to substantial microbial ingress, and whether or not rain has fallen, is apparently caused by translocation to belowground parts. The freeze- thaw cycle and the physical throughput of water remove up to 18% of the dry weight of overwintering leaves by the spring. During and prior to the period of leaching substantial microbial activity may take place, con- tributing to overall weight loss, but microbial contributions to weight loss in this and the previous phase of aboveground plant weight losses are undetermined. During the period from mid-August to the end of spring runoff, up to 30% dry weight may be lost from leaves of graminoids (Figure 9-7, Table 9-5). Three simulation models relate microbial activities quantitatively and unambiguously to the environmental phenomena that govern them. One model, gresp (Bunnell et al. 1977a), relates the respiratory response of microbial populations to changing temperature and moisture. The sec- ond, DECOMP (Bunnell et al. 1977b), expresses the respiration rate as a 310 P. W. Flanagan and F. L. Bunnell 100 c 90 - o » 80 Z 70 60 Va n. T — I — I — I — I — I — I — I — I — I — I — I— I — I — r ^ Live J L Dead J u _u JJASONDJFMAMJJASO FIGURE 9-7. Progression of leaf weight of Carex aquatilis and Eriophorum angustifolium, combined. Hatched bar shows the amount of material removed in vitro by warm water and 80% ethanol. Vertical bars indicate the standard errors. The dashed line indicates the weight loss before ingress of microorganisms is well under way. (Flanagan, unpubl.) function of substrate chemistry. The third, abisko ii (Bunnell and Scoullar 1975), integrates the effects of changing meteorological condi- tions and substrate chemistry within an ecosystem framework. The models document relationships between weight loss and microbial activi- ties, as they are influenced by abiotic variables and substrate chemistry and the relationship of the biomass of microbial populations to primary production and turnover of organic matter. Although the development of these models was based on tundra research, their predictive abilities have also been tested for conditions found in the taiga and moors (Bun- nell et al. 1977a, Bunnell and Scoullar 1981). Temperature, Moisture, and Microbial Respiration The function GRESP represents a formal statement and complex hy- pothesis of the manner in which temperature, moisture and substrate fea- tures influence aerobic respiration of microbes. It treats aerobic respira- tion as a function of the supply rates of water, oxygen and organic nutri- ents. The critical features of the hypothesis are presented: R{T.^4) = [M/f^a, +M)][ff2/(ff,+M)] «3a4'^-'°'^'° where R{T,M) = ^A CO2 respired (g substrate)"' hr"' at temperature T and moisture M Microflora Activities and Decomposition 311 T = temperature, °C M = moisture, percent dry weight a,,...,ai = substrate specific parameters. The rationale of the gresp function has been presented elsewhere (Bun- nell and Tait 1974, Bunnell et al. 1977a) and only a summary is repeated here. Microbial respiration is assumed to be related to the moisture po- tential of the substrate via two saturation processes. The first process is related to the metabolic water requirements of decomposer organisms and embodies a convention of soil mycologists, that is, the expression of water content on a relative basis, or as a percentage of the value when the soil is saturated (Griffin 1966). This process is expressed as M/{a, +M), where M represents the percent water content on a dry weight basis and fir, represents the percent water content at which the substrate is "half- saturated" with water or respiratory activity is at half its optimal level. The second saturation process occurs at high moisture levels. It is assumed to represent the effect of water on gas exchange with the atmos- phere either of oxygen, carbon dioxide or both. The simplest formula- tion is employed (Bunnell and Tait 1974). Since the degree to which gas exchange is inhibited can be expressed as M/(a2 + M), the degree to which it is not inhibited can be expressed as: l-[M/(a2-hM)] or 02/(02+ M). Again M represents the moisture content, and 02 represents the percent water content at which gas exchange is limited to half its optimal value. The third and fourth factors, temperature and substrate characteris- tics, are treated as a substrate-specific Qio relationship: where a^ is the substrate specific respiration rate that occurs at 10 °C when neither moisture nor oxygen are limiting and a^ is the Qio coeffi- cient. Alternative formulations for both moisture and temperature influ- ences on rates of nutrient supply are discussed by Bunnell et al. (1977a); the treatment of substrate characteristics is pursued later in this chapter. According to the gresp function any one of the major determinants of rates of respiration (moisture, oxygen, temperature and substrate) can effectively reduce the rate of microbial respiration independently of the other factors. Thus, the rate determinants are combined multiplicatively. Evaluation of the complex hypothesis represented by the ores? function indicates that it predicts carbon dioxide evolution more accu- rately from aboveground substrates than from tundra soils. In the some- what more aerobic soils of the taiga the model accounts for 78 to 84% of the variability in respiration rates (Bunnell et al. 1977a). The generality 312 P. W. Flanagan and F. L. Bunnell 500 500 SD-C.Q.-I 500 500 500 SD-E.a-2 o _f- Litter 500 Measured Simulated FIGURE 9-8. Measured and simulated rates of microbial res- piration in relation to moisture and temperature for: 1-yr-old standing dead o/Carex aquatilis (SD-C.a.-l); 2-yr-old stand- ing dead ofC. aquatilis (SD-C.a.-2); 2-yr-old standing dead of Eriophorum angustifolium (SD-E.a.-2); and mixed graminoid litter. (After Bunnell et al. 1977a.) Microflora Activities and Decomposition 313 TABLE 9-2 Parameter Values Giving the Best Fit of the gresp Func- tion to Data from Seven Substrates and Coefficients of Determination No. of Oi 02 (Ml O, O4 Substrate obs. water (% dry wt) g-' hr-) (Q.o) r' rl* Carex ( 1 yr old) 96 59 NE 204 3.33 0.79 0.86 Carex (2 yr old) 45 213 822 296 2.18 0.63 0.71 Eriophorum (newly dead) 54 416 416 777 8.79 0.65 0.96 Eriophorum (1 yr c .Id) 79 74 NE 184 2.56 0.64 0.72 Eriophorum (2 yr c Id) 65 19 NE 107 2.79 0.72 0.77t Dryas (2 yr old) 38 160 160 158 1.75 0.36 0.43 Total Barrow litter 346 116 2820 232 3.74 0.66 0.71 * rl is an estimate of the accountable variation explained by a nonlinear model. t An optimum solution was not found. NE — No estimate; number of measurements at moisture contents >500"7o moisture (% dry wt) are <10. Therefore 02/(^2 + A/) was set equal to 1. Source: After Bunnell et al. 1977a. and accuracy of the model in predicting measured microbial respiration from aboveground substrates, assessed visually (Figure 9-8) and statistic- ally (Table 9-2), suggest that the hypothesis is applicable to a diversity of aboveground substrates. Coefficients a^ and ai assess the moisture range over which mi- crobial respiration is little affected by changes in moisture levels. The range encompassed by these coefficients extends beyond the optimal moisture regime determined for individual species, particularly at the up- per end. Excluding the poorly constrained values (NE in Table 9-2) the weighted mean of the mid-point between a, and ai for aboveground sub- strates is 461% moisture on a dry weight basis. For the microbial com- munity in toto, respiration from dead vegetation is depressed far more as amounts of moisture approach the lower thresholds than near the upper thresholds. Within plant litter where aquatic fungi are more abundant, the upper limits to the moisture range may be greatly extended (Table 9-2; Flanagan and Veum 1974). Some substrates, particularly in taiga soils, have computed half- saturation values that are equal for both a, and az (e.g. newly dead £"^0- phorum angustifolium, Table 9-2). The resulting model with a, set equal to 02 has only three parameters but accounts for 96°/o of the variation in respiration from that substrate (Bunnell et al. 1977a). However, Bunnell et al, (1977a) noted that equal values of a, and 02 are likely an artifact of the model. Given the multiplicative form of the gresp function, a 314 P. W. Flanagan and F. L. Bunnell narrow-peaked response of respiration versus moisture can be obtained only by having a, equal 02. Despite an uneven data base, trends in coefficients a, and 02 are re- vealing. With increasing age and pitting within the substrate, the mois- ture range for effective respiration appears to broaden (see E. angusti- folium. Table 9-2). Bunnell and Tait (1974) stated that the volume of water relative to the amount of organic matter was critical. Thus, they predicted that the moisture range over which respiration was uncon- strained would broaden with age in aboveground substrates and narrow with increasing depth and bulk density below ground. The trend below ground has been documented most rigorously for aspen forest floor in the taiga, and does show a gradual decrease in the effective moisture range with depth (Bunnell et al. 1977a). Coefficient a^ represents the respiration rate at 10 °C when moisture and oxygen are not limiting. It is assumed to be a measure of substrate quality and as such should decline with the age of the substrate. Bunnell et al. (1977a) documented the expected pattern within the taiga forest floor; among substrates of the Biome research area it is evident among the E. angustifolium age classes (Table 9-2). The exponential response of respiration with temperature, defined by coefficient a^, assumes Q,o values ranging from 2.2 to 8.8, with younger substrates showing a higher Qio than older substrates (Table 9-2). There are two possible reasons. 1) Newly senescent substrates have not experienced a winter and may not be well colonized by psychrophilic organisms; thus they would show lower rates of respiration at lower temperatures and higher Qio values. 2) Younger substrates contain greater proportions of constituents of low molecular weight which appear to have higher Qio values associated with their utilization as discussed earlier in this chapter. Over the range 0° to 10 °C the weighted average of Qio values for all aboveground substrates tested at the Biome research area is 3.65. Observations suggest that while younger aboveground substrates have the chemical potential for higher respiration rates than do older substrates, respiration is more likely to be constrained by temperature and moisture. The high Qio values from younger substrates imply a pop- ulation poorly adapted to low temperatures. The narrower moisture range suggests that drying by wind frequently may reduce realized respir- ation. These environmental constraints act to ensure that nutrients pres- ent in newly dead standing vegetation are not released into the system un- til the spring thaw. The same clear pattern of carbon dioxide evolution with tempera- ture and moisture is not observed for decomposition processes below ground. The best fit of the gresp function to Gilson respirometry meas- ures of wet meadow soil accounts for only 10 to 20% of the variation. The computed Q,o is near 2.0 for a variety of soils and the optimal mois- Microflora Activities and Decomposition 315 1 ' ^ FIGURE 9-9. Seasonal courses of the evolution of carbon dioxide from soils of four microtopographic units and soil temperature near the surface in the ba- sin of a low-centered poly- gon. Carbon flux meas- ured by KOH titrations in darkened lysimeters. Data are ten-day running means for 1973. (Benoit, unpubi) ture level is about 75 to 80%. Few measures incorporating higher mois- ture levels were available and the estimate is likely low. Rates of soil res- piration as measured by Gilson respirometry can be extrapolated only tenuously to estimate decomposition rates in the field. At best they repre- sent a potential rate of decomposition that may not be realized. Another estimate of field decomposition rate and the factors which control it can be made from in situ measurements of carbon dioxide evo- lution although these measures include respiration of roots as well as microflora. Field data were obtained by daily potassium hydroxide titra- tion of gas collected from plastic cores sunk into the soil in 1973. Mea- sures of carbon dioxide evolution from different microtopographic units all peak in early August (Figure 9-9). Although the peak appears correl- ated with temperature, early August is also the time of maximum above- ground biomass of vascular plants and intense activity by soil fauna. Logarithmic regression of the daily evolution of CO2 m"% as measured by lysimeters in wet meadow soils, against mean daily soil temperature estimates a Qio of 1.89: R = 1465.6x1.89 Tmean/lO r' = 0.37, a < 0.01 where R is ml CO2 m ^ day ' and Tmean is the mean daily temperature. 316 P. W. Flanagan and F. L. Bunnell Linear regression with the same data produces: R = 1476+1 17.4 xTmean r' = 0.39, a = 0.01. For all microtopographic units, linear regressions of carbon dioxide evolution versus mean daily temperature and maximum daily tempera- ture consistently provided higher coefficients of determination, r\ than regressions involving 2^^ '"and e^'^'" (where T = temperature). The ob- served linear response to temperature may be a result of the summation of a number of exponential responses. The contribution of plant roots to observed carbon dioxide produc- tion was analyzed by comparing undisturbed soil cores with cores effec- tively stripped of primary producers. Linear regressions of carbon diox- ide evolution from stripped cores also provide better fits than do expo- nential models for the effect of temperature. The relative response (the predicted response at 10 °C divided by the predicted response at 0°C) is higher for stripped cores than for cores on which the graminoid and moss canopy was left intact. Removing the plant cover increased the relative response from 2.08 to 3.1 in basins of low-centered polygons and from 1.59 to 1.89 on rims of low-centered polygons. The intercepts of equa- tions for stripped cores of basin and rim soils were 155 and 634 ml CO2 day', about half the level of the intercepts of untreated cores (353 and 1121 mlCOjday-'). Direct comparisons between treated and untreated cores should be viewed with caution because microbial populations in treated cores are not experiencing the same environment as the controls. Further, there may have been increased root respiration associated with clipping. How- ever, the results do suggest that the near-linear response of carbon diox- ide evolution to temperature changes in soils is in part due to the influ- ence of primary production. The lower relative response of soils on rims, which support higher primary production than do soils in basins, corrob- orates the suggestion, as do the high Qio values observed for microbial respiration in standing dead and litter substrates (Table 9-2). In short, microbial activity below ground appears to respond more strongly to changes in temperature than do processes of primary production below ground. Undoubtedly some of the differences in carbon dioxide evolution observed among microtopographic units (Figure 9-9) are associated with differing primary productivity. Rims of low-centered polygons and poly- gon troughs evolve approximately twice as much carbon dioxide as ba- sins of low-centered polygons and support considerably greater primary production. Evolution of carbon dioxide from meadow soils is still great- er, possibly reflecting the higher biomass of bacteria in these soils. We cannot distinguish between the influences of soil moisture and primary Microflora Activities and Decomposition 317 production on measured respiration, but acknowledge that the difference in soil respiration among microtopographic units is greater than that ex- pected from differences in primary productivity alone. In taiga soils coefficient a^ and total respiration decrease with depth in the profile, while the measured Q,o increases (Bunnell et al. 1977a). Such findings suggest that the relative contribution from primary pro- duction in the taiga decHnes with depth and that the effect of tempera- ture on microbial populations increases with depth. In both tundra and taiga, microorganisms in the litter layer continue respiration at lower temperatures than do organisms at depth in the soil. Surface layers are subject to wider ranges in both temperature and moisture than are deeper layers. Thus it is not surprising to find deeper communities apparently adapted to narrower ranges of temperature and moisture than surface communities. The tendency within tundra soils for more linear relation- ships with temperature may reflect an adaptation to low temperatures. The sensitivity analyses of the model based on the function gresp that were conducted by Bunnell et al. (1977a) have been extended using other substrates from tundra. In all cases the model is most sensitive to coefficient a^, which determines the predicted Qio. Relative sensitivities to other coefficients vary with the substrate. For those substrates where the overall fit of the model is close (r^ 5= 0.80) we observe that the broad 100 96- c a E 92 0) q: a) 88 84 T 1 ~~~-a-.,Soil ""^ -..^tanding Dead \ -r - S,Litter \ \ X - - Jun 1 1 1 1 1 Aug '70 Aug '71 Aug '72 Aug '73 FIGURE 9-10. Progression of weights of cellulose placed in surface soil, litter and standing dead mate- rial. Vertical bars give standard error. (After Bun- nell et al. 1977b.) 318 P. W. Flanagan and F. L. Bunnell response of microbial respiration in substrates from tundra and taiga is most sensitive to temperature, then to substrate chemistry, and least sen- sitive to amounts of moisture, particularly high levels of moisture. Des- pite the variable response to moisture, elimination of moisture effects from the model reduces its predictive ability by a minimum of 23 to 31% (Bunnell et al. 1977a). Also, it is important to note that while respiration rates are relatively insensitive to moisture over the range of respirometry data collected, moisture levels in the field may become high enough to reduce respiration significantly. Temperature, moisture and oxygen are all important modifiers of the rate of respiration of tundra microorganisms. The overall effect of the microbial population and micrometeoro- logical factors can be seen in the relative decomposition rates of a uni- form substrate, cellulose, placed in three different microhabitats (Figure 9-10). Over several years, the weight loss from the cellulose was greatest in the litter, intermediate in the standing dead, and lowest in the soil. As both soil and litter have more cellulolytic decomposers than standing dead (Figure 9-1), the results support the hypothesis that moisture may be limiting in the standing dead (Chapter 8). The considerable decrease in decomposition rate noted for the soil suggests that conditions in the soil are less favorable overall to decomposition than those above ground. Substrate Chemistry and Microbial Respiration Microbial respiration is assume J to be influenced by substrate chem- istry as well as by temperature and moisture. This assumption is broadly accommodated by coefficient aj in the function gresp. Coefficient O} represents the respiration rate at 10°C when neither moisture nor oxygen are limiting, and is employed to establish the upper level or amplitude of the response surface for respiration. The "quality" of the substrate is thus directly proportional to the magnitude of a,, which in turn is di- rectly correlated with the percentage of ethanol-soluble compounds or percent glucose (Bunnell et al. 1977a). The significance of a, to broad patterns of respiration is estimated by the preceding sensitivity analyses. Earlier works (Henin et al. 1959, Minderman 1968) proposed that observed rates of weight loss result from the summation of rates from specific chemical components, but did not relate these rates to microbial activities. Bunnell et al. (1977b) extended these earlier models of decom- position to encompass not only the observed patterns of weight loss, but the microbial activities producing these patterns. Ethanol-soluble com- pounds disappear five to six times as fast as other constituents of natural substrates (Figure 9-11). Combination of these two chemically defined groups produces the common departure from a simple exponential Microflora Activities and Decomposition 319 IOOf: 80 60 <:7> c E 1) 40 ZO- IC 8 6h t • ■■ I . lnsolubl"e *•--- ^^ • "^ ~"\ & - o ~^--^Soluble o O ~~~- A ..... 1 J 1 , 1 . 1 9- 0 I 2 3 Time, yrs FIGURE 9-11. Percentage of ethanol-soluble and ethanol-insoluble compounds remaining in decom- posing Carex aquatilis (A,, A) and Eriopiiorum an- gustifolium (m, o). Percentages are based on weight per unit area. (After Bunnell et al. 1977b.) illustrated by many decomposing substrates (Surges 1958, Minderman 1968, Satchell 1974). Here we address the manner in which the phenomenon of substrate "quality" represented in the function gresp by the single coefficient Oj can be related to substrate chemistry more directly. In accordance with earlier workers different constituents are assumed to have their own chemical-specific rates of decomposition. In addition the decomposition rate of each chemical constituent is assumed to be a function of the tem- perature and moisture-dependent respiration rate of the microflora, and thus changes seasonally or even daily. The observed decomposition rate of a substrate is assumed to be the sum of the temperature-moisture- chemical-specific rates of utilization times the amount of each chemical constituent present. Observed rates of decomposition thus change with the meteorologically induced changes in rates of microbial respiration and the changing capacity of the substrate to provide energy to the mi- crobial population. As developed by Minderman (1968), the decay rate of a substrate can be expressed as: (dY/dt) = I {dy,/dt) i= 1 (dY/dt) = I -k,y^ i= 1 where Y = total weight of the decomposing substrate 320 P. W. Flanagan and F. L. Bunnell /; y, = weight of substrate component /; 1 y, = Y 1 = 1 k, = decay rate of substrate component /'. The above equation states that each substrate component decomposes at a constant rate of decay specific for that substrate constituent and inde- pendent of the amount of other substrate components present. The rela- tion of decomposition to temperature and moisture discussed earlier sug- gests that the above equation can be written RiT,M) = i, r,{T,M)y. where R{T,M) = respiration rate of the total substrate at temperature T and moisture M r,{T,M) = respiration rate of substrate component /, a function of temperature and moisture y, = amount of substrate component / present, as before. As expressed in the above equation, the model of decomposition not only accounts for the observed differences in response surfaces of res- piration versus temperature and moisture for substrates of different chemical composition, but also accounts for the observed differences in rates of loss of different chemical components from a substrate in the field. To document the relative contributions of different chemical con- stituents to total respiration, the influences of temperature and moisture must be reduced or removed. Bunnell and Tait (1974) proposed several methods for separating the temperature and moisture effects from chemical-specific effects. Three methods have been evaluated by Bunnell et al. (1977b). Their evaluation suggests that dominating influences on the pattern of respiration for any specific constituent are levels of tem- perature and moisture. Only 1 to 4% of the variation in instantaneous respiration rates is due to chemical composition. These observations are not incompatible with the preceding sensitivity analyses, which suggest that the overall response surface of microbial respiration versus tempera- ture and moisture is sensitive to substrate quality. The general amplitude of the response surface, and thus its overall shape, are sensitive to substrate quality (Bunnell et al. 1977a). Variations in temperature and moisture, however, account for more of the variation of the total surface (Figure 9-8) as it rises to the amplitude set by substrate quality. When summed over a year, even small differences associated with substrate chemistry will produce distinctly different annual rates of loss. In their evaluation of the relationships expressed in the last equa- tion, Bunnell et al. (1977b) initially treated five different substrate com- ponents: ethanol-soluble cellulose, lignin, pectin, starch and volatiles. Microflora Activities and Decomposition 321 Several important points emerge from their analyses. The coefficients associated with the volatile and lignin components were occasionally neg- ative, implying negative respiration. Analyses yielding negative coeffici- ents for respiration usually had a disproportionately large number of Dryas substrates. The observations imply 1) that respiration rates of total substrates from any botanical taxon cannot be predicted consistently from independent consideration of the five substrate components men- tioned above, and 2) that some substrate components may have an inhib- itory effect on the respiration rate of other components. In addition to generating inhibitory effects, specific substrate com- ponents also may provide energy for the degradation of more recalcitrant components. For example, the observed rates of weight loss from pure cellulose filter papers placed in the field are lower than rates of loss of cellulose from natural substrates. As well as indicating the failure of the five selected chemical constit- uents to contribute independently to microbial respiration, the analyses of Bunnell et al. (1977b) indicate that the predictabihty of regression equations within temperature-moisture classes is little altered by ignor- ing amounts of pectin, starch and volatiles in the substrate being re- spired. When broader chemical groups (e.g. percent ethanol-soluble and percent ethanol-insoluble) are employed, and Dryas substrates are omit- ted, regression coefficients for the two substrate components are consis- tently different and significantly greater than 0 (a < 0.(X)1). Rates of respiration of ethanol-soluble components are about 5 to 10 times greater, depending on temperature, than the rates associated with other chemical constituents. It is noteworthy that ethanol-soluble compounds make a greater contribution to total respiration at higher temperatures. The discussion of microfioral potential to utilize substrates noted that several fungal isolates show higher temperature optima for utiUz- ation of substances of lower molecular weight than they do for utiliza- tion of the more recalcitrant substances such as cellulose. The analyses associated with the last equation suggest that the phenomenon is general within the mycoflora of the coastal tundra at Barrow and applies to asso- ciated rates of respiration as well as to the physiological potential to de- grade these substrates (Bunnell et al. 1977b). At lower temperatures tun- dra fungi not only maintain their competence to degrade more recalci- trant chemical constituents (Figure 9-1), but also have a greater propor- tion of their respiration associated with these constituents. Microbial Respiration as a Measure of Weight Loss Implicit in the preceding discussion is the assumption that the gener- al form of the response surface for microbial respiration against temper- 322 P. W. Flanagan and F. L. Bunnell ature and moisture is characterized by the physical and chemical nature of the substrate. The model gresp closely mimics this response surface for a variety of substrates (Figure 9-8, Table 9-2). Ignoring hysteresis ef- fects and microbial succession, the assumption can further be made that given the initial characterization for a specific substrate one can predict the instantaneous respiration rate under any temperature and moisture condition during a year. Unfortunately, in situ measures of respiration from the soil are confused by plant and invertebrate activity, and there are no measurements of the pattern of weight loss during a specific an- nual cycle. However, annual measures of weight loss from a variety of substrates under markedly different meteorological conditions are avail- able. These data were collected from different International Tundra Biome research sites. Bunnell et al. (1977a) and Bunnell and Scoullar (198I)have compared measures of total weight loss from different litters with losses due to microbial respiration as simulated by the model. To project simulated weight losses due to microbial respiration they first estimated coefficients a^ through a^ (p. 310-311) from respirometry data collected at the specific site using the non-linear optimization techniques described by Bunnell et al. (1977a). The gresp model was then used to project rates of microbial respiration during the year. Temperature and moisture data used in the projection were those collected from the appropriate site. Computations of weight loss assume that the substrate is 45*Vo carbon and has a respira- tory quotient of I.O. The annual weight losses computed from the simulated daily respi- ration rates are compared with the rates of weight loss as measured by lit- ter bags (Table 9-3). Bunnell and Scoullar (1981) discuss the implications of computed coefficients a, through a^ for each substrate and site. The hypothesis relating microbial respiration to measures of temperature and moisture accounts for 71 to 98% of the variation in rates of respiration measured from a variety of natural substrates and predicts annual loss rates under a wide range of environmental conditions that are 70 to 90% of the measured loss from litter bags. The values for loss of weight seem reasonable, given the additional losses due to leaching and physical reduction. The equation on page 320 explicitly encompasses the additional in- fluence of substrate chemistry. In nature, rates of weight loss from ethanol-soluble compounds and ethanol-insoluble compounds show a ratio of 5.75:1. When the effects of temperature and moisture are incor- porated according to the central equation of gresp, rates of respiration from the two chemical groups show ratios ranging from 5.1:1 to 8.7:1 (Bunnell et al. 1977b). Thus the rates of microbial respiration associated with the two chemical constituents show approximately the same ratio as rates of weight loss from these constituents in the field. If the simpUfying Microflora Activities and Decomposition 323 TABLE 9-3 Annual Weight Losses of Various Litters Measured and Predicted from the Simulated Microbial Respiration Weight loss Simulated as Research (% of initial weight) a percentage area Substrate Measured Simulated of measured Abisko, Sweden Rubus chamaemorus leaves 32' 23.3 73 Barrow, Alaska Dupontia fisheri leaves 15^ 13.4 89 Carex aquatilis leaves 14.6' 13.4 Moor House, Calluna vulgaris shoots 15-20' United Kingdom Calluna vulgaris stems 8' l.V 92 Rubus chamaemorus leaves 36-38' 20.1 81 Note: Abiotic and respirometry data for the research areas were collected during develop- ment of the ABISKO model (Bunnell and Dowding 1974). 'Rosswall (1974). ^Benoit et al. (unpubl.). 'Benoit (unpubl.). 'Heal and French (1974). 'Shoots and stems combined in model. Source: Bunnell et al. (1977b). assumption is made that both chemical groupings contain the same per- cent carbon as the parent material, 45*^0, the respiration rates can be con- verted directly to weight loss. The rate of weight loss of ethanol-soluble compounds is: weight loss hr-' = 3.S GRESP (T, M)l .19 WE where GRESP = central equation of the model (p. 310) with co- efficients appropriate to the specific substrate 1.19 = conversion factor from ^1 CO2 to g substrate WE = weight of ethanol-soluble compound in grams. Following the same approach the rate of weight loss of ethanol-insoluble compounds is: weight loss hr-' = 0.16 GRESPiT,M)l .19 WNE where WNE represents the weight of ethanol-insoluble compounds in grams. The annual rate of weight loss is given by: 365 1-n (l-A:.) / = i 324 P. W. Flanagan and F. L. Bunnell TABLE 9-4 Measured and Simulated Chemical Composition After One Year and Rates of Weight Loss of Erio- phorum angustifolium Standing Dead Material Percentage Measured* Simulated! Chemical composition of the substrate after one year: ethanol-soluble 7 10 ethanol-insoluble 93 90 Loss per year: • ethanol-soluble 49 48 ethanol-insoluble 11 12 Total weight loss 27 31 ♦Measured values use changes in identified age classes. tSimulated values used abiotic data from 1973. Source: Bunnell et al. (1977b). where /:, is the computed daily rate of weight loss due to respiration. The three equations above make up the model decomp. Using meteorological data from the Barrow site, Bunnell et al. (1977b) computed daily values of the function gresp as determined by temperature and moisture and applied these to the chemical-specific coefficients. They obtained annual rates of weight loss of -0.66 and -0.13 for ethanol-soluble and ethanol- insoluble constituents respectively. Measured weight losses from the total litter in the field are -0.69 and -0.12 g g"' yr"'. The rates of weight loss computed from temperature-moisture-chemical-specific rates of respira- tion are thus in very close agreement with measured rates of weight loss. Because the loss rates are chemical-specific (Bunnell et al. 1977b), the model decomp can project not only total substrate weight loss but also chemical composition, which is also very similar to the observed value (Table 9-4). The model DECOMP provides a framework that permits extrapola- tion of laboratory measures of microbial activity to predict total loss of substrate weight and changing composition of substrate. The concepts of decomposition embodied in the model appear supported by combined lab- oratory and field evidence. The reasonably close agreement between sim- ulated and measured values for standing dead material (Table 9-4) sug- gests that at least during the initial period of decomposition many of the changes in substrate weight and chemical composition result from chang- ing rates of microbial respiration which are chemical-specific and inde- pendently influenced by temperature and moisture (Bunnell et al. 1977b). Microflora Activities and Decomposition 325 7 160 _ 1 1 ' ' ' ' ' 1 '"TT E ,,-'^,^ o _ ,-'''^-'^ 1 120 3 - I 80 o - o _ -= 40 o ^^ .- ^ ... - 3 -''' '^ E ~ ^^ 3 '^,^— ' ■"■" " 0 /-i-*;.??*^- ' V , , , , 1 1 I 1 J , 1 20 10 20 10 20 10 Jun Jul Aug Sep FIGURE 9-12. Cumulative COj release over a single growing season, simulated for soil microbes, mi- crobes plus roots and the whole system, and meas- ured by KOH titrations in darkened lysimeters. Root respiration includes that of microbes associated with the rhizosphere. Whole-system CO2 evolution in- cludes the flux from soil microbes, roots, above- ground decomposers and respiration of aboveground plant biomass. (After Bunnell and Scoullar 1975.) For the more advanced stages of decomposition, such as below- ground substrates, the relationships of microbial activities to decomposi- tion are not as well quantified. No data are available on the chemical composition of belowground substrates. Furthermore, measures of mi- crobial activity are confused by the activities of invertebrates and vascu- lar plants. To evaluate the hypothesis concerning decomposition below ground, a broader approach incorporating more ecosystem components but less chemical detail has been employed. The model abisko ii (Bunnell and Scoullar 1975) incorporates the temperature and moisture influences expressed but also simulates contributions from dying roots and respira- tion of vascular plants. Bunnell and Scoullar (1975) have evaluated the model abisko 11 for the Biome research area comparing in situ measures of carbon dioxide evolution with the cumulated totals of simulated respiration from rele- vant components of the system (Figure 9-12). The whole-system respira- tion of Figure 9-12 includes carbon dioxide evolution from soil microor- ganisms, roots, aboveground decomposers, and the growth plus main- tenance respiration of aboveground live material. Simulated microbial respiration follows a pattern very similar to that of the field measures of respiration, but shows a greater depression early in the season. The early season depression is most evident in the 326 P. W. Flanagan and F. L. Bunnell pattern of simulated microbial plus root respiration (Figure 9-12). There are at least two reasons for this early season disparity between simulated and measured values. The soil temperature employed in the model is that at 5 cm depth. Early in the season field temperatures near the surface permit respiration while simulated temperatures at 5 cm depth are too low to allow significant respiration. Because the lower threshold for root respiration is higher than that for microbes, the disparity between simu- lated and measured values is more obvious when roots are considered. The release of carbon dioxide trapped during freeze-up is not simulated but will appear in field measures (Benoit, pers. comm.; Coyne, pers. comm.), and this also contributes to the early spring disparity between field and simulated values. Over the 85-day sample period the accumulated totals of measured respiration and simulated whole system respiration are 159 and 165 g C m"^ respectively. The difference between measured and simulated values over this period is thus 6 g C m"^ or 3.797o of the measured values. A dis- parity of less than 5% is well within the sample error associated with data for root biomass. Thus the simulated dynamics of respiration of mi- crobe, root and other contributing compartments must be assumed real- istic within the accuracy of available data. The proportion of simulated total soil respiration that originates with the roots on any given day varies between 33 and 70%, and lies at the lower end of the range of re- ported values of 50 to 93% (Billings et al. 1978). Errors in different pro- cesses might compensate to produce an invalid sense of accuracy. How- ever, the generally realistic behavior of the model for other tundra areas, including Devon Island, Moor House and Abisko (Bunnell and ScouUar 1981), suggests that microbial respiration for specific substrates in the upper 10 cm of soil follows the relationship expressed by the function GRESP. The dynamics of microbial respiration at depth are much less clear and are confused by anaerobic conditions and poorly understood changes in substrate quality with advancing age. In summary, the concepts of decomposition discussed above appear sufficiently comprehensive to allow laboratory measures to be related ef- fectively to field measures through the vehicle of simulation models. The manner in which microbial respiration responds to temperature, mois- ture and broad chemical groups predicts not only weight loss but chemi- cal composition of aboveground substrates. Upper limits on decay rates are established by chemical composition, but are modified by abiotic var- iables. Respiration is most sensitive to temperature and the microflora has responded by extending its capabilities to grow, respire and utilize substrates at low temperatures. Respiration declines with both increasing and decreasing moisture levels. At low moisture levels degradation and loss of chemicals from standing dead vegetation is temporarily sus- pended; at high moisture levels respiration becomes the province of bac- Microflora Activities and Decomposition 327 teria which cannot degrade compounds of larger molecular weight. De- composition of such compounds remains poorly understood, but it ap- pears that accumulation of organic matter in tundra is more a product of high moisture levels than of low temperatures. Microbes and Turnover of Organic Matter and Nutrients The preceding models concentrate on rates of microbial activity and ignore microbial biomass. The accuracy of the predictions made by these models suggests that concentration on rates of processes is an insightful approach. That in no way obviates attempts to examine the consistency of the measures of microbial biomass with estimated turnover of organic matter. Growth, respiration, production efficiency, and maintenance demands of the microbial biomass in specific substrates can be related to weight loss from that substrate. Using the chemostat model of Marr et al. (1963) it is possible to examine compatibility between field and labora- tory data and to evaluate the influence of microbial activities on turnover and accumulation of organic matter. Several workers (Babiuk and Paul 1970, Gray and Williams 1971, Flanagan and Bunnell 1976) have used this approach in attempting to balance budgets of energy or carbon in a variety of ecosystems. The biomass equation of Marr et al. (1963) is ex- pressed as: idx/dt) + ax = Y(ds/dt) where x = microbial biomass 5 = substrate available for microbial growth Y - yield coefficient, g microbial tissue (g substrate)"' a = specific maintenance rate — g microbial tissue required to maintain 1.0 g microbial tissue for a specific time, e.g. 1.0 hour. When the rate of growth is zero, the above can be expressed: ds/dt = ax/Y'. The yield coefficient Y' is not identical to y because at zero growth there is no actual yield. However, material for maintenance is believed to be utilized at very nearly the same level of efficiency as material assimilated for production of new tissue. Gray and Williams (1971), following the approach of Babiuk and Paul (1970), utilized values for Y' anda of 0.3 g g"' and 0.001 g g' hr"' respectively. Their values for x and a were derived from field data col- 328 P. W. Flanagan and F. L. Bunnell lected at Meathop Wood in Great Britain. Utilizing these values they found that the annual maintenance demands of the microbial biomass in Meathop Wood were such that no organic matter would be available for microbial growth, or for growth or maintenance of any other soil organ- ism. In commenting on the obvious unreality of the situation, Gray and Williams (1971) suggested a number of possible sources of error: 1) the yield coefficient used was too small, 2) the estimate of maintenance re- quirements was inflated, 3) microbial biomass was overestimated, and 4) primary productivity was underestimated. In an analogous study, Flanagan and Bunnell (1976) attempted to minimize the potential errors inherent in the biomass equation by incor- porating laboratory determinations of the specific maintenance rates and yield coefficients of major fungal species in the coastal tundra at Barrow. Values for a and Y ' were calculated as 0.32 x 10"' g g"' hr"' and 0.35 g g"' respectively from data on growth and respiration. The value of Y' ob- tained by Flanagan and Bunnell (1976) is nearly identical to that calcu- lated in studies of organisms from temperate regions, but the value of a is only one-third of that calculated by Marr et al. (1963) in chemostat studies and used by Babiuk and Paul (1970) and Gray and Williams (1971). Using these coefficients Flanagan and Bunnell (1976) estimated that the average standing crop of fungi in the standing dead grew, main- tained and renewed itself at a cost of approximately 10 g substrate m"^ yr"'. This estimate is compatible with the weight loss from standing dead leaves calculated by converting carbon dioxide respired annually by mi- crobes in the tissue to organic matter (Flanagan and Veum 1974). The following discussion is an attempt to 1) quantify the annual maintenance demands of an average standing crop of microorganisms in a unit area of Carex-Oncophorus meadow, 2) estimate the yearly poten- tial for microbial production on the same area, and 3) examine whether microbial potentials for substrate utilization during growth and mainten- ance are compatible with estimates of the amount of organic matter pro- duced and decomposed each year. Microbial Maintenance and Production in the Coastal Tundra at Barrow Direct counts of bacteria from the soils are sparse. Calculations of microbial utilization of substrate are constrained to estimates based pri- marily upon biomass, maintenance demands and yield coefficients of fungi. Utilizing the biomass equation and replacing x by 18.1 g m"^ (the average standing crop of microbial biomass to 7 cm depth at the Carex- Oncophorus meadow), a by 0.32 x 10"' g g"', and Y' by 0.35 g g"', calcu- Microflora Activities and Decomposition 329 TABLE 9-5. Standing Crops and Annual Weight Loss of Five Organic Matter Pools Standing Turnover Organic crop rate Weight loss residue (gdw m"^)' (gg" yr"') 0.300^ (gdw m"^ yr"') Moribund leaves 90 27.0 Standing dead 50 0.075 4.0 Litter 75 0.100 8.0 Dead roots, 0-7 cm 100 0.025' 2.5 Soil organic matter, 0-7 cm 8000 0.020' 180.0 Total 8315 221.5 'Average values from data collected over 4 years at three sites. 'The losses are due primarily to the downward translocation of materials to roots in fall and to leaching in spring by meitwater. 'Estimated from decay rates of roots and cellulose paper in litter bags inserted in the soil, 1972-1977 (inclusive). lated maintenance demands are 0.0166 g m~^ hr"' or approximately 48 g m"^ for the period of microbial activity, which is about 100 days. The annual primary productivity of the Carex-Oncophorus meadow both above and below ground, including phanerogams and cryptogams, is estimated to be approximately 200 g m"^ If 48 g of organic matter m"^ yr"' is necessary to maintain the average standing crop of microorgan- isms, 152 g m'^ yr"' remains to be removed by microbial and invertebrate activity or organic matter will accumulate. Litter-bag estimates of annual rates of decomposition for major substrates in the upper 7 cm of soil (Table 9-5) indicate that 222 g m"^ is decomposed annually. This is approximately one-half the rate for an ungrazed short-grass prairie (Van Dyne et al. 1978). Minimal annual maintenance requirements of microbes (48 g m"^) and invertebrates (7 g m"^ Chapter 1 1) could be met easily and would leave about 167 g m"^ an- nually for production of microbial and invertebrate tissue. Ignoring in- vertebrates, the potential number of microbial generations possible can be estimated using the equation of Gray and Williams (1971): Yis + Nx) = Nx where s = total substrate available to microorganisms (total minus maintenance) TV = number of generations of average microbial biomass Y and x are as in the two previous equations. Allowance is made for recycling of microbial tissue towards its own pro- 330 P. W. Flanagan and F. L. Bunnell duction at a rate governed also by the yield coefficient Y, in the term Y(Nx). Solving for N, the total number of average microbial standing crops possible per year, gives 5.0 and 4.3 generations based on measures of total decay and primary production respectively. Observed rates of turnover for fungi in Carex-Oncophorus meadows were markedly lower, 1.5 to 3.3 times, varying with depth (Table 8-3). In the troughs, however, observed turnover rates were similar to those calculated, 3.6 to 6.2 times, but fungal biomass was significantly lower. Only in the most productive areas do fungi exhibit values comparable to calculations based on labora- tory measurements. In these areas the accumulation of organic matter is lowest (Chapter 7). The observation that litter bags in meadows estimate an annual rate of decomposition greater than annual input from primary production may result from overestimation of the decay rates of untested but apparently more recalcitrant material at depth. The observations do suggest that in the most favorable microhabitats decomposition can ap- proximate primary production. Douglas and Tedrow (1959) observed similar variability in rates of decomposition of organic matter from tun- dra soils. Highest rates of decomposition were observed from the half- bog soil (Dupontia meadows and polygon troughs, Table 1-4). Those rates, 190 g m■^ are comparable to the most rapid rates we have esti- mated and again suggest no accumulation of organic matter in sites most favorable for decomposition. If all 222 g m"^ yr"' of decomposable tissue undergoes dissolution to carbon dioxide, water and minerals through microbial processes, then, based on an average carbon content of 0.45 g g"' organic matter, decom- posers should release carbon dioxide to the atmosphere at an average rate of about 152 mg m'^ hr"'. If annual primary production is matched by decomposition the estimate of average carbon dioxide evolution is 144 mg m"^ hr"'. Root and Microbial Respiration The annual weight loss from belowground substrates plus surface litter is about 190 g m"^ yr"' (Table 9-5). If all of this matter, or an equiv- alent amount minus the average standing crop of microbial biomass below ground (190-17.5 = 172.5 g m"^ yr"'), is consumed annually by microbial metabolism, maintenance and growth, it is possible to calcu- late an average release of CO2 m"^ hr"' from Barrow soils that does not include the CO2 emission of roots. The average carbon concentration of the soil organic matter in the top 10 cm is about 45% so 172.5 g of organic matter could release 78 g C or 286 g CO2. The amount of carbon dioxide released annually (100 days) by microbes would be equivalent to an average respiration rate of 1 19 mg CO2 m"^ hr-'. Microflora Activities and Decomposition 331 The question concerning root contribution to total soil respiration is partially answered for Barrow tundra soils. Billings et al. (1977) esti- mated total soil respiration ranging from 75 to 125 mg CO2 m"^ hr"' in mid-June 1972. During the last week in July 1972, total soil respiration as measured by Billings et al. (1977) ranged from 150 to 300 mg CO2 m"' hr"'. The values of Billings et al. are similar to calculations based solely on microbes, and are somewhat lower than actual field measurements made on root-free soils at Barrow (Benoit, unpubl.). Flanagan and Veum (1974), using respiration data measured in situ in Barrow tundra soils, calculated the average release of CO2 from these soils to be in the range 147 to 235 mg CO2 m'^ hr"'. The calculated rates do not include root respiration. They are about five times less than meas- urements made on temperate forest floors (Witkamp 1966) and about three times less than the average measurements made in tundra soils of the Taimyr peninsula, USSR (Aristoskaya and Parinkina 1972), which include root respiration. The comparison between the Soviet and U.S. data above suggests that microbes may contribute around 30% of total soil respiration. This speculation is in general agreement with rates of respiration simulated by ABISKO 11 (Figure 9-12) which indicate that root respiration may contribute from 33 to 70% of total soil respiration on any given day (Bunnell and Scoullar 1975). This range is somewhat lower than the 50 to 93% contribution by root respiration calculated by Billings et al. (1978). In summary, the observed rates of microbial turnover in the most productive areas very nearly account for the total input of organic mat- ter. Estimated rates of decomposition below ground are compatible with measures of soil respiration, and indicate that some microtopographic units may not be accumulating organic matter. Estimates of carbon diox- ide evolution by decomposition above and below ground range from 2.8 to 5.6 g CO2 m"^ day"' (Flanagan and Veum 1974), in general agreement with estimates of carbon dioxide incorporation by atmospheric flux. Microorganisms and Mineral Nutrient Cycling The distribution of nitrogen and phosphorus in various ecosystem components can be calculated using data on nutrient content of live fungal tissues from Flanagan and Van Cleve (1977), and nitrogen and phosphorus content of soils and decaying matter from Chapter 12 and Flint and Gersper (1974). Flint and Gersper estimated that the wet mea- dow tundra required 6.4 g N m"^ yr"' for plant and animal growth. Ac- cording to the calculations above, gross release of nitrogen by decompo- sition is about 7.5 g m"^ yr"', while the combination of exchangeable and dissolved inorganic nitrogen amounts to 8.0 g m"^ (Figure 9-13). 332 P. W. Flanagan and F. L. Bunnell E 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 fJ3'" Nitrogen xn. ABODE 14 1.2 1.0 0.8 0.6 0.4 0.2 0.0 rS'" Phosphorus n ABODE FIGURE 9-13. Total and available pools and potential release rates of nitrogen and phosphorus in the soil cal- culated for the wet meadow. A) Total amount of nutri- ent to 20 cm. B) Amount available including dissolved and exchangeable (resin-extractable) phosphorus (Flint and Gersper 1974, Barel and Barsdate 1978). C) Amount of nutrient in an average standing crop of mi- crobial tissue to 7 cm. D) Potential annual release from tissue assuming 18.1 g m~^ microbial biomass, 5.0 generations, and no internal microbial mineral re- cycling. E) Potential annual release calculated from the nutrient content of the material undergoing decay. Without including annual input to the system from rain and nitrogen fix- ation (Chapter 7) there appears to be more than adequate nitrogen for plant growth in the coastal tundra at Barrow. The amount of nitrogen immobilized by the average standing crop of microorganisms (0.8 g m"^) is almost insignificant in terms of available nitrogen plus that generated by decomposition. The situation is quite different, however, in the case of phosphorus. The average amount of available phosphorus is small, as is the amount of phosphorus released annually in decay processes. The phosphorus im- mobilized in an average standing crop of microbial tissue is greater than the sum of labile pool plus the annual input via decay (Figure 9-13), sug- gesting a profound influence of microorganisms on availability of phos- phorus in the system. Because tundra microorganisms, except for their resistant propagules, die off each year, they 1) release a relatively large quantity of phosphorus to the soil annually, and 2) are taking phos- phorus from the system at levels that are greater than the normal size of Microflora Activities and Decomposition 333 the available pool. These observations indicate a pool of available phos- phorus which turns over very rapidly and/or possible limitation to plant metabolism by microbial competition for and immobilization of soil phosphorus. SUMMARY Field and laboratory measures have been combined to provide an overall picture of decomposition. The ability of the tundra microflora to utilize substrates varies spatially, with aerobic decomposers showing a marked increase in capacity to degrade cellulose and phenols in the soil, as compared to the phyllosphere. This gradient in potential utilization is accompanied by an increase in anaerobes, and a marked decrease in both zymogenous forms and general microbial biomass. Fungi are better able to utilize cellulose and phenolic substrates than are bacteria. Unfortu- nately, the enzyme capabilities of the anaerobic microflora remain un- known, but no anaerobic decomposers of aromatic compounds are known from other areas. Respiration rates of the microflora are governed in a predictable fashion by temperature, moisture, and substrate chemistry. Respiration rates are shown to be the dominant influence on weight loss from sub- strates. Substrate chemistry establishes a potential maximum rate which is modified by abiotic variables. Ethanol-soluble compounds generally are respired 5 to 7 times more rapidly than non-ethanol-soluble com- pounds, but some substrates (e.g. Dryas leaves) apparently contain sub- stances inhibitory to microbial respiration. Both bacteria and fungi show adaptations to cold. Microbial respir- ation continues to - 7.5 °C and fungal growth is still positive at 0°C, in- dicating greater levels of activity at low temperatures than are observed in vascular plants. Many microorganisms in colder strata of the environ- ment show linear rather than exponential responses with increasing tem- perature, suggesting adaptation to cold through more rapid response to small increases in temperature. Cold-adapted microorganisms, especially fungi, increase in numbers from the phyllosphere into soils, while in the upper regions of the soil microbial populations display a wider range of temperature optima and mesophilic forms are more prominent. Among the fungi, psychrophiles, thermotolerant psychrophiles and cold-tolerant mesophiles retain the capacity to utilize structural plant carbohydrates at temperatures below 0°C, while aerobic bacteria are largely restricted to non-structural plant components at 0°C. As bacteria can use the pro- ducts of fungal decomposition of large molecules, it is possible that co- evolution has permitted the development of different enzymatic re- sponses to low temperatures. 334 P. W. Flanagan and F. L. Bunnell Moisture levels above 20% dry weight are necessary to initiate microbial metabolism, while levels much above 400% dry weight attenu- ate microbial activity. Oxygen, carbon dioxide, and temperature rela- tions interact with moisture levels to obscure definition of optimal mois- ture levels for decomposition. Shifting bacterial: fungal ratios along oxy- gen gradients indicate that oxygen availability and/or moisture alters the numbers and character of participants in decomposition. Large numbers of facultative anaerobic bacteria in litter and soil reflect a commonly oc- curring niche. Although present, obligate anaerobes are major contribu- tors to decomposition only in heated soils. Perhaps because of microbial adaptations to low temperatures, or- ganic matter does not appear to be accumulating in some microtopo- graphic units. Only the fungi are capable of degrading the larger com- pounds, particularly at low temperatures. The fungi, however, are re- stricted by high levels of moisture to the upper 7 to 10 cm of soil, and or- ganic matter may accumulate at depth. The nutrient dynamics within the soil also suggest that phosphorus, but not nitrogen, immobilized within the standing crop of microbial tissue may be a factor limiting nutrient availability to vascular plants. 10 The Herbivore-Based Trophic System G. O. Batzli, R. G. White, S. F. Maclean, Jr., F. A. Pitelka, and B. D. Collier INTRODUCTION The tundra is well known for its conspicuous and abundant animal populations. Indeed, tundra may be better characterized by caribou, wolves, lemmings, snowy owls, ptarmigan and hordes of flies than by any other feature, at least in popular literature. The next two chapters consider the composition and organization of animal communities, and their participation in the energy and nutrient dynamics of the coastal tun- dra ecosystem. Ultimately, all heterotrophic activity, animal and microbial, de- pends upon the energy and nutrients fixed by green plants in net primary production. The amount of net primary production (Chapter 3) sets a limit upon the abundance and production of heterotrophic organisms. Two more or less distinct trophic systems based upon this net primary production may be recognized in virtually all ecosystems — a herbivore- based system that begins with the consumption of living autotroph tis- sue, and a detritus-based system that begins with the consumption of dead organic matter (Figure 10-1) (BatzH 1974, Heal and MacLean 1975). The distinction corresponds broadly to an aboveground and below- ground division (perhaps reflecting a paucity of information on below- ground herbivory). The two trophic systems may converge to some ex- tent, particularly at the top carnivore level, and a single animal popula- tion may function in both trophic systems; this is a categorization of tro- phic functions rather than animals. Several important conceptual differences distinguish the two trophic systems. The herbivore-based system begins with the consumption of liv- ing plant tissue, and thus impacts directly upon plant production, growth, and reproduction. Thus herbivores (or, indirectly, carnivores preying upon herbivores) modify the rate of input of chemical energy 335 336 G. O. Batzli et al. r R r R Carnivore 1 Carnivore 1 1 r R '' r' /-' =1 2 Consumers [Carnivore V- Carnivore Microbivore[ . 1 R ' J. R 1° Consumers 1 Herbivore \- Animal Saprovore Microbial Decomposer — Gross Primary _ 1 r' =1 1 Input Production Primary PrnHi more Dead Organic 1 "^^■•' 1 ■' Herbivore- based trophic system Detritus-based trophic system FIGURE 10-1. A generalized trophic structure for terrestrial ecosystems, showing the distinction between herbivore- based and detritus-based trophic systems. Arrows represent the flow of energy and materials; R represents respiratory loss of energy. (After Heal and Mac Lean 1975.) (fixed carbon) into the ecosystem. The detritus-based system is based upon the consumption of dead organic matter. Saprovores and microor- ganisms influence the ecosystem through their control of the rate of decomposition and cychng of mineral nutrients. There is ample reason to believe that, in tundra, this is particularly important (Chapter 5). Other distinctions between the herbivore- and detritus-based trophic systems are discussed by Heal and MacLean (1975). Herbivore-based food chains in arctic regions contain relatively few taxa. Whole groups of invertebrates that are common in grasslands at lower latitudes, e.g. insects and mollusks, have few representatives in the Arctic (MacLean 1975a). The most abundant herbivores are homeo- therms, probably because they can maintain high rates of activity and growth at low temperatures. Herbivorous birds, especially ptarmigan {Lagopus spp.) and geese {Anser albifrons, Branta canadensis and Chen hyperborea), use the North American tundra as a breeding ground during summer, but their occurrence and impact appear to be patchy. While avian herbivores gen- erally migrate south in late summer, some ptarmigan do overwinter on inland tundra where they consume mostly willow buds and twigs (West and Meng 1966). The Herbivore-Based Trophic System 337 Some mammalian herbivores also occur sporadically on the tundra. Ground squirrels {Spennophilus parryii) may reach impressive densities along river banks and beach ridges, where substrate suitable for con- structing their winter hibernacula can be found. Hares {Lepus othus, = timidus, and L. arcticus) rarely reach significant numbers on tundra, though summer herds of 100-150 occasionally appear (Batzli 1975a). The two remaining groups of mammalian herbivores, the microtine rodents (Lemmus, Dicrostonyx and Microtus) and the ungulates {Rangifer and Ovibos), frequently reach high densities over wide areas of tundra, and represent most of the biomass in the first link of the herbivore chain in tundras. Differences in body size and mobility of the microtine rodents and ungulates lead to different tactics for dealing with the severe arctic cli- mate (Batzli et al. 1981). The large ungulates have more insulation and can withstand lower temperatures (Scholander et al. 1950). When winter snow conditions or temperatures become intolerable, they can travel long distances to more favorable habitats. In contrast, the small microtines have poor insulation and must rely upon increased metabolism to main- tain body temperature. Because they lack the ability to move long dis- tances, they must select or create favorable microhabitats in order to sur- vive severe winter conditions. Large body size gives ungulates the oppor- tunity for a long life span and iterative reproduction. However, two other characteristics associated with large body size — a longer develop- mental period and an increased parental investment in each offspring (Pianka 1970) — reduce the ability of ungulates to respond rapidly to fa- vorable conditions by reproducing, and populations remain relatively stable. Small microtines, by comparison, suffer greater mortality in se- vere environments, but their shorter developmental time and greater fe- cundity allow them to respond rapidly to favorable environmental condi- tions. The following exposition describes and compares the most impor- tant components of the herbivore-based trophic systems at Barrow, dom- inated by microtine rodents, and at Prudhoe Bay, dominated by ungulates. HERBIVORY AT BARROW— LEMMINGS Introduction In the coastal tundra at Barrow the brown lemming {Lemmus siber- icus, =trimucronatus) is the dominant herbivore. The density of trap- pable animals (post-weanlings) may reach 225 ha"'. Collared lemmings {Dicrostonyx torquatus, = groenlandicus) are usually scarce, about 0.1 per hectare, though densities on elevated ground may be higher. They 338 G. O. Batzli et al. have reached substantial numbers only once in the last 20 years, 27 ha"' in 1971 (Figure 10-2). No other vertebrate herbivores regularly inhabit the Barrow peninsula, but a few caribou (Rangifer tarandus) and ptar- migan {Lagopus lagopus) visit occasionally. A discussion of herbivory can therefore center on a single species: the brown lemming. Population Dynamics and Demography Changes in the lemming population have been monitored for 20 summers, from 1955 to 1974. During that time densities have fluctuated between peaks of up to 225 trappable lemmings ha"' and lows of 0.02 ha"', with three to six years elapsing between peaks (Figure 10-2). Although reliable estimates of population size were not made before 1955, high densities were also observed in 1946, 1949 and 1953 (Pitelka 1957b). These fluctuations have traditionally been called cycles, largely because of their great amplitude (3 or more orders of magnitude), even though all aspects of successive cycles are not alike. 240 200 160 ,; 120 o 80 40 l\ • Lemmus oDicrostonyx ; ?'Lj ? it I', ■-I — ^ i#^ V OO 00 ^^. 551 1601 1651 1701 174 Year FIGURE 10-2. Estimated lemming densities averaged for all habitats in the coastal tundra at Barrow for a 20-year period. The question marks indicate numbers based upon observations other than trapping. The Herbivore-Based Trophic System 339 Mean densities for the entire tundra at Barrow are mainly useful for considering annual trends (Figure 10-2). The densities were calculated by calibrating results from extensive snap-trapping done at seven sites in five habitat types (Pitelka 1973). Local densities may depart markedly from overall densities, but the general trends from year to year were simi- lar in all habitats. A description of the sequence of events during a standard cycle can begin with the development of a high population. During the pre-high winter, lemmings reproduce in nests constructed out of dead grass and sedges and placed at the base of the snowpack. The population grows rapidly and reaches a peak in late spring. Breeding ceases during May, so few young are still in the nest during snowmelt, but juveniles continue to be recruited into the trappable population until early June. Before snow- melt there may be signs of stress. Many lemmings burrow to the surface and wander about, sometimes dying (Rausch 1950, Thompson 1955b). During snowmelt massive cHpping of graminoids and disruption of moss and lichen carpets are revealed, and lemmings scurry everywhere. Large numbers of predators attack the exposed lemmings. Particularly promi- nent are pomarine jaegers {Stercorarius pomarinus), snowy owls {Nyctea scandiaca) and least weasels {Mustela nivalis). During the summer lem- ming survival declines, and the population crashes to a low level, where it remains for one to three years. While this may be the general scenario, careful analysis of trapline data indicates that each cycle has peculiarities of its own (Pitelka 1973). In 1956, 1963 and 1969 populations increased under the snow, but de- clined before it was possible to measure maximum densities. In 1956, considered a peak year, an early snowmelt began in May, exposing the lemmings to avian predators. In 1963 and 1969 predation under the snow by weasels was unusually heavy (Pitelka 1973, MacLean et al. 1974), and normal peak densities of more than 100 ha'' were never reached. The highest recorded density occurred during the 1960 peak, which lasted through the summer despite heavy predation and widespread destruction of habitat. In contrast, during 1965 the population declined to less than 1 % of its initial density during the course of the summer. The decline fol- lowing the population peak in 1971 was not as great, and, unlike all other post-high summers, in the summer of 1972 lemmings were present in moderate numbers. The population did not reach its usual low density of less than 0.5 ha"' until 1974. During the pre-high summer of 1964 densi- ties reached unusually high levels, but the population merely doubled during winter to produce the 1965 high. The pre-high summer of 1970 represents the other extreme: densities remained low, and the population increased by a factor of 250 during the ensuing winter. The combination of events since 1965, which has been especially peculiar compared with previous cycles, has been described in detail by Pitelka (1973). 340 G. O. Batzli et al. Year of Cycle FIGURE 10-3. The percentage of female lem- mings pregnant in each age class during the course of a cycle. Sample sizes range from 10 to 746 and include all females collected by Pitelka (1973) during 1952-65. Data were collected for only one winter (1962-63). The shaded bars sepa- rate summer and winter and indicate times when mean air temperatures are near 0°C. (After Osborn 1975.) Demographic changes accompany these population fluctuations. Suppression of breeding, indicated by a low incidence of pregnancies, oc- curs in May just before snowmelt and in September during freeze-up (Figure 10-3). When lemmings do breed at these times, nests lie exposed on the surface because burrows are filled with ice or water or the snow cover is not well developed. Summer breeding appears to decline regu- larly in late August, although less so in pre-high years, but the resump- tion of breeding in early summer varies, depending upon temperature and the timing of snowmelt (Mullen 1968). During these breeding pauses the population structure shifts toward the older age classes, and density declines. Once the summer season begins, the population reproduces maximally — nearly every female is pregnant by mid-July — and the popu- lation structure shifts toward the younger age classes. If survival is high, the population increases rapidly. In general older females become preg- nant more frequently than younger ones. The breeding intensity of adults varies little from summer to summer, but juveniles and subadults breed much less during the summer of a high population (Figure 10-3). Little is known about the winter breeding season except that it lasts The Herbivore-Based Trophic System 341 to 6 c o 0 Jan I Feb |Mar | Apr |May | Jun | Jul | Aug | Sep | Oct | Nov | Dec FIGURE 10-4. Mean litter sizes for different age classes of lemmings throughout the year for the period 1952-65. Vertical bars represent ± 1 SE. (After Osborn 1975; based upon data of Pit el ka 1974.) from November through April. Far fewer females are pregnant in mid- winter than in summer. Breeding intensity varies more during winter than summer, probably depending upon the snowpack, thermal condi- tions and the availability of food, but there is little direct evidence for this. Lack of winter nests, lack of placental scars in adult females in spring and decreases in the population indicate that during some winters Httle, if any, reproduction occurs (MacLean et al. 1974). Litter sizes differed among age classes of lemmings and among sea- sons (Figure 10-4). Older females have larger litters during summer; the mean litter size is eight for adults, seven for subadults and six for juven- iles. During midwinter mean Htter size declines to three. Although sta- tistical analysis (ANOVA, p < 0.05) indicated significant, but minor, dif- ferences in average litter sizes from summer to summer, they did not ap- pear to be related to the phase of the cycle (Osborn 1975). Dramatic changes in survival rates from summer to summer do oc- cur, from 70'Vo per 28 days to 10% per 28 days for adults in July, but our knowledge of survival rates throughout a cycle is scanty. Changes in sex ratio indicate that survival of males tends to be lower than survival of fe- males during the summer and higher during the winter, possibly as a re- sult of differential predation. Osborn (1975) developed a computer simu- lation model that allows survival rates and density of sucklings, which are not trappable, to be estimated by a trial and error procedure, given information on age-specific reproductive rates and on population struc- ture. Using the reproductive rates and litter sizes discussed above, to- gether with field observations of age structure and population size during summer, he estimated density of sucklings and survival rates for each 342 G. O. Batzli et al. Males Females (/> >> o T3 00 CVJ > 80 40 OL 80 40 0 80 40 ^^'^ Adults ■^^iJ^ Subaduits > BOn 8 - Juveniles \J^- 40- Sucklings 0*1 I I I I I I I 1 1 I I I I I I I I I I I I I I I I I I I lYlTl I I loop Juveniles F A J A 0 D Low F A J A 0 D Prehigh F A J A 0 D High Year of Cycle FIGURE 10-5. The density and demography of brown lemmings through a standard three-year cycle calculated by a population model. (After Osborn 1975.) The Herbivore-Based Trophic System 343 age, sex and reproductive class for a standard cycle based upon data from 1961-65 (Figure 10-5). In the model, survival rates for October through April were not varied with time and should be viewed as average values over the winter. One of the most striking results of these simulations was the small increase in survival of lemmings required to build a peak population (Figure 10-5). Sucklings accounted for about half of the population dur- ing the pre-high summer and one-quarter of the peak population, During this time, reproductive rates were no better than the previous year, but suckling and female survival improved slightly. Even with high reproduc- tive rates survival rates of suckhngs and juveniles during the pre-high summer needed to be relatively high for the population to increase as ob- served in the field. During the population crash the survival rates of young animals remained very low so that at times the population became almost completely adult. Survival rates for sucklings are based only on litters whose mother also survives; when a nursing female dies her litter also dies. Thus, improvement of female survival has a double effect on population growth. Habitat Use Lemmings do not make equal use of various habitat types, so any analysis of herbivory must consider spatial patterns of habitat utiliza- tion. We use two scales in our analysis, mesotopographic units and mi- crotopographic units (Figure 10-6). Snap-trap lines run by Pitelka for 19 years in the Barrow area pro- vided information on habitat utilization. These lines were divided into five habitat types, which can be arranged along a moisture gradient and in Ez.v n.iD Ez.s yi Vegetation Type i2,y,2i 211 y,3n in ni.iz Small, low-centered polygon (Polygon grid) Meadow (Meodow grid) Meodow Large, low-centered polygons ond ponds (Pond grid) Higti-centered polygons (Upland grid) FIGURE 10-6. The relative utilization by lemmings of habitat types on the Barrow research area. The numerals above the topographic profile indicate vegetation types (Table 1-4). 344 G. O. Batzli et al. correlated with the vegetation types given in Chapter 1 (Table 1-4): 1. Ridges: well-drained raised areas with Salix heath vegetation (I, II). 2. High-centered polygons: polygons with well-drained centers and with wet troughs and ponds between them (I, III, V, VI). 3. Low-centered polygons: polygons with mesic centers and with wet troughs between them (II, III, IV, V, VI). 4. Graminoid flats: wet meadow with mixed graminoid vegetation (V, VI). 5. Carex flats: wet meadow dominated by Carex (VI). Although densities varied, the pattern of habitat use was fairly con- sistent from year to year. In early summer, well-drained areas dominated by high-centered polygons were the most heavily used, and twice as many lemmings were captured there as in any other habitat. By late summer the use of such areas decreased markedly, while use of areas dominated by low-centered polygons and of low-lying meadows increased. This pattern reflects the fact that low habitats are flooded in early summer but be- come drier as the summer progresses. The extremes of the moisture gra- dient, ridge and Carex meadow, received substantial use only when pop- ulation densities were high. Collared lemmings, when present, usually were found in ridge or high-centered polygon habitats (Figure 10-6). Density estimates and signs of activity on four 2.25-ha live-trapping grids in the immediate vicinity of the Biome research area gave a slightly different picture of habitat use. During winter nest densities were greater in high-centered polygons than other habitats, but population densities at snowmelt and winter clipping of vegetation were greater in low- centered polygons (Table 10-1). Clipping rates indicate the amount of winter foraging, while nests may reflect the intensity of reproduction. All indicators of activity support the notion that the polygon and pond habi- tats, which had relatively uniform vegetation heavily dominated by Carex, were used least by lemmings. Summer population densities and the rates of clipping indicate that the low-centered polygons and mea- dows were used most heavily, particularly in the summer of 1972 and the winter of 1973. At lower densities, in the summer of 1973 and the winter and summer of 1974, use of high-centered and low-centered polygons was more similar. Use of the meadow depended upon moisture conditions. The sum- mer of 1972 was relatively dry, so lemmings moved into the lower, wetter portions of the grid as standing water receded, and the overwintering population was similar to that on low-centered polygons (Table 10-1). The summer of 1973 had late rains, however, and much of the meadow was flooded at freeze-up. As a result, the overwintering population dur- ing 1974 was only a third of that on low-centered polygons. The Herbivore-Based Trophic System 345 TABLE 10-1 Summary of Indicators of Brown Lemming Activity in Four Habitats at Barrow 1972 1973 1974 Winter nest density (no. ha') Higii-centered polygons — 24.9 18.5 Low-centered polygons — 18.2 12.9 Meadow — 17.2 4.2 Polygons and ponds — 2.7 0.4 Summer population density (no. ha-') High-centered polygons 3.4-14.7 3.1-4.0 0.3-0.6 Low-centered polygons 12.9-46.1 3.1-4.9 0.3-0.6 Meadow — 2.5-3.1 0.9-1.2 Polygons and ponds 1.6-7.7 0.3-1.5 0.0-0.6 Percentage of graminoid tillers clipped High-centered polygons 14.8 12.3 2.4 Low-centered polygons — 25.1 6.7 Meadow — 24.3 1.9 Polygons and ponds — 5.4 0.3 Note: Densities are seasonal extremes. Habitat types are illustrated in Figure 10-6. These observations support several generaHzations. First, brown lemming activity, both summer and winter, tends to be concentrated in polygonal terrain, which has a mixture of relatively dry and wet habitat. Second, low, wet areas with vegetation heavily dominated by Carex and dry ridges with Salix heath are the areas least utilized by brown lem- mings. Third, the use of low-lying meadows with mixed graminoids varies depending on seasonal and annual moisture conditions. Lemming activity is neither randomly nor uniformly dispersed with- in the larger topographic units discussed above, and the use of local mi- crotopographic units must also be considered. Habitats with well- developed high-centered and low-centered polygons have the greatest mi- crotopographic relief and support the greatest winter nesting. Ninety- three percent of 139 nests examined in 1974 were located in polygon troughs. Winter clipping of vegetation is also concentrated in troughs, where the density of shoots is higher, but patches of clipping occur in all microtopographic units. In 1973 we found clipping percentages of 25 to 50% in troughs, where graminoid shoot densities were about 3000 m'^ 15% on rims with 2000 shoots m'\ and 5% in basins with 1000 to 1500 shoots m'^ Summer activity patterns have been more intensively studied than winter patterns. Using techniques of radio tracking. Banks et al. (1975) showed that individual lemmings may move over 1 km day"'; however, most lemmings live in home ranges of highly variable size. Females tended 346 G. O. Batzli et al. TABLE 10-2 Home Ranges (Mean ±1 SE in Hectares) of Lemmings on Low-centered Polygons 1972 1973 Males 1.33±0.28 (12) 0.88 ±0.28 (13) Females 0.68 ±0.37 (6) 0.41 ±0.31 (11) Note: Sample sizes are given in parentheses. Source: Banks et al. (1975). to have smaller home ranges than males (Table 10-2), but the variability is such that differences were not statistically significant. Differences also occurred from year to year, but again variability was high. Analysis of the frequency of movements of 5 m or more indicated that males moved more than females and that more movement occurred in 1972 than in 1973 (ANOVA,/7< 0.01). The greater movement of males probably results from the mating system of lemmings, viz. promiscuous polygamy. Males can increase their fitness by touring the habitat in search of females, whereas females restrict their movements, usually moving between foraging sites and nest- ing burrows. A combination of increased energy requirements and vul- nerability to predation when moving probably limits male movements. The greater movement of lemmings in 1972 than in 1973 was correlated with higher population density (Table 10-1), hence greater social interac- tion, and with a lower rate of predation in 1972. Summer activity in polygonized terrain takes place primarily in the troughs. Lemming burrows are concentrated on the sides of troughs above the water line or on the sides of high-centered polygons. Summer nests are usually located in these burrows or under elevated patches of moss {Sphagnum spp.). Runways follow troughs, particularly the frost cracks associated with them. Foraging activity is concentrated near the burrows and along the sides of runways. About 95 to 10097o of graminoid shoots in the immediate vicinity of burrows are repeatedly clipped while the burrow remains occupied. Away from the burrows, clipping is patchy, and intensity decreases as distance from the runways increases. Cheslak (pers. comm.) found only from 0 to 30% of shoots clipped at a distance of 0.5 m from runways when lemming densities were moderate, about 10 ha"' . Maps of the locations of radio-tagged lemmings show pat- terns that match the patterns of polygon troughs (Banks, pers. comm.). In nonpolygonized terrain the association with microtopographic features is not so clear, and lemming activity appears to be located in more randomly distributed patches, but, again, clipping is concentrated near runways. The Herbivore-Based Trophic System 347 Although collared lemmings are not abundant on the coastal tundra at Barrow, they are more common than brown lemmings at many other tundra sites (Bee and Hall 1956, Krebs 1964, Fuller et al. 1975, Batzli and Jung 1980), and they present an interesting contrast. Collared lemmings generally prefer higher, drier habitats than brown lemmings, particularly areas where dicotyledonous plants are common. They excavate more elaborate burrow systems than brown lemmings and do not use runways, which are not required in habitats without dense graminoid growth. Nutrition and Energetics Diet The food habits of lemmings, as of most herbivores, vary with sea- son and habitat. In general, however, collared lemmings and brown lem- mings specialize on different food types. Collared lemmings take primar- ily dicotyledons supplemented by graminoids, whereas brown lemmings take primarily graminoids supplemented by mosses. Salix appears to be the most important dietary item of collared lemmings in summer (40 to 50%) and may be even more important in winter. Dicrostonyx feces from winter nests contain large amounts of willow leaves. Dietary specializa- tion reflects more than differences in habitat preference because dietary compositions remain distinct even at sites where both species occur (BatzH 1975a). We have examined the food habits of brown lemmings in some de- tail. Significant seasonal changes in diet occur, from about 80 -20 -10 a. FIGURE 10-7. The relationship of energetic require- ments of non-reproducing lemmings (solid lines) and percentage of energy used for thermoregulation (dashed lines) to ambient temperatures. Subscripts give the body weight of lemmings in grams. which assimilated energy can be transformed into tissue. Growth rates of lemmings have been summarized by BatzH (1975a) and allow for the de- velopment of sexual dimorphism in size beginning as subadults, 30 to 60 days old. In order to calculate the total energy used for growth, 25% of the energy stored in new tissue must be added, given an efficiency of tis- sue growth of 0.80 (Blaxter 1967). Assuming that growth is independent of ambient temperature, growth adds a constant energy increment to the average daily metabolic rate. Energy requirements for maintenance and growth as a function of ambient temperature differ markedly for juvenile and adult lemmings (Figure 10-7). The average daily metabolic rate increases significantly with declining ambient temperature; for the 20-g juvenile lemming it in- creases nearly 3-fold over the temperature range -1- 17 ° to -25 °C, the an- nual range of temperature usually encountered by a lemming. The average daily metabolic rate equation separates the energy cost 350 G. O. Batzli et al. of thermoregulation, 1.87 kJ °C"', from the cost of maintenance. The cost of thermoregulation expressed as percent of total energy require- ments, average daily metabolic rate plus growth, is zero at +17°C, which is the lower limit of thermoneutrahty, compared with 36"7o of the total for the 80-g lemming at -25 °C and 60% for the 20-g lemming at -25 °C. Energy required for reproduction includes that invested in growth of fetuses and production of milk for sucklings. Fetuses grow from 0 to 3.3 g over the 21 -day gestation period, an average growth rate of 0.25 g day"'. Given an average summer litter of seven (Figure 10-4), the repro- ductive female must support an average fetus growth of 1.75 g day"'. Ac- tually, the growth is concentrated in the latter phases of gestation. After a tissue growth efficiency of 0.80 is applied, the cost of pregnancy is about 75% greater than that for growth of the 20-g lemming (Figure 10-7). The cost of suckling growth must also be supported by the breeding female. Applying a growth rate of 0.8 g day"' to a litter of seven gives a value of 5.6 g of suckHng growth per day. Since 1 g suckling live weight has an energy content of 4. 19 kJ, this is equal to 23.5 kJ. This must be di- vided by 0.3, the value for efficiency of conversion of milk to suckHng tissue, and by 0.7, the value for efficiency of milk production by the mother (Brody 1945, Hashizume et al. 1965). The energy requirement is therefore 1 13 kJ day"', equal to the average daily metabolic rate of a 40-g lemming at 5 °C. Thus, lactation plays an immense role in the energetics of lemmings. Securing this additional energy requires more activity by reproducing females, which further increases energy demand. Lemmings usually are able to satisfy this demand during summer, but during winter, when the cost of thermoregulation is high and forage is sparse, litter size declines to three (Figure 10-4). Reproduction would not be possible at all during winter without the construction of nests. The value of the nest to the lemming was explored by MacLean and Thomsen (pers. comm.) using a heat flow model. The model regards the lemming as a homeothermic body of temperature 71 and radius b, proportional to the cube root of body mass. The lemming must produce heat at a rate q that is equal to the heat flow from the warm lemming to the cold surroundings. The nest forms an insulating layer of inner radius b and outer radius a around the lemming. The model shows that an equiUbrium heat distribution is rapidly es- tablished in the inner layer of the nest around the lemming. At this time g = 4nkiT,-T^)[ab/(a-b)] where k is the thermal conductivity of the nest material in J s"' cm"' °C"' and Ta is the temperature of the snow around the nest. Heat flow is determined by the temperature gradient, the radius of the lemming and The Herbivore-Based Trophic System 351 210 TO c a) a. X UJ 0) c UJ 2 4 6 Number of Pups FIGURE 10-8. The simu- lated daily energy expend- iture by a reproducing female lemming with a winter nest in relation to litter size, weight of pups and ambient temperature T. (MacLean and Thom- sen, pers. comm.) the radius of the nest. Under winter conditions heat flow from the nest is considerably less than the average daily metabolic requirement without a nest. For a 60-g lemming at an air temperature of -30 °C the average daily metabolic rate maintained over the entire day gives 199 kJ of energy used. The time that can be spent in a nest depends upon energy reserves, which can be estimated from the stomach capacity equation discussed below. Assuming 30-minute bouts of foraging with metaboHsm at the average daily rate, interspersed between bouts of nest use, total energy use by a 60-g lemming with a nest of 12-cm radius is 129 kJ day"'. A large nest is crucial for winter reproduction. It reduces the energy cost to the female and prevents rapid cooling and death of the sucklings prior to the development of homeothermy at 10 to 12 days of age. The physiological processes involved in reproduction are energy-demanding, especially lactation. However, much of the heat from respiration by the female goes into the nest, and thus contributes to homeothermy. Before the sucklings develop homeothermy the major costs of reproduction come from 1) the increased radius of the nest contents, the female plus young, which leads to greater heat flow from the nest; 2) the need to re- warm cooled sucklings following a period of absence; and 3) the growth of the sucklings. The estimated energy expenditure of a lemming sup- porting a maximum reproductive load, six 11-g heterothermic sucklings at an ambient temperature of -30 °C, is 186 kJ (Figure 10-8), or 144% of s: <3 l^ <3 5 5 C3 ^ ^ o to O ••^ -wa (i) C50 ^ <^ c tlj "Q s: ST" Q S ^ 5 5 o ^ k: C =s a q h. c "q o ■•^ •■■** tn *».* ^ t3 s O C5, S^ o Q, ,^ ^ f-^ o ■**# s: 5j •S2 § w o <3 >. Cl, 5^ 1 o o I d s 1^ 00 r- ^ S o o d 1 O 1 oo 1^ p- ^ »N 1 1 1 o 1 O O) 0^ d in 8 d (N — — 0\ (N — Cv o I s d — _ _2 s V3 u > a a I/: U u> k. '5 c« TJ C 3 O 3 CB W5 L- V J= o 3 8 c 2 c 2 to _c V k. (8 t/i u 3 ca rN 0\ — — — (N O O O odd w^ w^ O _ -r ■* >C ^ 3 o o o o c ^ s nl Q < >> CO „ to to e 00 c ■a 5 -^ 5 ^ O ^' O CO -o ^ c - 7 O 7 CO 'w C j: CO ^ to = 7 C O CO O m QQ c _ _ = < E U Of O \ c "" 3 H c o ■2 S & ^ O 0^ Ou 'C c c o Q c J^ c 3 c o c « '5 « 1) CO o JJ oo >• o T3 c« CO E =* _ o o E — ^ 2 (1> ■a •n c m CO N) UU CO • 03 CO a> > > 00 JC C CO 5 2 o z 352 The Herbivore-Based Trophic System 353 the energy used by a nonreproductive female with a similar nest of 12-cm outer radius. The aditional energy cost of reproduction is rather low compared to earlier calculations because nonreproductive lemmings have a high rate of energy use at such low temperatures, even with a nest. Total population energy demand for a high year (September of a pre-high year through August of a high year) has been calculated by sum- ming energy requirements for the individuals of each category present each month. Monthly changes in density, population structure and re- productive intensity were estimated for a standard cycle in which trap- pable densities reached 150 ha"' (Batzli 1975a). The low population was assumed to have a similar structure, but average densities for the whole year were 400 times less than for the high population (Table 10-3). Grodziriski and Wunder (1975) reported that production, the sum of all energy deposited in new tissue, averaged 2.3% of respiration for ro- dent populations in general, a value slightly higher than the 2% predicted by Turner (1970) for vertebrate homeotherms. When rising to a peak, production for a population of Lemmus at Barrow was 5% of respira- tion. This value is the highest known for a homeotherm and occurs des- pite the high energy requirements of thermoregulation. The high produc- tion results from nearly year-round reproduction and the high popula- tion turnover rate of Lemmus. Values calculated for Dicrostonyx are more similar to those for other small rodents. Digestion and Ingestion In order to calculate ingestion rates for the population, we must know the mean com.position of the natural diet and the digestibility of natural forage as well as energetic demand. The information on the food habits of lemmings presented above has been averaged across habitats, and overall summer and winter digestibilities calculated (Figure 10-18). A mean digestive efficiency of 33 Vo for energy is derived from these values and applied to assimilation to give an estimate of ingestion. The digestive efficiency of Lemmus is strikingly low, much lower than that of Dicrostonyx and most other herbivorous mammals (Batzli and Cole 1979), and dramatically elevates the ingestion rate. The materi- al within the plants may be separated into structural carbohydrate, com- posed of cellulose, hemicellulose and lignin, and nonstructural carbohy- drate. Nonstructural carbohydrate is material contained in the cell cyto- plasm and should be much more easily digested than structural carbohy- drate. Total nonstructural carbohydrate (TNC) in graminoid shoots var- ied between 30 and 40% of the biomass for most of the growing season on moist meadows (Chapter 5, Figure 5-5). Lipid concentration found in shoots varied from 5 to 15%. Therefore, the observed 36% digestibility of graminoids can be accounted for by digestion of TNC and lipids. Although digestibility of tundra graminoids by brown lemmings is 354 G. O. Batzli et al. low compared to the 45% to 55% digestibility of temperate grasses by other microtine rodents (Batzli and Cole 1979), tundra graminoids are higher in total nonstructural carbohydrate and lipids than are temperate graminoids (Chapter 5). Thus, the higher digestibilities achieved by tem- perate microtines must result from the breakdown of a significant por- tion of the structural tissue. For lemmings, however, the rate of energy and nutrient assimilation is maximized at the expense of efficiency. Food is passed rapidly through the gut and only the most easily digested frac- tion assimilated. Melchior (pers. comm.), in laboratory feeding trials us- ing graminoids as food, showed that hunger reached maximal levels after two hours of food deprivation, and we found that guts were virtually empty after three hours. Following a change of diet, fecal pellets at- tributable to the new diet appeared in 35 minutes. The rapid passage of food through the gut and the high daily energy demand require that a significant amount of each day be spent foraging. During summer adult lemmings spend about 60 to 70% of their time out of the burrow (Banks et al. 1975, Peterson et al. 1976). Melchior (pers. comm.) estimated the stomach capacity of adult male lemmings by feed- ing animals to satiation; food consumed was approximately 10% of body weight. This gives a stomach capacity of about 125 usable joules per gram of lemming: Stomach capacity = 0.1 H^x0.20x 18,900x0.33 = 125 J g"' where Wis the body weight of the lemming in grams, 0.20 is the propor- tion dry weight in forage, 18,900 represents the average number of joules per gram dry weight of forage and 0.33 is the proportion of joules assimi- lated. Given this stomach capacity, and assuming that the value derived for adult males holds for all age classes, we may calculate the number of times the gut must be filled each day to satisfy the energy requirements of average daily metabolic rate plus growth. Since both stomach capacity and body weight are proportional to the volume of the animal, we assume that stomach capacity increases linearly with weight, whereas average daily metabolic rate increases with ^P ". Gut capacity increases more rapidly with size than does metabolism, and the number of gut fill- ings needed to satisfy average daily metabolic rate falls. Growth rate also decreases with size, further reducing the required number of fillings per day. Small lemmings (20 g) require 22 to 52 fillings per day while large lemmings (80 g) require only 14 to 22 as temperature varies from -i- 15 to -25 °C. The large number of fillings does not represent the number of foraging bouts, however, as lemmings may spend an hour or more out of the burrow at a time, at least during summer, and probably refill the stomach before it is empty (Peterson et al. 1976). Perhaps the most revealing expression of energy demand for the The Herbivore-Based Trophic System 355 -10 0 10 Temperature, "C FIGURE 10-9. The minimum feeding times required to meet the energetic demands of different-sized lemmings at several temperatures, assuming high availability of for- age (no search time). Ratios of feeding times for large and small lemmings are shown by the dashed line. lemming is the time spent foraging. Each foraging bout requires the lem- ming to leave the warmth and protection of the nest or burrow and risk exposure to predators and/or lower temperatures. Melchior (pers. comm.) found that, with grasses and sedges provided ad libitum, mean food intake (± 1 SE) was 0.14 ± 0.02 gdw min"'. At this rate, a 40-g lemming could fill its stomach in slightly less than six minutes. Since no search was involved this rate is limited only by handling time, and pro- vides an estimate of maximum ingestion rate. Minimum daily foraging time may now be calculated as a function of body weight and ambient temperature (Figure 10-9): Min. feeding time = Stomach capacity (min day"') (gdw) X Required energy (J day-') Max. ingestion rate" (g min"') Stomach capacity"' (J) Thus, small lemmings require more separate fillings, but less time to fill the gut and less feeding time even at low temperatures. Of course, 356 G. O. Batzli et al. rate of intake may increase with size, but an 80-g lemming would have to eat 66% faster than a 20-g lemming at low temperatures and 156% faster at high temperatures to spend similar amounts of time feeding. No rela- tionship between body size and maximum feeding rate could be found in Melchior's data. Actual foraging times would include search for and selection of food and would be much greater than minimum feeding times. Further- more, during winter lemmings must leave the nest to forage under lower subnivean temperatures. While they are away from the nest, nest temper- ature falls. More time spent foraging increases energy demand, which in- creases the necessary foraging time. A positive feedback relationship ex- ists and accentuates winter energy requirements. Nutrient Relationships One of the most intriguing aspects of lemming nutrition is its rela- tionship to plant nutrient concentration and nutrient flux. Nutrient con- tents of plants vary widely with plant species, plant part, site, season and year (Chapter 5). Changes in forage quality have serious implications for lemming nutrition. For instance, the preference for Dupontia fisheri in midsummer may be related to the fact that it tends to grow in nutrient- rich areas. Indeed, the general propensity of lemmings to use vegetation more in troughs and wet meadows than in drier areas may be related to the higher nutrient status of graminoids in those areas. Interest in the role of nutrients in lemming population dynamics led to examination of the nutrient dynamics of lemmings in relation to food habits and nutrient content of the forage. Of the energy ingested by the population as a whole, 1.5% is retained as production (Table 10-3): P/I - A/IxP/A - 0.33x0.045. The balance is lost as urine and feces, or through respiration. Since the energy and nutrients come from the same food, we may calculate the P/I value for nutrients as P//,^, = 1.5(L/F[^,xl.2-') = 1.3L//^,A,, where P/I[n] is the percentage of a nutrient retained by the lemming population, 1 .5 is the percentage of energy retained, L/Fi^vi is the ratio of nutrient concentration in lemmings to forage, and 1.2 is the ratio of energy concentration in lemmings to forage. Calculated values for forage (Table 10-4) represent typical values and ignore wide variation for some nutrients, so only major trends The Herbivore-Based Trophic System 357 TABLE 10-4 Relative Concentration of Energy (kJ gdw'^) and Nutrients (mg gdw'J in Forage and Lemmings Concentration Concentration Concentration in forage in lemmings factor P/1 (F)* (L)t (L/F) (%) Energy 18.9 21.8 1.2 1.5 N 25 103 4.1 5.3 K 20 8 0.4 0.5 P 2 27 13.5 17.6 Mg 3 5 1.7 2.2 Ca 2 31 15.5 20.2 Na 1 5 5.0 6.5 •Data from Tieszen (unpubl.); Chapin et al. (1975). tData from Bunnell (unpubl.). Note: Percentage of forage nutrients retained in lemming production given by ratio of production to ingestion (P/f) calculated from equation given on p. 356. should be considered. Relatively large amounts of the calcium and phos- phorus in forage are retained as lemming production, and these nutrients may be limiting for lemmings. If assimilation is low with high fecal loss, or if recycling within the body is inefficient with high urinary loss, the animal is in danger of entering negative nutrient balance, particularly during pregnancy and lactation when nutrient demands are high. A simulation model of lemming nutrition was constructed to explore the nutrient balance in lemmings (Barkley et al. 1980). The model calcu- lates the amount of nitrogen, phosphorus and calcium that a lemming could absorb from a given diet; the amount of nutrients required, given the lemming's body size, growth rate and reproductive condition; and the changes that occur in the lemming's nutrient pool. Normal nutrient pools were calculated from data of Bunnell (pers. comm.). Because little information was available on lemmings' ability to absorb nutrients or on their endogenous loss rates, which determine minimum nutrient re- quirements, this part of the model relied heavily upon nutritional infor- mation for the laboratory rat. In rats, and presumably in lemmings, ab- sorption and losses are functions of ingestion, fecal output, metaboHc rate and nutritional condition. Relatively high absorption rates and low loss rates were used, so the model was conservative and biased against the production of nutrient deficiencies. The total amount of forage consumed was based upon lemmings' energetic requirements, food habits and forage digestibility. The stand- ard model used a diet consisting of 80% graminoids and 20% mosses during midsummer and 60% graminoids and 40% mosses during mid- 358 G. O. Batzli et al. winter. Continuous shifts in the diet mimicked the field situation. Nutri- ent concentrations in graminoids were those reported for the Carex- Oncophorus meadow for 1970 when nutrients were low and for 1973 when nutrients were high (Chapter 5; Chapin et al. 1975). Early runs of the model, using a combination of starting ages, dates and nutrient concentrations, showed that only reproducing females would experience serious deficiencies — depletion of normal body pool by more than 25%. A series of runs were then conducted to compare the ef- fects of different litter sizes, nutrient contents of forage, digestibility of forage and food habits of reproducing females. In years when there were low concentrations of nutrients in their forage, adult females (120 days old) could barely support litters of seven during midsummer, four during winter and two during late August. No offspring could be supported dur- ing late June of such years. In winter, reproduction was limited by nitro- gen, calcium and phosphorus, but in early and late summer only calcium and phosphorus were limiting. During summers when there were high levels of nutrients in the vegetation, 30-day-old subadults could raise lit- ters of eight. As expected, smaller litters and slower growth in older ani- mals improved the nutrient status of the lemmings. The simulation indicated that the condition of nutrient-deficient lemmings could be improved during summer by an increase in the intake of mosses or a reduction in digestive efficiency for energy and dry matter (Figure 10-10). In winter these tactics also improved calcium and phos- phorus deficiencies, but they caused worse deficiencies in nitrogen. Re- duced digestibility allowed the extraction of minerals from more total dry matter because the lemming had to eat more to meet energy require- ments. These resuhs help to explain some unusual aspects of lemming bi- ology. The low digestibility of forage by lemmings, compared to temper- ate microtines, appears paradoxical since existence in an arctic environ- ment increases energy requirements. However, the results of the model suggest that lemmings confront a nutritional situation where calcium and phosphorus availability are more critical than energy. Low digestive effi- ciency for energy requires greater food intake and thus increases nutrient availability. Adjusting the model so that energy digestibility was im- proved caused reproducing lemmings to become calcium- and phos- phorus-deficient, even when nutrient levels in the forage were high. Cal- cium and phosphorus concentrations in arctic forage appear to be low compared to those in temperate forage (Table 10-5). Low digestive effici- ency for energy by lemmings may have evolved to assure an adequate in- take of inorganic nutrients. The lower limit of digestibility must be deter- mined by the actual ability of lemmings to find enough food and pass it through the gut. The actual digestibilities are the resuh of several con- flicting pressures. The model also helps to explain the presence of relatively high 100 The Herbivore-Based Trophic System 359 Adult 9, Litter of 7 July, Low Nutrients 100 o 100 Subadult 9, Litter of 8 July, High Nutrients T — 7-r a> o u » a. +20 40 0 Doy of Run FIGURE 10-10. Simulated nutrient pools in repro- ductive female lemmings with normal digestibility of energy in diet (S), 20% higher digestibility ( + 20) and 20% lower digestibility (-20). The assumptions of the model are given in the text. The shaded area represents lactation. (After Barkley 1976.) TABLE 10-5 Concentrations (Mean ± 1 SE) of Nutrients in Shoots of Mature Graminoids from Four Temperate Grasslands and from Coastal Tundra at Barrow Habitat No. of species Nutrient (Vo dry wt) N Ca Reference Great Basin 5 Northern Great 16 Plains Tall Grass 4 Prairie Mown Grass- 6 lands Coastal Tundra 7 1.07±0.22 0.53±0.11 0.19±0.07 Harner and Harper (1973) 1.43 ±0.34 0.28 ±0.08 0.14 ±0.05 Johnston and Bezeau (1962) 1.13±0.39 0.39±0.26 0.19±0.08 Gerloff et al. (1964) 1.47 ±0.38 0.42 ±0.06 0.29 ±0.08 National Academy of Sciences (1969) 1.86±0.07 0.16±0.01 0.13±0.01 Chapin et al. (1975) 360 G. O. Batzli et al. amounts of moss in lemming diets, up to 40% in winter, even though the digestive efficiency for energy of mosses is low. Lemmings fed only mosses reduce their intake and quickly starve. ResuUs from the model suggest that mosses serve as nutrient supplements because they are 10 to 20% higher than graminoids in phosphorus concentration and 200 to 300% higher in calcium. The low digestive efficiency for energy of mosses requires a larger food intake, and hence assures a larger intake of nutrients. But mosses cannot be used exclusively as forage because of low digestible energy or other nutritional deficiencies. Finally, reproduction during winter, when energy demand is already high, becomes more understandable with the hypothesis that growth and reproduction are limited by nutrient availability rather than energy availability. In fact, the high energy demand of thermoregulation may assure an adequate intake of nutrients, and conditions for reproduction may be nearly as favorable in winter as they are in late summer, if suffi- cient forage is available to meet energy demands. PREDATION ON LEMMINGS Introduction Predators are conspicuous in the tundra at Barrow during the sum- mer of a lemming high, and their populations have received considerable attention (Pitelka et al. 1955, Maher 1970, 1974, MacLean et al. 1974). A separation of avian and mammalian predators also distinguishes migra- tory predators from those remaining through the winter (Table 10-6). Their relative abundance throughout a standard lemming cycle is shown in Figure 10-11. During the high winter, when the major increase to a high lemming population occurs, the only predators are the arctic fox and two species of weasels. Occasional snowy owls are seen in a pre-high winter, but they are so scarce as to be negligible. The immigration of snowy owls begins in late winter (April) of a high year and nesting commences in mid-May, well before snowmelt. At snowmelt they are joined by pomarine jaegers, glaucous gulls and, in some years, short-eared owls (Pitelka et al. 1955). The total intensity of predation rises dramatically and reaches its max- imum at snowmelt, when lemmings are both most numerous and most exposed. A period of intense territorial activity follows, both within and between species. Nonbreeders are forced to marginal habitat, if they re- main at all. The breeding population, with some attrition of unsuccessful birds, remains through the rest of the season. Migration of jaegers occurs in August, and snowy owls follow in September and early October. Following the departure of avian predators, foxes and weasels are again ■= E Q 2 - o y5 ■ — O — 11 x: O 3 — 03 o ■o c op !5 O •a op o C3 00 c E E op o > n I- •— o m c. ^ 1^ 00 •— 00 c ^ 2 XI — P 00 E i >■ <^ O 3 >. > C/5 00 c 2. 'E 1) L. c c x> 00 7\ -r. .C --=00 c i( OO c .-^ C OO — E ^ i> E ^ -5 c^ E o c a> O "" c t: 00 .2 o 2 B3 .■= O '■■> S ns O H - £ o o 05 ID T3 C ^ 'i_ E w ^ 5 > ■- re •— o E 3 OO c o 00- • — :/: .= s p o O 03 ? - - — o c c c o c So o o OO "* re E o re c o 8 2 Q. 1> 3 ^ o >. . >- X) x: =^T3 .2P re -a „ P _3 .E «J o -o o b o t) .s: 53 00 Ci. Is o o b o t. 1) 00 re • re T3 i— a. 00 > re "3 O T3 c c n CO re {/; P 00 = 00 • ■S •" ^ o -a p XI .- c f re > o re o re >. ^ re -3 — X — re _ — re ~ E re -o > t: — X 1) re ly-i C. ? S S P U ■a o >. re re 3 OO p E E >. >. _ — re P 00 c m p P X •^3 * - u: = ■o -o re 5 "" E re D ^ > X c o o — (N c D. 3 c OO > p 're E _>^ _« 3 00 o 00 I o p o o 'J 00 > P re — 00 P r- r^ 3 00 P o= P ° Cl E >^ E p u T3 00 . P >. i 5 E re i2 2 E -o II c c re X re a> ^ a- a > re u. c X 5j IS c •- O 00 I- re £ ^ 3 T3 O O m — o b o t, p £ £ 3 o o LT) o I o so 2; = •* OO ~j_ ■■A -^. 3 -s: R -^ 3 S re ?3 -J so a §■ o . JS - o F OO >" So -J o I o p > ■n o J^ re p >- 5 OO = E I = re o s = O >. o I .Et 361 362 G. O. Batzli et al. Trappable Lemmings (0 c m a 0.10 0.05 -t-t- -t— t- \HX- I . Snowy Owl 2. Pomarine Jaeger 3. Other Joegers 4. Glaucous Gull 5. Short-eared Owl ^ Arctic Fox 2. Leost Weasel 3. Ermine Prehigh Summer High Winter fc^=£ a High Posthigh Winter Posthigh Prehigh Winter Summer Summer FIGURE 10-11. The estimated densities of predators during the course of a standard lemming cycle for the coastal tundra at Barrow. Periods of snowmelt and freeze-up are indicated between summer and winter. (After Pitelka et al. J 955, and unpublished observations of authors.) the only predators. Presumably, these remain until the lemming popula- tion has declined to the point that it will no longer sustain predators. Foxes then switch to alternative food, especially dead waterfowl and sea mammals, but weasels may disappear from the Barrow peninsula alto- gether between lemming highs. Avian predators are usually scarce in the post-high summer. Abundance of Predators — Numerical Response The rate at which predators take lemmings is the product of the nu- merical response (number of predators per unit area) and the functional response (number of prey taken per predator) to prey density (Holling 1959). If lemming density is insufficient to support breeding when jae- gers and owls arrive at Barrow, they quickly move on, and only nomadic individuals are seen throughout the summer. If the density of the lem- mings is intermediate to high, both snowy owls and pomarine jaegers es- tablish breeding territories, and in both species the density of the terri- tories depends upon the abundance of lemmings. Maher (1970) con- cluded that pomarine jaegers did not establish territories if the density of The Herbivore-Based Trophic System 363 trappable lemmings was below about 12 ha"'. At higher lemming densi- ties the density of breeding jaegers increased and approached an asymp- tote of about 7.3 nesting pairs km"^ at lemming densities above 100 ha"'. The lemming density required for breeding is higher for snowy owls than it is for jaegers. Snowy owls have not bred on the Barrow peninsula in the absence of breeding jaegers, but jaegers have bred during periods of low to moderate lemming density when snowy owls did not. The density of owls during breeding is much lower than that of jaegers, and the owls' numerical response to the number of lemmings is less consistent. Pitelka et al. (1955) estimated the density of snowy owls on the Barrow peninsula to be 0.2 pair km"^ during the 1953 lemming high, and Pitelka (1973) esti- mated a density of about 1 pair km"^ during the 1960 high. The difference in territory size in these two species may result from a difference in timing of nesting. Snowy owls arrive, establish territories, and begin their clutches well before snowmelt. The only exposed ground at this time is on ridges, bluffs and high-centered polygons, where the combination of wind and sublimation of snow produces small snow-free areas. Owls watch for lemmings while sitting on these vantage points. Thus, the extent of exposed ground in the spring may help to determine breeding density, along with amount of lemming movement over the snow. Pomarine jaegers generally arrive later and establish territories around the time of snowmelt, when the tundra is rapidly becoming ex- posed and lemmings are maximally exposed. Jaegers search for lemmings while patrolling or hovering about 10 m above the ground. The strong numerical response in jaeger populations suggests that territory size is set by food supply. The territories are larger than necessary in June, but by July the demand for food is higher, since the young must be fed, while the lemming population has usually declined. For the first few weeks of summer, there are often many more pom- arine jaegers and snowy owls present than eventually establish territories and breed. At low to intermediate lemming densities, jaegers engage sporadically in territorial behavior. Unless mating and nesting ensue, the birds quickly abandon the territory and leave the area. In contrast, non- breeding snowy owls may remain localized in an area for long periods of time. Even when lemming densities are high, some jaegers do not breed. Pitelka et al. (1955) estimated jaeger densities at 15 to 25 km"^ during ear- ly summer 1953; the breeding population later that summer was 6.9 pairs km"^ Maher (1974) estimated the nonbreeding population as 25 to 50% of the breeding population in 1956 and less than 25% in 1960, two other peak years. Thus, in a pre-high season, the density of snowy owls may in- crease during the summer while jaeger density always declines after a spring peak (Figure 10-11). The numerical response of other avian predators is less clear. Short- 364 G. O. Batzli et al. eared owls bred at a density of 1.0 pair km"^ in the 1953 lemming high, were absent during the 1956 and 1960 highs and bred at low densities dur- ing the 1965 and 1971 highs. Parasitic and long-tailed jaegers occur pri- marily as nonbreeding nomads in late June and July, although a few par- asitic jaegers may breed when lemming populations are high. These smaller jaegers are excluded by the aggressive behavior of the pomarine jaeger (Maher 1970, 1974). Once breeding densities have been established, the density of avian predators during late summer depends upon breeding success. Jaegers lay no more than two eggs; clutches with one egg may occur when lem- ming densities are low to moderate (Maher 1970). The clutch size of snowy owls varies considerably. Nine nests in 1952-53 had a mean of 6.3 eggs per clutch and a range of 4 to 9 (Pitelka et al. 1955). Clutches as large as 14 have been observed in Lapland (Wasenius, cited by Watson 1957). Watson (1957) indicated that the clutch size of the snowy owl on Baffin Island was a function of lemming density, and this appears to be the case in the Barrow area as well. The fledging success of both pomarine jaegers and snowy owls is highly variable and depends upon the continued availability of lemmings through the breeding season (Pitelka et al. 1955, Maher 1970). In each of the high years of 1956, 1960 and 1965 lemmings were sufficiently dense at snowmelt for jaegers to establish territories at their maximum density of about 7 pairs km'^ In 1956 lemming density fell rapidly to less than 2 ha'' in August, and jaeger breeding success was 4%. In 1960 lemming density remained high all summer, with an estimated density of 215 ani- mals ha"' in August, and jaeger breeding success was 55%. In 1965 lem- ming density fell from 150 ha"' in June to less than 1 ha"' in August, and almost no jaegers fledged (Maher 1970). Thus, a high lemming popula- tion in June does not guarantee that significant jaeger recruitment will occur. Indeed, Pitelka et al. (1955) found higher breeding success in the pre-high year of 1952 than in the high year of 1953. Given the tendency of the lemming population to increase during a pre-high summer and de- cline rapidly in the summer of a population high, breeding success of jae- gers may often be better during years of moderate lemming density. Snowy owls begin incubation with the first egg laid, resulting in an interval between hatchings of about 40 hours (Watson 1957). Hence, in a brood of eight chicks the youngest and oldest will differ in age by about 2 weeks. Since the parents tend to feed the more active and aggressive young first, the older chicks survive at the expense of the younger if lem- mings become scarce. This mechanism allows a close adjustment of owl breeding success to changes in lemming density through the season. Pitel- ka et al. (1955) estimated fledging success of snowy owls to be less than 50% in 1952 and 1953. Even with prefledgling mortality of 75%, the mean clutch reported by Pitelka et al. (1955) of 6.3 eggs would produce The Herbivore-Based Trophic System 365 1.6 fledged young, well above the maximum of 1.1 fledged young per jaeger nest (2 eggs x 55% success rate) reported by Maher (1970) for 1960. Mammalian predators are difficult to observe, and major popula- tion changes occur in winter, so less is known about their population fluctuations. The role of the arctic fox is modified by the proximity of the ocean and by human activities. Although foxes hunt lemmings, they also have access to an abundant supply of carrion, primarily carcasses of marine mammals and eiders that are crippled and lost by hunters. Foxes are trapped commercially by man during the winter. As a result, fox den- sity shows little correlation with lemming density (Figure 10-11), although a strong numerical response is reported elsewhere (MacPherson 1969). An adult female fox regularly hunted and caught lemmings on the Biome research area throughout the summer of 1974, but foxes are usu- ally more abundant in winter than in summer. The main denning areas are inland, possibly because of the low density of breeding waterfowl in the coastal tundra. In some areas foxes are known to take large numbers of waterfowl eggs (Underwood 1975). Of all predators the least weasel appears most closely tied to the lem- ming. Least weasels may be absent from the coastal tundra at Barrow during years when the lemming population is low, but they appear during lemming peaks. Presumably, recolonization occurs from areas to the south where the lemming cycles are of lower amplitude and where local asynchrony of fluctuations of coexisting microtine species may allow maintenance of a more stable predator population (Pitelka 1957b). Nev- ertheless, the strong numerical response that occurs during most high winters results from reproduction. The reproductive cycle of Mustela nivalis differs from that of other small mustelids in that delayed implant- ation does not occur. This allows the least weasel to make a rapid repro- ductive response to increasing or high lemming populations. Juveniles collected at Barrow on 18 May and 6 June 1963 attest to the occurrence of winter breeding. Four pregnant females collected during summers with high lemming densities contained 15, 12, 12 and 3 (x = 10.5) em- bryos, compared with an average of 4.8 for temperate zone females (B. Fitzgerald, pers. comm.). Thus, a dramatic reproductive response to lemming density appears to be present. By immigration and reproduction least weasels can increase from nearly zero to maximum densities over the course of a pre-high summer and high winter (MacLean et al. 1974). Thompson (1955b) estimated the peak density at 25 km"^ during the 1953 lemming high, a value that Maher (1970) considered conservative. Maximum densities since 1969 have been far less than this. MacLean et al. (1974) reported 59 least wea- sels collected during summers following winters when lemming densities increased (1953, 1956, 1960, 1963, 1965, 1969), but only seven specimens in all other summers. 366 G. O. Batzli et al. Ermine likewise show a numerical response to lemming density, al- though the response is less than that of the least weasel (MacLean et al. 1974). Ermine usually occur on the Barrow peninsula when lemming populations increase, but the density of ermine is probably less than 10% of the density of the least weasel. Weasels differ from avian predators in their response to declining lemming populations. Low lemming density leads to reproductive failure and reduction of the population of adult jaegers and owls. By early fall, regardless of nesting success, all avian predators have departed. Weasels are less mobile; they remain through the summer into the post-high win- ter, and exert considerable predation pressure upon the declining lem- ming population. Eventually, by death or emigration, weasel density falls to undetectable numbers that characterize the low phase of the lem- ming cycle. Nutrition and Energetics of Predators — Functional Response Predators characteristic of the coastal tundra at Barrow are categor- ized as obligate or facultative lemming predators (Table 10-6). Obligate predators are those whose presence depends upon an adequate popula- tion of lemmings. Such predators show the largest numerical response and the absence of a strong functional response. Facultative predators are able to maintain a population at times of low lemming density by use of alternative food sources. Their numerical response to lemming density is small or even inverse, but they show a marked functional response. This section will consider two aspects of the functional response of predators: 1) changes in prey selection associated with changes in the density of primary and ahernate prey, and 2) bioenergetic factors which determine rate of prey capture and consumption. Clear differences in the food habits of jaegers are evident between years and between species (Maher 1974). Remains of lemmings were found in 98% of the pellets of pomarine jaegers taken in the high lem- ming years of 1956 and 1960, indicating the strong dependence of breed- ing pomarine jaegers upon lemmings. Lemmings were found in 10097o of 75 pellets collected in 1959, when a few pairs of pomarine jaegers at- tempted breeding during a period of much lower lemming density. Even when lemming populations were low, lemmings were found in 41 % of 56 stomachs of transient, nonbreeding pomarines collected in 1957 and 1958. This slight functional response was of little importance because the birds were rarely seen after mid-June Parasitic jaegers preyed more on birds and their eggs than did the other jaegers, particularly when lemming densities were low (Maher 1974). However, during the 1956 lemming high, lemmings accounted for The Herbivore-Based Trophic System 367 75% of the food items in pellets of the parasitic jaeger. Data were avail- able for the long-tailed jaeger only for the 1955-58 period when this spe- cies took fewer birds and many more insects than did the other jaegers. Long-tailed jaegers are most common in July, when adult craneflies (Tipulidae) are abundant on the tundra surface. Comparable data on prey selection by snowy owls in relation to lem- ming density are lacking. Pitelka et al. (1955) noted that Lemmus makes up the bulk of the prey, but they also found owl pellets containing the re- mains of a variety of birds, ranging in size from the Lapland longspur to the old-squaw duck, as well as the remains of a least weasel. Examination of pellets over many years shows that the fraction of the diet consisting of lemmings is as large for snowy owls as for pomarine jaegers and may be even larger at low to moderate lemming densities. Arctic foxes depend not only on lemmings but also on carrion dur- ing the winter and on birds and their eggs during summer. Using the lem- ming population dynamics model, a lemming mortality equivalent to 13.5 kg ha"' was estimated during the winter (September-May) of a lem- ming high. These carcasses could constitute an important food source. Mullen and Pitelka (1972) investigated the disappearance of lemming carcasses by placing dead lemmings on the tundra in autumn. They found that virtually all carcasses disappeared by the following spring. In some cases other lemmings were implicated, but weasels were few, and winter observations indicated that foxes were primarily responsible for the removal of carcasses. No data are available on the summer diet of foxes in the Barrow re- gion, but in the Prudhoe Bay region Underwood (1975) found remains of Lemmus and Dicrostonyx in 86% of 50 fresh scats in an inland area and 75% of 24 scats in a coastal area, despite generally low densities of mi- crotines during the 1971 summer. Birds were found in 50% of the scats from inland and 63% of the scats from the coast. Underwood (pers. comm.) reports finding as many as 30 lemming carcasses in a single fox den at other sites on the Arctic Slope. It is likely that the functional re- sponse of foxes to lemming density is much greater than their numerical response. Data on the winter diet of both species of weasel derive from obser- vations of remains and scats at lemming nests (MacLean et al. 1974), and thus may be biased in favor of lemmings. However, in the absence of al- ternative prey during winter, it seems safe to say that virtually the entire diet consists of lemmings. During summer weasels also take birds and their eggs, and a functional response to lemming density is probable. Selective predation upon sex or age classes within the lemming pop- ulation could influence the population dynamics of the lemmings. Since both jaegers and owls regurgitate pellets containing the bones of their prey, it is possible to estimate the frequency of capture of various classes 368 G. O. Batzli et al. of prey. Maher (1970) analyzed the remains of lemmings in pomarine jaeger pellets collected in the high lemming years of 1956 and 1960. The proportion of prey in the smallest size class, corresponding to nestlings, was greater in 1956, when lemming density declined rapidly during the summer, than in 1960, when lemmings remained abundant throughout the summer. Osborn (1975) compared the 1960 distribution of prey in jaeger pellets with estimates of relative abundance of different age classes and sexes from snap-trapping data. Little or no difference was evident. These observations do not rule out selectivity because snap trapping probably overestimated the frequency of males in the population. Greater movement by males than females exposes males to more traps, just as it exposes them to more predation. Although more male lemmings were found in jaeger pellets in 1960, Maher's 1956 sample contained a greater proportion of females (54%). Thus, there seems to be no consis- tent selection of males by jaegers. Pitelka et al. (1955) determined the sex of 76 lemmings accumulated at a single snowy owl nest in June 1953; 25 were females and 51 were males. A similar preponderance of males was found by Thompson (1955c) in snowy owl pellets. The greater vulnerabiHty of males to owl predation seems clear. During the winter weasels prey heavily upon lemmings and live in the nests of their victims. MacLean et al. (1974) suggested that the large nests that are subject to the greatest predation are built by breeding fe- males and that weasel predation may be concentrated upon reproductive females. This hypothesis may be tested by comparing the sex ratio of lemming populations after winters with heavy predation with the sex ratio after winters with little predation (Table 10-7). The sex ratio signifi- cantly favored males after the winters of 1959-60 and 1964-65; a sex ratio favoring males after the 1962-63 winter is marginally significant. During all three of these winters weasel predation was heavy (MacLean et al. 1974). Breeding also occurred during these winters, so the stress of breed- ing might have contributed to the skewed sex ratios. The quantity of food removed per predator can be examined by cal- culating the number of prey needed to meet the energy requirement or by field observation of prey capture rate. The latter approach allows for the interaction of behavior and energy requirement in a natural setting, but such data are not easy to obtain. Gessaman (1972) studied the metabolic rate of snowy owls by mea- suring both the oxygen consumption and the food intake of birds con- fined in outdoor pens. He concluded that the daily consumption of an average adult owl was 6.6 60-g lemmings during the coldest period of the winter (T = -29 °C) and 4 lemmings in summer (Ty -5°C). Pitelka et al. (1955) observed food consumption by a captive immature snowy owl kept in an outdoor cage during summer. During the late phase of growth. The Herbivore-Based Trophic System 369 TABLE 10-7 Sex Ratios of Lemming Popula- tions in Spring Year Date o-o- 99 o'o-/99 X' 1959 1-15 June 70 97 0.72 4.37** 1960 16 May-15 June 520 278 1.87 73.39t 1961 16 May-15 June 46 49 0.94 0.09 1962 1-15 June 248 226 1.10 1.02 1963 16 May-15 June 245 204 1.20 3.74* 1964 1-15 June 104 106 0.98 0.02 1965 16 May-15 June 308 215 1.43 16.54t •p<0.10. ** p< 0.05. tp<0.001. Source: Data from Pitelka (unpubl.). it consumed an average of 7.5 60-g lemmings day"'. After its weight sta- bilized, food consumption fell to 5.4 lemmings day"', a value 35% greater than Gessaman found for summer. Maher (1970) estimated the food requirement of pomarine jaegers by direct observation of adults, which were unfettered, and chicks, which were penned in the field. Chicks near the age of fledging consumed about 3.3 60-g lemmings day"', about one lemming less than adults. Because of the very high insulating value of the winter pelage rela- tive to the summer coat, the maintenance energy requirement of arctic foxes does not change appreciably through the year (Underwood 1971). Deposition of fat during summer, use of fat during winter and changing activity patterns modify the annual energy budget greatly. Total assimi- lated energy requirements for a single fox varied more than five-fold over the year. Requirements were greatest in^ August, about 4609 kJ day"', equivalent to 14.7 60-g lemmings, and least in April, about 838 kJ day"' or 2.7 60-g lemmings. This remarkable difference may help to explain the greater density of foxes during winter than summer, even though food appears to be more available in summer. The energy requirements of arctic weasel populations have not been measured. But extrapolating from Brown and Lasiewski's (1972) equa- tions for Mustela frenata, MacLean et al. (1974) estimated that a 65-g least weasel living at an ambient temperature of -20 °C would have a rest- ing metabolic rate of 409 kJ day"', which would require consumption of 1.3 60-g lemmings day"'. However, weasels use lemming nests in winter, and maintaining a microclimate at 0°C reduces the resting metabolic rate to 281 kJ day"', requiring only 0.9 60-g lemming day"'. Estimates of winter predation by weasels can be based upon remains 370 G. O. Batzli et al. FIGURE 10-12. The impact of avian predators on lem- ming populations in 4 years as indicated by the percent- age of mortality accounted for by predators. Lemming densities are given in Figure 10-2. (After Osborn 1975.) found in winter nests. High predation rates — 20% of nests on Banks Island (Maher 1967) and 35*^0 of nests at Barrow (MacLean et al. 1974) — have been reported. The actual percentage of the population consumed was probably much higher since weasels usually consume more than one lemming per nest. But exact predation rates could not be calculated be- cause the number of lemmings using each nest and the number of nests built by each lemming are unknown. In order to assess the impact of avian predators on lemming popula- tions, Osborn (1975) modeled the numerical and functional responses of snowy owls and pomarine jaegers. His model contains the information discussed above, plus growth curves of Watson (1957) for the snowy owl and Maher (1970) for the pomarine jaeger. He calculated the mortality rates for lemming populations when they were high enough to ensure breeding by the two avian predators, and expressed predation as the percentage of mortality (Figure 10-12). In general, the model indicated that avian predation accounted for the greatest amount of mortality at snowmelt in early June when non- breeding pomarine jaegers were still present. In 1956 and 1965, when the lemming population declined rapidly from peak levels, predation became an important source of mortality again in late summer when lemming populations were low. In 1964 and 1973, when lemming populations in- creased in late summer, percent predation declined. The simulation for the early summer of 1972 showed the maximum impact of predation: 88% of total mortality. In 1960, when lemming densities were at their highest recorded levels, absolute predation rates were also highest (79 g ha"' day"'), but avian predation accounted for only 39% of the mortal- ity. Finally, Osborn found that in early June the amount of predation The Herbivore-Based Trophic System 371 relative to standing crop of lemmings peaked at lemming densities of 25 ha"' and dropped off rapidly at lower or higher densities. Osborn did not simulate predation for summers when lemming pop- ulations were low, but there is some evidence suggesting that predation relative to standing crop is quite high. Pomarine jaegers do not defend territories when lemming populations in early June are less than 10 ha"'. At these densities lemmings are distributed in small patches, upon which wandering predators prey. The percentage of lemmings taken may be greater than during years when lemmings are protected from other pred- ators by pomarine jaegers. During 1972 one of the live-trapping grids was located within an area that was defended by a nesting pair of pomarine jaegers and we ob- served minimum survival rates for adult lemmings of 40 to 70% per 28 days. On two other grids not defended by jaegers lemming survival rates were less than 20% per 28 days. During 1973 and 1974 no jaegers nested in the study area, and adult survival rates were less than 30% per 28 days on all grids. Wandering jaegers and owls were relatively common in all three years. Weasels were present in 1973, and in 1974 an arctic fox re- sided in the study area. Animals on trapping grids not defended by terri- torial jaegers appeared to be healthy, and all females were reproducing, so predation by other predators appeared to be the most likely cause of the high mortality. FACTORS INFLUENCING LEMMING POPULATIONS Abiotic Factors One of the most dramatic features of tundra is the rapid transition between the mild summer and the severe winter. Because they neither mi- grate nor hibernate, lemmings must function throughout even the most extreme conditions. During the summer, temperatures near the ground are above freezing, but they are nearly always below the lower limit of thermoneutrality (17 °C) of lemmings (Figure 10-7). Hence, even summer temperatures cause metabolic rate and food demand to be elevated above the minimum value, but they pose little direct threat to survival. Habitat flooding at the time of snowmelt and again, in some years, following July and August rains may be a greater danger to lemmings. Much of the low-lying habitat, including meadows, polygon troughs and basins of low-centered polygons, becomes inundated. Since water des- troys the insulation value of lemming fur, and the high specific heat of water makes it an effective heat sink, lemmings must keep dry most of the time. Slight flooding (2 to 3 cm deep) can make a habitat unsuitable except for occasional foraging. At snowmelt the subnivean habitat col- 372 G. O. Batzli et al. lapses and the lemming population is concentrated into small areas of suitable habitat; when populations are dense, a period of intense social interaction may ensue. Lemmings may be forced into unfamiliar habitat where they may become more susceptible to predation. Thus, spring is a particularly traumatic period during which breeding subsides and mor- tality is high. In years of high rainfall much of the low-lying habitat con- tinues to be flooded and unavailable to lemmings most of the summer. Polygon troughs and wet meadows, which are most susceptible to flood- ing, have the highest density of food plants and are preferred winter hab- itat. If these areas are flooded at freeze-up, they can remain unavailable to lemmings all winter long. The effect of late-season flooding may be greater than its spatial extent alone would suggest. In August and early September lemmings encounter the only marked diurnal change in light intensity and temperature that they exper- ience. The snowpack generally develops between mid-September and mid-October; delay poses two dangers for lemmings. First, if the vegeta- tion becomes coated by freezing rains, the energy cost of foraging greatly increases. Second, in the absence of snow the ground surface is exposed to increasingly low nighttime temperatures. The reduction of breeding at this time attests to the severity of the period in the annual cycle of the lemming. Fuller (1967) has suggested that weather conditions in autumn can be a major factor influencing the population dynamics of lemmings. The dense and shallow snow cover offers only modest protection from winter cold; very low temperatures are encountered at the ground/ snow interface. In 1970-71 the temperature at the ground surface under 30 cm of snow at the Carex-Oncophorus meadow dropped below -20 °C in early December and to -25 °C in early March (Chapter 2). In March 1972 the mean temperature was -26 °C under 29 cm of snow (MacLean et al. 1974). Temperatures as low as -32 °C were recorded under 15 cm of snow. Lemmings apparently take advantage of the higher temperatures that are found under deeper snow by concentrating their activity in poly- gon troughs. MacLean et al. (1974) found a significant correlation be- tween nest density and amount of topographic relief {r = 0.52, p < 0.01), and presented evidence that winter reproduction was inhibited in years of shallow snow accumulation. The structure of the snowpack may be just as important to lemmings as its depth. Once the snowpack accumulates, moisture is redistributed, and a strong structural profile develops (Chapter 2). The profile consists of two main layers: a fine-grained wind-packed layer of high density, and a large-grained layer of very low density (depth hoar). Where the snow is shallow the depth hoar layer may be thin or nonexistent. Although lem- mings can move easily through depth hoar, dense snow may exclude them from a portion of the habitat. Freezing rains or partial thaws after The Herbivore-Based Trophic System 373 the snowpack has developed can produce ice layers in the snowpack, and these also inhibit the formation of depth hoar. The subnivean atmosphere has been suggested as another important feature of the winter environment of lemmings. The dense, wind-packed snow offers high resistance to gaseous diffusion. Kelley et al. (1968) and Coyne and Kelley (1974) demonstrated a buildup of subnivean carbon di- oxide in fall, and again in spring. On two occasions, concentrations rose rapidly to 700 and 1500 ppm. While well above normal levels of 320 ppm, these values appear to be too low to influence the physiology of lemmings. Alveolar air in mammalian lungs is normally 5 x 10* ppm CO2, and that level could easily be maintained by a slight change in respiratory rate (Johnson 1963). Snow chimneys, tunnels dug by lemmings to the snow's surface, are frequent when lemming populations are high. Melchior (pers. comm.) found that they are used most during fall and spring, when release of gas under the snow is to be expected. Lemming tracks are often seen in fresh snow around the chimneys. Sometimes the tracks may be traced to the same or another chimney, indicating that the lemming returned to the subnivean environment. In other cases the tracks lead to a dead lemming or to signs of predation by foxes or owls. There are obviously risks asso- ciated with ventures above the snow. The advantages are unknown but may include escape from the toxicity of the subnivean atmosphere or dis- persal to new habitat by an easier route when the depth hoar layer is poorly developed. In summary, reproduction regularly subsides at snowmelt and dur- ing freeze-up before the snowpack develops, indicating that these are pe- riods of stress for lemmings. Mortality rates, particularly from exposure to predators, are high at snowmelt. Circumstantial evidence suggests that a shallow snowpack and low winter temperatures inhibit winter repro- duction, but we found no significant correlation between snow depth and spring population densities at Barrow over the past 25 years (r = +0.18; p > 0.25). If the depth hoar layer of the snowpack is poorly developed, winter populations decline, apparently because insufficient forage is ac- cessible to them. Adequate snow depth and a well developed hoar layer are necessary but not sufficient to allow a population increase during winter. Unfavorable abiotic factors may contribute to the decline of high populations, but they do not appear to be necessary for the decline. Poor winter conditions can reduce populations to extremely low levels and pre- vent recovery of the population, thereby altering the timing of peaks. In summer, flooding may restrict the habitat available to lemmings, but reproduction and survival do not appear to be directly affected. 374 G. O. Batzli et al. Predation The general characteristics and feeding patterns of predator popula- tions have been discussed above. In this section the impact of predation on the population dynamics of lemmings will be evaluated. Osborn's (1975) simulation indicated that in some years avian predators could ac- count for 88% of the early summer mortality of lemmings. The major impact came at snowmelt when lemming densities were about 25 ha"', and percent predation dropped off rapidly at higher densities. Maher (1970) calculated the total impact of all predators on lemming popula- tions during a high year. The calculations were based on the assumption that no young lemmings were weaned until mid-July and that mean litter size was six. Both assumptions are conservative (BatzH et al. 1974). He concluded that predation could not prevent population growth during the summer if lemming densities were greater than 65 ha"' at snowmelt. Since lemming densities during high years are usually greater than 100 ha"' at snowmelt, something in addition to predation must account for declines during the following summer. Predatory impact at high lemming densities is reduced by the protection afforded lemmings by territorial pomarine jaegers. During summers with lower population levels, when jaegers are not territorial, predation rates may be high enough to prevent population increases. Thus, predation contributes substantially to the rapid decline of lemming densities during some summers, and may even prevent population growth in others, but it is not sufficient to cause the decline of lemming densities from peaks. During winter weasels are the most important lemming predators, and there is some evidence that winter predation rates may be sufficient to restrict the growth of lemming populations. During 1963 and 1969, lemming populations reproduced under the snow, but by snowmelt the populations had been reduced, and there was evidence of intense preda- tion by weasels (Mullen 1968, MacLean et al. 1974). Maher (1967) re- ported a similar circumstance on Banks Island. Pearson (1966) argued that the most significant effect of predators on cycling populations of microtines is the reduction of populations to extremely low levels, which delays their recovery. Maher (1970) and MacLean et al. (1974) supported this view as it applied to lemming popu- lations. The increased mortality that adult lemmings experience during summers when populations are low lends further credence to such a role for predation. However, high populations of lemmings usually develop under the snow and not during summer, so the critical period is winter. If winter predation regulates lemming populations, intensity of predation should be negatively correlated with the change in lemming density. Our data for the winters of 1969 through 1974 do not show this trend. During The Herbivore-Based Trophic System 375 four out of five winters populations did not increase even though preda- tion was low. We conclude that high predation rates are not a necessary condition for the maintenance of low populations. In summary, predation contributes to population decHnes and may be sufficient to prevent increases at low densities, but it is not sufficient to account for summer declines following a peak. Furthermore, relaxa- tion of winter predation will not necessarily lead to population increases. Nutrition Weber (1950a) observed many dead lemmings and devastated vege- tation following the spring 1949 population peak; he therefore proposed that exhaustion of food supplies and subsequent starvation caused the decline. However, Thompson (1955a) noted that after the peak in 1953 vegetation grew rapidly, even though total primary production was only half that expected with no grazing. Most of the dead lemmings that Thompson found appeared to be victims of predation. He suggested that lemming populations declined because of high predation rates and low reproductive rates, which resulted from low vegetative cover and poor forage availability. Pitelka (1957a, b) supported and expanded Thomp- son's views to include the possibility of changes in forage quality as a fac- tor influencing lemming reproduction. Finally, Pitelka (1964) and Schultz (1964, 1969) proposed the nutrient-recovery hypothesis to ac- count for the cyclic nature of lemming population dynamics. According to the hypothesis the nutrient concentration in vegetation declines fol- lowing a lemming high and does not increase sufficiently to support good reproduction by lemmings for two or three years. The hypothesis is com- plex, involving several components of the ecosystem, and it will be con- sidered in detail. Schultz's descriptions contain a few gaps concerning the mechanisms which drive the nutrient-recovery hypothesis, so we have embellished it slightly in the following treatment and tried to make expli- cit all the major causal links. The fundamental interactions and mechanisms of the nutrient- recovery hypothesis are summarized in Figure 10-13. Intensive grazing takes place during the winter buildup of the lemming population and continues until the population crashes during the summer, apparently as a result of the combined effects of habitat destruction and predation. At snowmelt soluble nutrients released into the meltwater from urine, feces and clipped vegetation are rapidly taken up by growing plants. Thus, nu- trient concentrations are high in the early summer forage. Later in the summer this growth becomes standing dead material, and locks up some nutrients that would otherwise be available the following summer. In addition to releasing nutrients the lemmings' intensive grazing re- 376 G. O. Batzli et al. RECOVERY 14) Reproduction improves and population builds over two or three years 13) Plant roots shallow^wtiere nutrient concentration greater,and forage quality improves 12) Increasing standing dead increases insulation of soil wfiich decreases depth of thaw 1 1 ) Poor reproduction so popula- tion remains low the year follow- ing crash 10) Low grazing over winter and poor nutrient availability 'during following summer DECLINE HIGH LEIVIMING DENSITY 1) Intensive winter and early summer grazing 2) Nutrient release from feces, urine and litter at snow melt 3) Uptake high during early summer so high nutrient content in early forage 4) Nutrients locked in standing dead at end of summer LOW LEMMING DENSITY 5) Live biomass and standing dead material reduced 6) Low insulation leads to increased depth of thaw 7) Plant roots grow deeper in soil where less phosphorus 8) Low nutrient uptake produces poor quality forage in late summer 9) Poor quality winter forage (leaf sheath bases) so no winter repro- duction and dechne continues FIGURE 10-13. Summary of steps in the nutrient-recovery hypothesis. (Adapted from Schultz 1964, 1969.) duces the amount of standing live and dead plant material during the summer of a lemming decline. This in turn reduces the albedo of the soil and its insulating cover, and the depth of thaw increases. As the depth of thaw increases, plant roots penetrate more deeply to where soil nutrient solutions (particularly phosphorus) are more dilute, and the available nu- trients are distributed throughout a larger volume of soil. The low nutri- ent uptake of these roots leads to low nutrient concentrations in late sum- mer growth, so that the leaf sheath bases produced then provide poor winter forage for lemmings. Because of the poor quality forage little lem- ming reproduction occurs during the winter, and the decline continues. There is little winter grazing, so the nutrient pulse in the spring is weak, and the nutrient quality of the vegetation stays low during that summer. The Herbivore-Based Trophic System 377 Lemming reproduction remains low during the summer and winter after a decline, and the standing dead material begins to accumulate. The increase in standing dead material and litter improves insulation over the soil and reduces the depth of thaw over the next two or three summers. As the depth of thaw decreases, plant roots are confined to soil with higher nutrient concentrations, and forage quality improves. Lemming reproduction then increases, and the population grows until a new peak is reached, usually 3 to 4 years after the last. Our evaluation of the major premises of the nutrient-recovery hy- pothesis follows. Heavy grazing by lemmings can drastically reduce the standing crops of live and dead aboveground biomass (Dennis 1977) and increase feces and urine output (steps 1-3, Figure 10-13). The total con- sumption of graminoids by lemmings during a high year amounts to over 40 g Tn'\ nearly 50% of the annual aboveground production and 20% of the total net production. Consumption is less than 1 g m"^ yr"' when pop- ulations are low (Figure 10-14). About 70% of the dry weight consumed is returned to the surface as feces and urine. Except for nitrogen, potassium and sulfur, minerals are primarily returned in feces (Wilkinson and Lowrey 1973). While the urin- ary minerals are readily available to plants, those in feces may not be. Most fecal phosphorus probably occurs as calcium diphosphate, a form that is soluble in a weakly acidic (pH 5) solution (Barrow 1975). The rate of nutrient loss from feces will depend upon where they are located. For instance, feces in ponded troughs or basins of low-centered polygons should lose their phosphorus more rapidly than feces on rims. Since most lemming feces are deposited in places where standing water occurs, at least during snowmelt, and since the tundra soil solution is acidic (pH 4.5 to 5.5), fecal phosphorus should be readily available to plants. Prelimin- ary leaching experiments, using a solution that mimicked the soil solu- tion, showed that over 80% of phosphorus was removed from feces in 24 hours (Chapin et al. 1978). Standing dead plant material can amount to 40 g m"^ at snowmelt (Chapter 3), four to five times the dry weight of live material. Most of this represents the previous summer's production less those nutrients that have been removed by translocation and leaching. By felling standing dead over the winter, lemmings do add organic material to the tundra surface where it will decompose more rapidly, but its nutrient content is less than half of that from feces and urine deposited during the high win- ter. Disruption of mosses and lichens, which also takes place when the lemming population is at its peak, may also increase the rate of decom- position, but the amount is unknown. The total influx of nutrients produced by lemming activity may be considerable. The average amount of soluble phosphorus in the top 5 cm of soil is 0.5 to 4 mg m'\ whereas the amount deposited in lemming feces 378 G. O. Batzli et al. o c 200r a Trappable Lemmings High Year o E ~o x: 50 - b. Vegetation Consumed Graminoids - (high yeor only) y^ ^"N. - ^^^^...^^ Mosses N \ - ~ ^^--^^^^-^^ ^\ \ -4-- N ^-r r r V- 1 1 1 1 1 T- lOOOh c. Stood ing Crop of Graminoids 'o 0 A K o SI 1000 d.l -itter and Feces Higti Yeor^^^^ 0 ~L ow Ye or^-^ ^« Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug FIGURE 10-14. Idealized comparison of lemming density, forage consumption, standing crop of forage and deposition of waste products during years with high and low popu- lations. The calculations for lemmings are based on data presented earlier in the chapter. The density estimates in- clude only animals usually caught in traps (> 20 days old), but consumption and waste are calculated for the entire population. during a year when the population is high, assuming 90% solubility, is about 90 mg m"^ The phosphorus pool turns over very rapidly, and graminoids must absorb about 3 mg P gdw"' of plant material produced, assuming an average of 0.3% total phosphorus. Since the peak above- ground standing crop averages 80 g m■^ lemming feces would provide 35 to 40% of the required phosphorus. Most of the nutrient release from feces probably occurs during snowmelt in spring. That pulse spurs early nutrient uptake by plants and microbes since phosphorus uptake is pro- portional to phosphate concentration (Chapin and Bloom 1976). Hence, the first three premises of the nutrient-recovery hypothesis appear to be substantiated (Figure 10-13). The Herbivore-Based Trophic System 379 The idea that large quantities of nutrients may be tied up in organic matter during the summer following a lemming peak (step 4, Figure 10-13) is less tenable. Lemmings reduce the aboveground standing crop of vascular plants by about 50% at midsummer of a peak year (Figure 10-14) so nutrient storage in standing dead at the end of summer is also reduced. Furthermore, 15% of the phosphorus put into live biomass may be removed by translocation in late summer and by leaching the follow- ing spring (Chapin, pers. comm.). Although a few lemming carcasses may be found at snowmelt, large numbers of carcasses do not accumu- late on the tundra, apparently because they are eaten by predators and scavengers (Mullen and Pitelka 1972). Nutrients from all tissues other than bones are probably returned rapidly to the soil. Phosphorus and calcium are concentrated in lemming bones, but only 10 to 20% of the total phosphorus and calcium in lemming forage is retained in bone. In a high year lemmings consume about half of the aboveground vascular plant production, so no more than 5 to 10% of the total phosphorus and calcium content of forage could be sequestered in lemming bones. In some spots, where lemmings have grubbed for rhizomes, the standing crop of vascular plants may be reduced 90% (Schultz 1964, Dennis 1968). Thinning of the plant canopy does increase depth of thaw, although Schultz does not provide quantitative data. A simulation model suggests that complete removal of the canopy increases maximal depth of thaw by about 20% of normal, or about 5 cm, when the surface is satu- rated with water (Ng and Miller 1977). If the moss layer is drier, the ef- fect on thaw is somewhat less. When lemmings are excluded from patches of tundra for long periods of time, standing dead plants continue to accumulate, and thaw depth is reduced as much as 25% (Batzli 1975b). Thus, the fifth and sixth premises are supported, although the ef- fect on depth of thaw does not seem to be large. Schultz (1964) presented evidence that total calcium and phosphorus decreased with depth in tundra soils; however, soluble soil nutrients do not necessarily follow the same pattern (Chapter 7), although soluble in- organic phosphorus usually does (Bar^l and Barsdate 1978). The decline in nutrient absorption rates of temperate plants for 7 to 10 days after grazing, reported by Davidson and Milthorpe (1966), might also support the nutrient-recovery hypothesis. But the nutrient absorption rates of tundra graminoids increase following grazing under field conditions (Chapin 1980b). Moreover, the direct impact of grazing upon nutrient absorption rates would not last long. Schuhz's idea that roots would exploit greater soil depths does not seem likely when one considers that the plants could use the more con- centrated nutrients in the upper soil horizons. In fact, all roots of Du- pontia fisheri and the secondary absorbing roots of Carex aquatilis, the two most important forage plants for lemmings, are found in the upper 380 G. O. Batzli et al. soil horizon, regardless of thaw depth (Chapter 5). Furthermore, when soil temperatures and thaw depth increase as a resuh of human disturb- ance, nutrient availability and plant production increase (Bliss and Wein 1972, Challinor and Gersper 1975, Chapin and Van Cleve 1978), rather than decrease as the nutrient-recovery hypothesis predicts. For these rea- sons, the links between depth of thaw, nutrient availability in soil and nutrient concentrations in plants that are proposed by the hypothesis (steps 7-9, Figure 10-13) do not seem tenable. A more likely explanation for the decline in plant phosphorus concentration observed by Schultz (1964) in the years following peaks in the lemming population is that in- tensive grazing and grubbing for rhizomes sharply decrease plant phos- phorus reserves. Simulations suggest that plant nutrient reserves may be severely depleted by grazing (Chapin 1978). The involvement of other ecosystem components need not be invoked. Nutrient levels in forage may influence both litter size and the timing of reproduction of lemmings (steps 9-14, Figure 10-13). Phosphorus, calcium and nitrogen all have been implicated by a model of nutritional physiology of lemmings (see Nutrition and Energetics). Apparently lem- mings have adapted to low nutrient availability through high forage in- take rates, low digestive efficiency of energy and selection of mosses as a calcium supplement. Even if low nutrient quality of forage sometimes prevents lemming population growth, it probably is only one of several factors which can do so. Poor snow conditions and high weasel densities may also prevent population growth during winter. Hence, forage qual- ity may influence the rate of lemming population growth, but other fac- tors unrelated to nutrition may be equally important. In 1973 the depth of thaw averaged about lO^o less than in 1972. Low air temperatures caused peak standing crop of aboveground Du- pontia to be 25*yo lower, but concentration of phosphorus was 200*7o greater (Chapter 5, Table 5-4). These changes are similar to those ex- pected during the course of a lemming cycle according to the nutrient- recovery hypothesis. Yet in 1973 the lemming density was only 5 to 10 ha"', about half that of 1972, and the population declined during the winter of 1973-74. The nutrient-recovery hypothesis predicts that it should have increased. Our general conclusion is that the nutrient-recovery hypothesis, as developed by Schultz, should be modified. Lemming activity does not appear to control the nutrient concentration of forage by changing depth of thaw, nor do trends in lemming populations necessarily follow trends in forage quality, at least as indicated by phosphorus concentration. Nevertheless, the hypothesis has been valuable because it pointed out the importance of considering vegetational quality as well as quantity for herbivore populations. The quality of available forage is difficult to evaluate. About 40 spe- The Herbivore-Based Trophic System 381 cific nutrients are known to be required by rodents (National Academy of Sciences 1972), but the exact requirements of lemmings are unknown. Some nutritional work has been done on lemmings and their forage, which forms the basis for our tentative conclusions regarding the role of forage quality in population dynamics. Calculations of the energy requirements of lemmings during popula- tion buildup to a peak of 150 ha"' showed that in a normal high year suit- able forage would be completely utilized before snowmelt (Batzli 1975a). Some high population levels seem to reach 225 ha"' or more before de- clining, so insufficient available energy appears to be contributing to population decline in late spring of some years. Death may occur directly by starvation since the average level of body fat in carcasses we collected before and during snowmelt was about 2%, the level at which starving lemmings die in the laboratory. The continued decline of populations through the summer can not be related to lack of available food, but there may be continuing effects of earlier undernutrition. Reproducing females require considerably more energy than non- reproducing females. So insufficient forage relative to energy require- ments may also explain why there are fewer pregnancies and smaller lit- ters during the winter breeding season. The level of available graminoids then is one-tenth that of midsummer. Lemmings may not be able to maintain the necessary rate of forage intake on such a dispersed re- source. Recent experiments show that rate of forage intake increases line- arly with forage availability (Batzli et al., in press). Although there is considerable variation from year to year and site to site, tundra graminoids are often low in calcium and phosphorus (Table 10-6). Batzli (unpubl. obs.) found that the temperate microtine Microtus californicus does not reproduce well when fed a diet with levels of calcium similar to the highest amounts found in tundra graminoids. Lemmings, however, perform well on such forage; and metabolic experi- ments have shown that nonreproductive subadults are in slight positive balance for all minerals except sodium. But when lemmings are repro- ducing and when the nutrients in forage are at their lowest levels, this may not be true. Laboratory animals fed natural forage ate more ad libi- tum than was required to meet their energy needs, and fat levels rose to 15 to 20*^0 of body weight. In the field, where energy requirements are greater, fat averages only 3.5°7o of body weight. Apparently the ability to process large amounts of vegetation, which is related to low digestibility for energy, allows lemmings to do well on a diet that would not support temperate microtines. The simulation model of lemming nutrition (see Nutrition and Energetics) led to the same conclusion and indicated that reproductive success might be curtailed in years of poor forage quality. Schultz (1969) conducted experiments in which the nutrient status of tundra vegetation was changed. By fertilizing heavily he increased the 382 G. O. Batzli et al. protein, calcium and phosphorus levels in graminoids well above those in nonfertilized areas. Fertilization apparently increased winter reproduc- tion in 1968; there were about 75 winter nests ha"' in the fertilized area and none in the control areas. The effect continued through 1971, al- though by then it was less dramatic. Melchior (pers. comm.) reported 14 nests ha"' in the fertilized areas and 2 ha'" on the control plots. These re- sults suggest that, at least during some winters, reproductive perform- ance of lemmings can be stimulated by improving vegetation quality. In summary, lemming populations often increase up to a Hmit im- posed by their food supply and begin to decline when there is not enough food to meet energy demands. However, in summer primary production exceeds the lemmings' requirements, and lack of food cannot explain the continued population decline. Several lines of evidence suggest that lem- mings can survive on low quality forage because of nutritional adapta- tions, but the lack of nutrients may still reduce winter reproduction. Intrinsic Factors All the factors influencing lemming populations discussed so far are extrinsic, residing outside of the population itself. Several investigators have suggested that intrinsic factors such as behavior, physiology and genetics may be equally important and that social interactions and ag- gression increase with increasing population density. Christian and Davis (1965) proposed a hormonal imbalance to account for a population's de- cline. Chitty (1967) and Krebs et al. (1973) argued that some types of lemmings emigrate more or die sooner than others, thus changing the genetic composition of the remainder of the population. Changes in gen- otypic frequencies within the population are held to be responsible for changing reproductive and survival rates. Some work has been done on physiological changes in lemmings at Barrow. Mullen (1965) looked at blood glucose and formed elements of blood during four summers. There was no evidence of physiological changes associated with population density. Krebs and Myers (1974) found no evidence that physiological stress played a role in the produc- tion of microtine cycles. Andrews et al. (1975) reported changes in adre- nal activity and kidney disease associated with population density and climatic factors, but often these do not appear to be consistent or statis- tically significant. Thus the consequences of these endocrine adjustments for population dynamics of lemmings are unclear. Using a process of elimination Krebs (1964) concluded that genetic changes influenced lem- ming populations at Baker Lake, Canada, but gave no direct evidence. No studies of emigration or genetics have been done on lemmings at Barrow. The Herbivore-Based Trophic System 383 Summary Even though Httie can be said about the role of intrinsic factors in lemming population dynamics, it seems clear that extrinsic factors can exert a strong and overriding influence. In order for lemming popula- tions to reproduce and grow during winter, good quality forage must be available, the snowpack must be suitable and mammalian predators must o (0 Surface temperature Snowpack Soil moisture Nest sites Vegetation (quantity and quality) Forage consumption rate T" Decomposition Soil nutrients Predation rate ,. w O c Hormonal state Reproduc- tive rate Nutritional state Genetic state aEnxTiTH] Growth rate Survival rate Ei^ Dispersal rate Density (Biomass) I Aggression FIGURE 10-15. Diagram of relationships among factors influencing lemming population density. 384 G. O. Batzli et al. be scarce. Only when all three of these conditions prevail can a high pop- ulation be attained. A high population may begin to decline because of inadequate availability of forage, but only high mortality, resulting from predation or some other factor, and reduced recruitment can make the decline continue through the summer. Therefore, it does not seem that any single extrinsic or intrinsic factor can explain the population dynam- ics of lemmings. Rather, lemming populations respond to a number of factors that act and interact concurrently to determine the timing and amplitude of fluctuations (Figure 10-15). Dramatic fluctuations in the lemming population occur because the high reproductive output of lemmings during a single favorable winter allows them to increase their density by 100-fold or more, and these high levels cannot be sustained. The population usually peaks every three to six years, which implies that winters meeting all the necessary conditions occur every three to six years. Why favorable winter conditions are spaced at that interval is not clear. The underlying causes of the cyclic pattern are not likely to be random. But considering the variability in cycles during the past 25 years (Figure 10-2), and the relationships be- tween factors that influence population density (Figure 10-15), it seems clear that random factors, particularly weather, strongly influence the population dynamics of lemmings. HERBIVORY AT PRUDHOE BAY— CARIBOU Introduction The herbivore community in the Prudhoe Bay region is more varied than that at Barrow. There are caribou, willow ptarmigan and ground squirrels in addition to two kinds of lemmings. Each of the three small mammals requires a different habitat (Figure 10-16). Ground squirrels prefer stream bluffs, stabilized sand dunes near rivers, and pingos, where soil conditions allow construction of deep burrows. Within their home range the density and biomass of ground squirrels is high, but overall density is low (Table 10-3) because of their patchy distribution (Figure 10-16). Collared lemmings live around pingos, on stream banks and on polygonal terrain (Feist 1975, Batzli, unpubl. obs.). Brown lemmings prefer wetter habitats, as they do at Barrow. They are usually found in polygonal terrain with high-centered polygons and well-developed troughs where the vegetation is dominated by a variety of graminoids. The range of density and biomass of lemmings in the Prudhoe Bay region is relatively low (0.01 to 10 animals ha"' or 0.2 to 150 g dry wt ha"') com- pared with Barrow (Table 10-3). Much of the Prudhoe Bay vegetation is dominated by Carex and Salix spp., which grow on low, flat areas, often The Herbivore-Based Trophic System 385 FIGURE 10-16. Relative habitat utilization by herbivores near Prudhoe Bay. The numerals above the habitats refer to vegetation types identified by Webber and Walker (1975) for the Prudhoe Bay region. with large low-centered polygons, a habitat little used by lemmings at Barrow. Furthermore, the areas most favored by brown lemmings are also heavily utilized by caribou, whose trampling disturbs the habitat. Both factors may account for the modest lemming populations in the Prudhoe Bay region. Caribou represent the largest biomass of herbivores in the Prudhoe Bay region (Table 10-3). Because of their mobility caribou utilize a wide variety of landforms and vegetation types, and a large study area must be considered. During 1972-73 a resident population of 200 to 500 animals inhabited the Prudhoe Bay region, a 2340-km^ area of Coastal Plain bounded by the Kuparuk and Sagavanirktok Rivers in the west and east and by Prudhoe Bay on the north and the White Hills toward the south. These caribou constitute a portion of the Central Arctic caribou herd which has been identified in recent years by Cameron and Whitten (1979). During summer migratory herds of up to 3000 caribou may also pass through the region, and when they are under severe insect attack thousands may be concentrated in the coastal sand dunes associated with the river systems. 386 G. O. Batzli et al. Habitat Utilization by Caribou Migratory caribou move into the Prudhoe Bay region from herds that overwinter in Canada or south of the Brooks Range. The calving areas, located in the northern Foothills, consist of undulating terrain, frequently intersected by small streams. The first snowmelt north of the Brooks Range usually occurs in these areas (Hemming 1971). Snow is gone by the time caribou arrive for calving, and graminoids {Eriophorum spp.) have begun to grow, although green vegetation is very sparse. Calv- ing commences as early as 25 May and usually ends by 20 June (Lent 1966, Skoog 1968, White et al. 1981), and the main influx of migratory caribou reaches the Prudhoe Bay region by late June. The resident caribou calve on the Coastal Plain and the Foothills (Cameron and Whit- ten 1979). During the summer caribou graze either in small nursery groups of 2 to 10 cows with their calves and an occasional yearling, or in groups of 3 to 20 bulls and yearlings (White et al. 1975). Early grazing is concen- trated on exposed ridges and pingos that are dominated by dicotyledons. As the snow melts, caribou begin to graze on polygonal terrain and drained lake beds whose vegetation is dominated by graminoids and dwarf willows. Some of the poorly drained centers of low-centered poly- gons and lake beds do not dry until late summer, and the graminoid- dominated vegetation of these areas is utilized then. For most of the sum- mer, caribou prefer to graze on stream banks where the biomass and spe- cies diversity of the forage is high (White et al. 1975). The general sum- mer movement pattern appears to be determined by the phenological progression of vegetation types and associated changes in their nutri- tional status. As the season progresses caribou move back from the Coastal Plain to the Foothills, but daily movement patterns are less distinct due to the overriding effects of harassment by mosquitos {Aedes spp.) and warble flies (Oedemagena tarandi). During the warmest periods vegetation asso- ciated with standing water is avoided, presumably because it is prime mosquito habitat. Caribou gain some relief from mosquito harassment by grazing and walking into the prevailing wind. Or they move to the coastal sand dunes where it is cooler and windier than inland. Trail net- works generally join the preferred grazing areas with those areas where they seek relief. Where several trails join at the shores of larger lakes the soil becomes deeply rutted and completely devoid of moss and vascular plants. Once the temperature drops and insect harassment abates, cari- bou graze slowly through the vegetation towards preferred habitat. Little use is made of the trail systems at this time (White et al. 1975). In late October groups of caribou gather in herds of over a thou- The Herbivore-Based Trophic System 387 sand, and the animals migrate south to the wintering grounds during Oc- tober through December. Groups of the Central Arctic herd overwinter on the Coastal Plain and northern Foothills. Occasionally an early snow- fall in the Brooks Range prevents the migration of most of the caribou, and large herds overwinter on the Arctic Slope (Lent 1966, Hemming 1971). Some of the surviving calves and yearlings of these herds may become adjusted to overwintering on the Arctic Slope and add to the nonmigratory component of the northern caribou populations. Population Dynamics and Demograpliy The number of caribou grazing on the Coastal Plain varies accord- ing to migratory patterns, the number of resident caribou and the stage of their annual reproductive cycle. The number varies seasonally and an- nually in the Prudhoe Bay region. The estimates in Figure 10-17 are based on the total available area, not all of which is utilized by the cari- bou. Estimates based on the seasonal home range can be 5 to 15 times as high and give a good indication of habitat utilization during a season or year (Gaare and Skogland 1975). But the areas visited shift from year to 1 1 1 1 1 1 1 1 1 1 1 0.010 — A - / \/ Migrants - 0.008 o .c 6 c - / \ " >; 0.006 c 0.004 _o- L, — • — VResi dents 1972 r ^^^^.^^ 0.002 — 1973 f^ " * • 0 pCalving Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec FIGURE 10-17. The density of caribou in the Prudhoe Bay region during two years. The solid dots are based on observations of White et at. (1975); the open dots were calculated assuming reproductive performance equivalent to the Por- cupine herd as reported by Calef and Lortie (1973). 388 G. O. Batzli et al. year and estimates based on total available area may give a better indica- tion of the long-term population levels. The average resident caribou density in the Prudhoe Bay region was similar to recent estimates of the density of the Porcupine herd in 1972 (0.004 caribou ha''; Calef and Lor- tie 1973, Calef 1978), and the eastern Canadian Kaminuriak herd in 1973 (0.002 caribou ha"'; Parker 1972). These densities seem small, but the biomass they represent in the region is greater than that of lemmings, particularly when large numbers of migratory caribou move into the area (Table 10-3). In response to photoperiod, breeding activity commences with ag- gressive behavioral displays between adult males in mid- to late Septem- ber. Peak rutting occurs in late October to early November (Kelsall 1968, Whitehead and McEwan 1973). Estrus in female caribou begins in late September, and estrus cycles recur at 10-day intervals (McEwan and Whitehead 1972). The date of peak calving varies from year to year, which suggests that secondary factors such as nutrition and climate may modify the timing of both rut and parturition. Gestation lasts about 210 days. The effects of winter nutrition on the gestation period are uncer- tain; however, nutrition can affect birth weights of calves and milk pro- duction in lactating females (Skjenneberg, pers. comm.. White and Lu- ick, unpubl. obs.). Normally one calf is born; twins are rare. The age at which caribou first breed varies (Kelsall 1968). Female calves rarely breed, and frequently females are as old as 3'/2 years when they breed for the first time. Under good nutritional conditions up to 30% of females will conceive as yearlings, and caribou older than 4 years have peak pregnancy rates of 78 to 90% (Kelsall 1968). Female caribou breed until they are at least 16 years of age with little decline in fertility. Mature female caribou generally breed annually. However, when se- verely undernourished some females do not come into estrus, and lacta- tion continues through January (Reimers, pers. comm.). Under these cir- cumstances breeding in alternate years would be expected. Disease, as well as nutrition, may affect fertility. For example, brucellosis probably caused lowered pregnancy rates in the Western Arctic herd in 1961 (Lent 1966). In the Prudhoe Bay region, when calves were 4 to 6 weeks of age, 67% of non-yearling females had calves with them in 1972 and 31% had calves in 1973. Few caribou were observed in 1973, and the estimate may not be representative (White et al. 1975). The estimated number of fe- males with calves in the Porcupine herd to the east of Prudhoe Bay was 50% for 1972 (Calef and Lortie 1973). Caribou calves have a high mor- tality rate, which can be attributed to inclement weather, predation and accidents. By the end of the first year the cohort has normally been reduced 40 to 50% (Kelsall 1968, Parker 1972). Lack of data on age- specific mortality precludes the construction of life tables, but survivor- The Herbivore-Based Trophic System 389 ship curves have been compared. In reindeer herds with low early mortal- ity, the mean expected life span may be as high as 4 years, but a value of 2.5 to 3.0 years has been reported for the Kaminuriak caribou herd west of Hudson Bay (Parker 1972, White et al. 1981). Nutrition and Energetics During the summer months caribou graze on both graminoids and dicotyledons while moving slowly at 0.5 to 1.2 km hr''. Bouts of grazing are interspersed with periods of rumination during which they lie down. There are four to six grazing periods daily with a high degree of syn- chrony within each group. When not harassed by insects caribou spend 48 to 53% of the day grazing. Lactating females spend more time eating and less time search- ing and walking during a grazing period than do adult males, non-lactat- ing females, and yearlings (White et al. 1975, Roby 1978). The activity cycle is highly modified on days of heavy insect harassment when as little as 30% of the day may be spent eating. On days of intense warble fly ac- tivity, trotting and running may take up 25% of the day. From late June through early August caribou suffer insect harassment for up to 25% of the entire period, and attacks may last for over an hour (White and Rus- sell, unpubl. obs.). Thus, grazing periods are often interrupted, and trampling of the vegetation is increased. Although the mouth parts of the caribou are large, grazing is selec- tive. Rejected plant parts, particularly dead and coarse material, are ex- pelled from the rear of the mouth and drop back to the tundra almost continuously while the caribou is eating. On summer range, rejection may be as high as 20% of all vascular plants clipped; foraging and trampling may waste considerably larger amounts of lichens (Gaare and Skogland 1975, White and Trudell 1980). When feeding on willow, cari- bou nip leaf parts, buds and some current year's stems, but they exclude older stems and twigs. Dead material of low nutritional value forms 15 to 20% of the diet in the Prudhoe Bay region (White et al. 1975), indicating some inefficiency in the selection and sorting processes. Some selection of food results from selection of habitat type. Within the vegetation types there is further selection of plant species and parts. Early in the season an obvious preference is shown for the inflorescences of some dicotyledons, e.g. Pedicularis sudetica and Saxifraga spp. Be- cause of their low availability, inflorescences do not make up a large component of the caribou diet, but selective grazing on flowers may be important because of its influence on plant populations. Analysis of forage consumed by caribou in the Prudhoe Bay region showed that the dominant plant species in the diet were those that are 390 G. O. Batzli et al. a> to 0] O) o k_ CL 8o^ - Q 60- o c 40|- I 20h 0 80 60 40 20 Lemmus S Summer D Winter Rangifer Ml. I 0 ®Dry Maner(D) DEnergy (E) E=33 n D=3I Dicot Graminoids Moss 80 60 40 20 0 80 60 40 20 Summer Only m Caribou (C) □ Reindeer (R) I - D - E ■:• - i - 1 ft - * ft ft * ry Matter Only m h '^ = 57 fi R=49 1 i I 1 1 n ■■■ ■:•: is 1 1 Dicot Lichen Graminoids Moss Dead Material FIGURE 10-18. The composition and digestibility of diet for lemmings and caribou. The horizontal lines represent mean digestibilities for the overall diet. (After White et al. 1975, Batzli and Cole 1979.) most available in the vegetation type. These include the graminoids Erio- phorum angustifolium, Carex aquatilis and Dupontiafisheri, and several willows — Salix pulchra, S. arctica, S. ovalifolia and 5. lanata. This gen- eralized feeding was modified slightly by preferences for a few species of herbaceous dicotyledons and lichens (White et al. 1975). Dicotyledons made up a slightly higher percentage of the diet of caribou than did graminoids. Vascular plants contributed 92% of the diet (Figure 10-18). Mosses were eaten in such small amounts that intake may have been ac- cidental. Estimates of in vitro dry matter digestibility of hand-picked plant samples were used to calculate the mean digestibihty of dietary compon- ents (Person 1975, White et al. 1975, Person et al. 1980, White and Tru- dell 1980). Rumen inoculum was obtained from caribou and rumen- fistulated reindeer while they were grazing on tundra. The ranges in di- gestibility of individual species were large for graminoids (52 to 79%) and shrubby dicotyledons (21 to 71%), but the mean ± 1 SE digestibility of graminoids (54 ±3% to 64 ±3%, depending on species mixture) was not significantly different from shrubs (45 ±5%). Forb and lichen digestibility was similar to that of graminoids, but mosses had very low digestibility (Figure 10-18). Based on the relative occurrence of these dietary components mean summer estimates of digestibilities of all for- age consumed were 57% for caribou and 49% for reindeer. The lower di- The Herbivore-Based Trophic System 391 c UJ 60 Q 50 H 40 a. 20 40 60 Live Biomass, gdw m ' a. Live Biomoss, gdw m 80 -2 100 FIGURE 10-19. Consumption of forage by esophageal-fistulated rein- deer (a) and the percentage of the day spent eating (b) in relation to the available green forage. The dashed line represents sheep grazing on range similar in available biomass to that of the Prudhoe Bay region. The solid lines represent the extrapolation of the relationship to caribou. The shaded column represents available biomass of vascular plants in the Prudhoe Bay region in July. (After White et al. 1975.) gestibilities estimated for reindeer may reflect the effects of confining them to specific vegetation types. The potential digestibility of all material from the vegetation types in the Prudhoe Bay region varies considerably, but selection of green material would provide increased digestive efficiency. By following the phenological progression in vegetational development, caribou may be able to maintain maximum digestibilities of 57 to 63*^o throughout the summer (Person et al. 1975, White et al. 1975, White 1979). Whether or not caribou can select individual plants for digestibility is not known, but Klein (1970) suggested that they select for protein and minerals, particularly phosphorus. In vitro digestibility is inversely re- lated to nondigestible components such as lignin (White et al. 1975, Per- son et al. 1975). And selection for high digestibility should also provide higher intake of cell contents that contain most of the soluble protein and phosphorus. It seems likely that caribou select vegetational types and the plant species and parts within those types that are highest in general nu- tritional quality, but they avoid plant species and parts that contain toxic secondary compounds (White and Trudell 1980). Selection of vegetational types may also maximize the quantity of food eaten. Non-lactating female reindeer grazing on Carex-Eriophorum meadows were used by White et al. (1975) to estimate food intake. The availability of green vascular plants was an important factor controlling the rate of consumption of food (Figure 10-19a). Studies with grazing 392 G. O. Batzli et al. 200 r ? 160 X) ■o o o o Q 120 80 40 T 1 1 1 1 1 1 r -^ a A J 1 L 0 20 40 60 80 100 Liv/e Biomass, gdw m'^ FIGURE 10-20. Theoretical relationships between daily food intake and the availability of live biomass of vascular plants for lactating females (o), adult males (m), and non-lactating females (a). The relationships were calculated as the product of the relationships shown in Figure 10-19. The arrows indicate the daily food intake required to maintain body weight and, therefore, the amount of live biomass of available plants required before the caribou would gain weight. (After White et al. 1975.) sheep and reindeer have shown that the time spent grazing decHnes as available biomass increases (Allden and Whittaker 1970, Young and Corbett 1972, Trudell and White 1980). The curves for adult males and lactating females in Figure 10- 19b were extrapolated from the data for non-lactating females. Presumably, curves for caribou would be similar. The theoretical relationship between daily food intake and available plant biomass for reindeer can now be calculated (Figure 10-20, Trudell and White 1980). Food intake for lactating reindeer exceeds that for non- lactating reindeer because lactating females spend more time grazing. The theoretical relationships in Figure 10-20 are similar to actual obser- vations of food intake in relation to plant biomass reported for domestic sheep grazing in Mediterranean grassland systems (Arnold 1964, Arnold and Dudzinski 1967). Assuming that caribou and reindeer have similar grazing response functions, food intake of caribou in the Prudhoe Bay region during sum- mer is directly related to seasonal changes in standing crop of green vas- cular plants, which peaks in midsummer (Chapter 3). In particular, food intake declines rapidly when the availability of green biomass becomes less than 40 to 50 g m'^ Maximum food intake, when green biomass is greater than 50 g m'^ would be expected only for the month of July and early August in Carex-Eriophorum meadows. By selecting vegetation RADF The Herbivore-Based Trophic System 393 Relative Degree of Insect Relief Availability % of area In Vitro Digestibility % Peak Biomass gdw m"2 Biomass Ratio (liverdead) Sand Dunes ^ , , ^ I. Or Dryos Heatti/ Snowbed Dupontia Stream Meadow 60r 0 60r 0 80 O I 5r 0 I2(- 0 Enophorum Corex Meadow Meadow M E3 O _Ea_ j^_ Increasing Wetness of Habitat FIGURE 10-21. Characteristics of the vegetation types in the Prudhoe Bay region. The relative degree of insect relief is an index scaled to the wetness of the habitat. RADF (relative availability of digestible forage) = availability x digestibility x biomass x biomass ratio. (After White et al. 1975.) types highest in green biomass caribou could maintain a period of maxi- mum food intake from late June to mid-August. Each vegetational type has different attributes with respect to cari- bou grazing in the Prudhoe Bay region (Figure 10-21). The product of four measured parameters — availability of habitat, digestibility of for- age, peak green biomass and ratio of live-to-dead material — was used as a summary index of the relative availability of digestible forage for each habitat. The relative availability of digestible forage appeared to be posi- tively correlated with the distribution of caribou on days without insect harassment, particularly in groups of ten or fewer individuals (Figure 10-22a). Caribou group size was generally between one and ten individu- als following severe harassment as herds moved from relief areas to preferred grazing areas. On days with harassment, distributions changed noticeably, and there was no relationship to the relative availability of digestible forage (Figure 10-22b). 394 G. O. Batzli et al. ^ 60 « i 50 o .88 - a a. >< c o c o 3 o o o 3 (0 Q 0) 1 — r 1 — I — r |3l^-| Group Size (•)0-lO (o)>IOO 60 50 40 30 20 10 -b. 2,9 I- O -I — I — I — I — I — r Group Size 011-20 (0)21-99 T — I — I — r / / / / / / / / / — 2,9 16 ,12 / / 3,4 O 3,4 / I6« ^ — / OI2 1/ I I I I 1 I I 0 2 4 6 8 10 12 Relative Availability of Digestible Forage FIGURE 10-22. The rela- tionship of the distribution of caribou among habitats to the estimated relative availa- bility of digestible forage for those habitats when under no insect attack (a) or mild to se- vere attack (b). The numbers refer to habitat types given in Figure 10-16. (After White et al. 1975.) To investigate the possible influence of food quality on factors reg- ulating food intake a model of caribou rumen function (rumenmet) was constructed (White et al. 1981). The model interfaced factors regulating food intake with factors responsible for digestion and outflow from the rumen. The rumination time required to reduce the particle size of unfer- mented material, so that it could leave the rumen, limited the amount of time available for grazing, particularly when forage digestibility was low. Thus, although the availability of green plant biomass regulates the actu- al eating rate, the digestibility of the food controls the amount of time the animal must ruminate and the amount of time left for grazing. Re- sults from the model indicated that daily food intake may ultimately be regulated by the digestibility of the diet of caribou, a result well docu- mented for domestic ruminants (Baumgardt 1970, Baile and Mayer 1970). Seasonal changes in available plant biomass and digestibility of for- M The Herbivore-Based Trophic System 395 age (White et al. 1975) were used to generate estimates of food and energy intake and rumen function parameters with rumenmet. The number of grazing events, the time spent grazing, and the rate of ruminal volatile fatty acid production predicted by rumenmet agreed with field observations in the Prudhoe Bay region (White et al. 1981) when the rumen capacity was expressed as: Rumen capacity = 211^x18.9x0.57 = 226 kJ kg"' where 21 W is the dry matter capacity of the rumen in relation to body weight (g kg"'), 18.9is the energy content of the forage (kJ g"') and 0.57 is the proportion of energy assimilated. This capacity is about half the daily energy requirement of an adult non-lactating caribou during summer. At the end of summer, green biomass at Prudhoe Bay declines markedly to 10 g m'^ or less on all vegetational types except those on rims of low-centered polygons and pingos, where 20 g m'^ remains (Webber and Walker 1975, White et al. 1975). In most vegetation types a large amount of standing dead graminoid leaves is available. But this material is generally high in crude fiber and lignin, and expected digestibiUty would be only 30 to 40<^o (White et al. 1975). At this time caribou could continue to eat mostly green forage and maintain a relatively high digest- ibility, or they could consume larger amounts of dead material. RUMENMET was used to evaluate the effectiveness of these alternative tactics. If caribou continued to feed only on green material, then metab- olizable energy intake would be about 30*^0 of energy requirements. However, if both green and standing dead material were ingested, the model predicted that the daily metabolizable energy intake would be about 75% of energy requirements. Even if caribou could consume almost all available food on exposed ridges and at the base of feeding craters, the Prudhoe Bay region appears to be poor winter range. In interior Alaska good winter range for caribou is characterized by shallow snow and high lichen biomass (>100 g m"^; Hanson et al. 1975). Supplementation with frozen green graminoids is possible (Klein 1970, Hemming 1971), and maintenance requirements can probably be met. In the Prudhoe Bay region exposed ridges and pin- gos contain the highest lichen biomass. However, these areas constitute only 5% of the vegetated area (White et al. 1975), and the biomass of lichens is generally less than 10 g m'^ (WiUiams et al. 1975). Population energetics can be calculated from estimates of energy ex- penditure by age and sex classes within the population combined with estimates of productivity by age class. A flow chart was used to calculate energy expenditures of adult male and female caribou (Table 10-8). Average daily metabolic rate was computed using the model active (Bunnell et al., unpubl.), which was constructed to simulate the grazing 396 G. O. Batzli et al. TABLE 10-8 Flow Chart for Calculating Energy Requirements of Adult Grazing Caribou (kJ kg'° ''^ day'^) Standard Fasting Metabolism (FM) + Energy required for maintaining body function Resting Metabolism (RM) + Energy required for food inges- tion and digestion (energy cost of eating and specific dynamic effect) Maintenance Energy of Seden- tary Animal (MMr) Energy cost of locomotion and grazing activity Maintenance Energy of Grazing Animal (MMf), equivalent to Average Daily Metabolic Rate (ADMR) + Energy deposited in production (P) plus the energy cost of each process (tissue growth efficiency, Ec; efficiency of milk synthesis, E„; efficiency of fattening, Ef) \ Metabolizable Energy Require- ments (MER) Energy loss in urine and fer- mentation gases Assimilated Energy Requirement (AER) -I- Energy loss in feces (energy digestibility Df) Gross Energy Requirement (GER) or Ingestion FM, = 444 FM^ = 402 RM = 536 to 608 RM = 565 MM« = 2FM5 = 888 = 2FM»v = 758 McEwan (1970), caribou S = summer (1 June-31 Oct) Jf= winter (1 Nov-31 May) T. Hammel (unpubl.), reindeer White and Yousef (1978), White et al. (1975), reindeer AER = MER/0.82 Df = 0.57 to 0.63 GER = AER/D = 1608 to 1877 Kleiber (1961), general MMf,5,=2.8FM, MMf.H, =2.2 FM^ ADMR = 884 to 1244 Ec = 0.80 E„ = 0.70 Ef = 0.81Q„ + 3.0 = 0.50 to 0.70 MER = ADMR -hPc/Ec -I- P„/E„ + P,/Ef See text Agricultural Research Council (1965), general Qm = metabolizable energy of diet/gross energy of diet at maintenance Blaxter (1962), domestic ruminants White et al. (1975), caribou The Herbivore-Based Trophic System 397 behavior and activity patterns of caribou. The model indicated that ac- tivity significantly affected the average daily metabolic rate and that the energy spent in grazing and evading insects needed to be determined. Grazing involves almost continuous movement, so it was necessary to estimate the energy cost of walking on tundra (White and Yousef 1978). ACTIVE calculated that, compared to days with no harassment, the aver- age daily metabolic rate increased 1 .06 times during mild harassment and 1.6 times during severe harassment, and averaged 2.8 times the standard rate during summer. Energy expended on locomotion increases from \1% of the average daily metabolic rate on insect-free days to approxi- mately 60% during severe insect harassment. Because of their heavy insulation, thermoregulation is not a prob- lem for caribou in winter (White 1975), but the energy required for win- ter activities has not been determined. The assumption was made that energy expended in digging through snow was no higher than that ex- pended during mild insect harassment. The average daily metabolic rate was then calculated to be 2.2 times the standard fasting rate. Thus, dur- ing the year the average daily metabolic rate in non-lactating caribou var- ied from 884 to 1244 kJ kg"" ". Estimates of the production efficiency of each cohort of a caribou population were made from data of Krebs and Cowan (1962) and Kelsall (1968). In calves 3.0 and 4.6*^0 of gross energy intake were used for pro- duction by males and females, respectively. Efficiencies declined to ap- proximately 2% in animals between two and three years old and to zero in animals older than five years. Energy secreted in milk was taken as production of calves rather than a component of female production. The main reason for the low efficiency of production is the amount of energy required to support metabolism during winter, particularly from Decem- ber to June when productivity is negative, and the animals lose weight. The metabolic requirement for milk production during the first three weeks of lactation is very high— 10.5 to 12.6 MJ day"' or 40 to 50% of the average daily metabolic rate. This energy is required during May and June when primary production is negligible, and the predicted rates of energy intake would be low. Preliminary studies on reindeer grazing in shrub tundra in central Alaska indicate that the peak rate of milk secre- tion can vary from 0.8 to 2.2 liters day"' (5.4 to 14.7 MJ day"'), depend- ing on food intake (White, unpubl. obs.). Thus, lactation can be less than optimal, and the growth rate and survival of calves may be related to the diet. Data for grazing reindeer show that the growth rate of calves depends on milk production for at least 50 days. After 50 days milk pro- duction declines rapidly (Holleman et al. 1974), and females begin to rebuild the nutrient pools in their bodies (Cameron and Luick 1972). The general demographic pattern and data on production and gross intake were used to calculate population energetics for caribou (Table 398 G. O. Batzli et al. 10-3). At a density of 0.001 to 0.01 animal ha"' annual production was 0.24 to 0.78 MJ ha"' yr'. Energy retained, or productive energy, amounted to 1% of gross energy intake, 1.8^^0 of assimilated energy, and 2.1% of respired energy. Although calves make up only 15% of the population, they con- tribute 30% of the production. Seventy percent of the production is con- tributed by the 0- to 3-year-old caribou, which make up only 29% of the population. Thus, disturbances which affect the younger animals, e.g. adverse weather conditions, poor range quality or constant predation and harassment, have a marked effect on the population's productivity. Population energetics in caribou can be compared with previous cal- culations for elephants (Loxodonta africanus — Petrides and Swank 1966) and white-tailed deer (Odocoileus virginianus — Davis and Golley 1963). Average caribou biomass (14 MJ ha"') was much lower than that of elephants (297 MJ ha"') or white-tailed deer (54 MJ ha"'). Biomass and population turnover are reflected in absolute production of 0.78, 14 and 27 MJ ha"' yr"', respectively, for the three species. However, effi- ciency of production with respect to energy intake, assimilation and respiration was similar for all species. In addition to energy, caribou must receive sufficient nutrients from their forage to maintain a normal physiological state. Some insight into the relative importance of various nutrients for caribou can be gained by considering the degree to which nutrients in the diet must be retained. Similar calculations were made for lemmings. TABLE 10-9 Relative Concentration of Energy (kJ gdw') and Nutrients (mg gdw~') in Vascular Plants and Caribou Concentration in Concentration Concentration P/I*» summer forage (F*) in caribou (C)t factor (C/F) (%) Energy 18.9 28.9 1.5 1.0 N 25 72 2.9 1.9 K 20 6 0.3 0.2 P 2 15 7.5 5.0 Mg 3 1 0.3 0.2 Ca 2 30 15.0 10.0 Na 1 5 5.0 3.3 • Data as in Table 10-4. t Based on tissue values estimated for lOO-kg domestic cattle (Agricultural Research Council 1965). ** Estimates based on equation analogous to that used for lemmings (p. 356). Note: Retention of ingested nutrients in caribou represented by ratio of production to ingestion (P/1). I The Herbivore-Based Trophic System 399 Nearly all production of caribou derives from summer range, so nu- trient retention can be considered only in relation to summer forage. Re- tention of nutrients in the diet, the P/I ratio in Table 10-9, shows that ni- trogen, phosphorus, calcium and sodium must be concentrated strongly, but not as strongly as for lemmings (Table 10-4). Nevertheless, nutrient availability could be as important for caribou as for lemmings. No model has been constructed to explore the tactics open to cari- bou to maximize nutrient intake and to minimize nutrient loss by conser- vation and recycling processes as has been done for lemmings. However, if studies made on reindeer apply to caribou, then caribou may conserve nitrogen through urea recycHng (Wales et al. 1975), and this mechanism may conserve use of energy, water and glucose as well (White 1975). COMPARISON OF GRAZING SYSTEMS Now that the main features of the herbivore-based food chains have been described, it should be clear that microtine rodents and ungulates represent very different approaches to herbivory. This section will com- pare the main features of these two grazing systems and point out their consequences for the coastal tundra ecosystem as a whole. Population Characteristics Perhaps the most conspicuous difference between microtines and ungulates is body size. Although that may seem to be a trivial observa- tion, many life history characteristics of Lemmus and Rangifer, which determine characteristics of populations, appear to be a function of body size (Table 10-10). The ratio of body size between lemmings and caribou remains nearly constant from birth through adulthood; caribou weigh about 1500 times as much as lemmings. Both species need to grow by a factor of 20 from birth to adulthood, but lemmings grow relatively faster. Thus, lemmings double their birth weight in four days and reach adult weights in 120 days whereas caribou take four times as long. The association of higher meta- bolic and growth rates with smaller body size is well known (Kleiber 1961). Generally, the efficiency of growth does not change. Small and large animals produce the same amount of new tissue for each unit of energy digested, but small animals produce the new tissue more rapidly. Two other life history characteristics associated with small body size are a high reproductive rate and a short life span (Smith 1954). These relationships are dramatic in the lemming-caribou comparison. Because lemmings have a large mean litter size (seven) and a rapid 400 G. O. Batzli et al. TABLE 10-10 Life History Characteristics of Lemmus and Rangifer Ratio Lemmus Rangifer Lemmus/ Rangifer Body size Newborn (kg) 0.0033 5.0 0.0007 Weanling (kg) Individual 0.013 42 0.0003 Litter 0.091 42 0.0022 Adult (kg) 0.080 100 0.0008 Metabolic (kg" ") 0.15 32 0.0047 Growth Time to double birth wt (yr) 0.011 0.045 0.24 Total growth (kg) 0.077 95 0.0008 Adult/Newborn 24 20 1.2 Weanlings/Adult 1.1 0.42 2.6 Reproduction Litter size 7.0 1.0 7.0 Gestation (yr) 0.058 0.42 0.14 Lactation (yr) 0.041 0.44 0.094 Litters yr"' (max.) 9.0 1.0 9.0 Age at first reproduction (yr) 0.15 2.0 0.075 Survival Maximum life span (yr) 1.5 20 0.075 Expected length of life (yr) <0.1 to 0.3 2.9 0.069 Potential population growth* A„a, (yr) 1300 1.5 870 •Assume no deaths and that female lemming produces one litter per month (3.5 99 per litter), begins reproducing at two months and can produce nine months of the year. A female caribou produces one litter per year (0.5 99 litter"'). growth rate, the total weight of the Utter at weaning is 1.1 times that of the mother. A weanling caribou weighs only 0.4 as much as its mother. The relative investment of a female lemming in each litter must therefore be much greater than that of a female caribou. The same would be true during a life span because the number of litters produced is similar. Sur- viving to maximum age and reproducing at a maximum rate, a lemming could produce 14 litters and a caribou 18. During a normal hfe span both species might be expected to produce one or two Utters. A longer life ap- pears to compensate somewhat for slower development and lower repro- ductive rates in caribou so that the total production of litters is similar. It is the number of individuals per litter, resulting in the large biomass of the litter relative to the mother, that produces a greater investment in off- spring per female lemming. These results are consistent with those of Millar (1977), who concluded that litter size is the most important factor affecting the reproductive efforts of mammals in general. Just as the ratio of body size remains similar for lemmings and cari- bou at any stage in their life cycle, two important measures of survival. I The Herbivore-Based Trophic System 401 the maximum and the mean expected hfetime, are about 14 times greater in caribou than in lemmings. Remarkably, the usual ages at first repro- duction bear a similar relationship, which again suggests compensation between length of life and speed of development. Of course, the expected life span varies with any changes in the life table, so seasonal and annual differences will make its ratio between lemmings and caribou much more variable than that for maximum length of life. The consequences of life history differences can be immense when considered at the population level. For instance, if we assume that sur- vival is 100%, the potential population increase for lemmings during one year (A^^J is a factor of 1300 while that for caribou is only 1.5 (Table 10-10). High reproductive potential allows lemmings to respond quickly to temporarily favorable conditions, but such high local densities are reached that they cannot be maintained. Thus, population densities fluc- tuate wildly. Caribou cannot respond quickly to short-term changes in the environmental conditions, and their population densities change slowly in response to long-term changes in environment. Because of the great discrepancy in size of the major grazers, density figures (Table 10-3) do not give a good comparison of the two grazing re- gimes. The extreme changes in density of lemmings are modified some- what if the time of residence is included (line 2), but the best comparative figure of the amount of grazers present is probably biomass residence (line 4), which compensates for both body size and time of residence. Even this measurement shows that the annual grazing population on the coastal tundra at Barrow is much more variable than that in the Prudhoe Bay region. The annual biomass residence at Barrow may vary between twenty times less and five times more than at Prudhoe Bay. Energy flow through biomass is disproportionately large in lem- mings compared with caribou because of the small body of the lemming. Maximum annual respiratory rates of lemming populations at Barrow are 50 times those of caribou in the Prudhoe Bay region, and relative production rates are even higher. Lemmings do not produce young any more efficiently than caribou, but in relation to their size they produce more. Among lemmings, Lemmus produces more than Dicrostonyx because litters are slightly larger and breeding seasons are longer (Batzli 1975a). The result of these relationships is that even in the Prudhoe Bay region, where caribou account for most of the biomass of grazers, the respiration and production of lemmings is often greater than that of cari- bou. Population efficiency and turnover time reflect the same relation- ships; lemming production efficiencies are greater than those of caribou, and turnover times are about 40 times shorter (Table 10-3). 402 G. O. Batzli et al. Food Consumption and Foraging Patterns The grazing regime imposed by these populations depends not only on their biomass but also on their rate of food consumption and on the composition of their diet. Rates of food consumption must allow at least sufficient assimilation to supply energy requirements. Because of their high energy requirements small mammals often put more grazing pres- sure on tundra vegetation, even in the Prudhoe Bay region, than do cari- bou. Although Dicrostonyx and Rangifer have similar assimilation effi- ciencies (~0.55 to 0.65) Lemmus has much lower efficiency (0.33), which further increases ingestion. The gut capacity of Lemmus for digestible nutrients is 60% that of Rangifer in relation to body weight (Table 10-1 1), hence lemmings must fill their stomachs more often. However, for their size lemmings eat much faster, and the net result is that they need to spend only 20% as much time eating as caribou when forage is easily available. Interest- ingly, although the absolute turnover time of gut contents for lemmings is much less than for caribou, when corrected for metabolic weight (pro- portional to W'^^^ according to Kleiber 1961), the relative turnover times are almost equal. The distribution of grazing differs in its timing as well as in its inten- sity. Grazing by Lemmus becomes most intense during winter and at snowmelt when the vegetation lies dormant. Caribou migrate, and most of them leave the Prudhoe Bay region during winter, so the grazing pres- sure from caribou is highest during summer when plants are growing. The species and parts of plants taken by the grazers also vary. Lem- mus takes primarily graminoids with a supplement of mosses. During TABLE 10-11 Nutritional Characteristics of Mature Lemmus and Rangifer, Assuming Body Weights of 80 g and 100 kg. Respectively Ratio Lemmus Rangifer Lemmus /Rangifer Stomach or rumen capacity 125 226 0.56 (J g ' body wt) Fillings (no. day"') 14 2.2 6.4 Turnover time (hr) 1.7 11 0.15 (hr kg" '') 3.2 3.5 0.91 Maximum eating rate (g min ') 0.14 6.0 0.023 (gkg-""min~') 0.93 0.19 4.9 Foraging time (min day ') 160 740 0.22 Note: Energetic calculations are for summer (15°C). The Herbivore-Based Trophic System 403 summer caribou take equal amoums of the aboveground parts of dicoty- ledons and graminoids, about 15 to 20% of which may be dead material. During winter caribou in the Prudhoe Bay region probably take mostly dead plant material. Dicrostonyx specialize on dicotyledons, mostly Salix spp., throughout the year, but graminoids form a supplement of about 10 to 20"7o of their diet (Batzli 1975a). Ground squirrels eat primarily vegetative and reproductive parts of dicotyledons (Batzli and Sobaski 1980). The combined effects of these diets are that graminoids are most heavily grazed on the coastal tundra at Barrow, while dicotyledons usual- ly receive equal, if not more, grazing pressure than graminoids in the Prudhoe Bay region. Impact on Habitat Five major characteristics of the grazing systems in the coastal tun- dras at Barrow and Prudhoe Bay can be compared (Table 10-12). Be- cause small mammals at Barrow have been more intensively studied dur- ing the International Biological Program and because the literature on TABLE 10-12 Comparison of Mammalian Grazing Systems in the Coastal Tundras at Barrow and Prudhoe Bay Characteristic Barrow Prudhoe Bay 1. Diversity of grazers 2. Dominant grazer a. Biomass b. Consumption 3. Grazing pressure a. Annual b. Seasonal 4. Forage taken 5. Major impacts of grazers a. Microtopography b. Vegetation c. Soil Two species One family One order Lemmings (Lemmus) Lemmings (Lemmus) Light to heavy (cyclic) Winter > summer Graminoids and mosses Burrows, runways and hummock formation. Favor graminoids and in- crease productivity. Speed nutrient cycling, in- crease depth of thaw and change dispersion pat- terns of nutrients. Four species Three families Two orders Caribou (Rangifer) Lemmings (Dicrostonyx and Lemmus) and caribou Light to moderate (stable) Winter = summer Graminoids and dicotyledons Trampling, trails and burrows. Undocumented. Undocumented. 404 G. O. Batzli et al. them is more extensive (see Batzli 1975a for review), ideas regarding their impact on tundra are more complete. The impacts can be considered in relation to three interacting components of the ecosystem, viz. microtop- ography, vegetation and soils. While some inferences can be drawn re- garding the Prudhoe Bay region, they are largely speculative. Lemmings at Barrow construct their burrows on elevated sites, e.g. centers of high-centered polygons or rims of low-centered polygons, which have favorable drainage. As a result, much of the drier tundra is riddled with burrows, and barren areas are formed at burrow entrances by deposition of soil. Runways connect the burrows, and hummocks often develop between runways owing to erosion of fine materials that are not stabilized by vegetation. In areas denuded by frost heaving, there seems to be progressively greater hummock development associated with increasing vegetation growth and lemming activity. Unfortunately, this sequence has not been documented by long-term observations in one area. In the Prudhoe Bay region, where lemmings are less abundant, their impact is not as clear, although burrows and runways may be conspicu- ous in some areas. Ground squirrel diggings on river banks, sand dunes and pingos do create dramatic series of holes and mounds, but ground squirrels occupy only a small portion of the tundra. Caribou generally spread out as they graze, so compaction of soil caused by their trampling is not obvious except for systems of trails leading to sand dunes where they seek refuge from insect attack. Of course, the effects of grazers on microtopography presumably are reflected in the soil characteristics, such as soil temperature and depth of thaw. Exclosure studies at Barrow, originally started in 1950, indicate that the elimination of lemming grazing causes several changes in vegetation (Batzli 1975b). At well-drained sites, carpets of mosses and lichens devel- op and graminoids become sparse. In low, wet sites graminoids continue to dominate, but standing dead material accumulates and productivity declines. Apparently, heavy lemming grazing disrupts mosses and li- chens, which recover slowly. Graminoids, however, have their meristem- atic tissue under the moss layer and can replace shoots rapidly by draw- ing upon reserves in underground rhizomes. Chronic grazing by caribou during summer may have less effect on vegetation than the intensive grazing during winter and spring when lemming populations are high. Grazers can affect vegetation indirectly as well as directly (Batzli 1975b). The bulbet saxifrage, Saxifraga cernua, appears to be concen- trated around old lemming burrows and trails as does an acrocarpous moss, Funaria polaris (^. Murray, pers. comm.). Saxifraga is probably there because lemmings disperse their sticky bulbets, while Funaria may simply specialize on lemming feces as a substrate. Herds of caribou may have significant local effects owing to trampling and deposition of man- The Herbivore-Based Trophic System 405 ure (Bee and Hall 1956, Steere, pers. comm.), but this impact has not been measured nor have long-term patterns been studied. The moss Voitia hy- perborea is associated with musk ox and caribou dung (Steere 1974). Heavy mats of the grasses Arctagrostis, Alopecurus and Calamagrostis cover tops of sand dunes and pingos occupied by ground squirrels, ap- parently in response to disturbance and manuring. Caribou and ground squirrels also may influence the reproductive success of some dicotyle- dons, particularly Pedicularis and Saxifraga, since they seem to be espe- cially fond of their flowering heads. A particularly interesting interaction between lemmings, soil and vegetation may be occurring in areas with low-centered polygons. Micro- topography produces large differences in soil moisture within a few meters. The deepest portions of polygon troughs may contain water all summer long, whereas the basins of low-centered polygons contain water only in early summer, and their rims are never submerged. Soil organic matter is greatest in the basins (bulk density < 0.5 g cm"') and least in the troughs (bulk density > 1 g cm''). Exchangeable phosphorus is greatest (240 Mg g'') in the troughs and least (90 ^g g"') in the basins (Barel and Barsdate 1978) Differences in soil conditions produce different vegeta- tional communities. Graminoid shoots are most robust and dense (^^3000 m"') in troughs and least robust and dense in basins (1000 to 1500 m"^). Phosphorus concentration in plant tissues and plant production are both highest in troughs. Decomposition and, therefore, nutrient cycling ap- pear to be most rapid in troughs because that is where production of organic matter is greatest and accumulation is least. The highest activity rates of bacteria and the highest standing crops of soil invertebrates also occur in troughs, thus accounting for high rates of decomposition. Lem- mings also concentrate their activities in troughs, apparently because the most palatable and nutritious food is concentrated there. All these observations are consistent with one another, and they allow the microtopographic units to be ranked in order of decreasing bio- logical activity: troughs, rims and basins. The one factor that seems most likely to account for the differences in biological activity is soil phos- phorus. Higher levels of phosphorus allow greater production of more nutritious vegetation, which stimulates both decomposition and herbi- vory. Moisture conditions may also influence decomposition, and the troughs maintain greater soil moisture during the warmest part of the summer. But why should phosphorus be concentrated in the troughs? One explanation might be that phosphorus is leached from the rims of low-centered polygons to the troughs. But if this were the only factor, phosphorus would accumulate on both sides of the rims, in the basins as well as the troughs. Then the basins and troughs would be expected to show similar levels of available phosphorus, which is not the case. A se- cond possibility is called the nutrient-transport hypothesis: that lemmings 406 G. O. Batzli et al. transport nutrients from both basins and rims to troughs. Polygons form in drained lake basins, which have relatively uniform topography and sediments and, therefore, an even distribution of soil nutrients before polygon formation begins. The nutrient-transport hy- pothesis provides an explanation for the development of the current pat- terns of biological activity as polygonal ground develops. The sequence of events can be hypothesized as follows. Troughs develop over ice wedges, and the wedges continue to ex- pand to produce the rims that surround the basins (Figure 1-10). Drain- age is impeded in the central basins and the deepest parts of the troughs. Although the basins of low-centered polygons hold water during early summer, they are higher and relatively drier than the troughs by midsum- mer. Snow cover is deeper in the troughs during winter. Lemmings place their winter nests in troughs where deeper snow improves the microhabi- tat. Foraging lemmings move out from the troughs under the snow, but most feces and urine are deposited in the troughs near the nests. Nutri- ents accumulate in the troughs as a result of lemming activities. Nutrients are depleted in the basins where lemmings remove forage but deposit few wastes. Summer (June through September) burrows and nests are located on the relatively dry rims. The rims attain an intermediate nutrient status because they are the site of summer nests and near the winter nests. Ac- cumulation of nutrients is associated with higher primary production and higher concentration of nutrients in the trough vegetation (Barel and Barsdate 1978, Tieszen, pers. comm.). Improved forage reinforces the preference of lemmings for troughs. Nutrient depletion has the opposite effect in the basins. Increased activity of decomposers and soil in- vertebrates occurs in response to higher quality of litter and concentra- tion of soil nutrients in troughs (see Chapter 1 1 for details). Accumula- tion of organic matter in the soil is slowed, and rates of nutrient cycling increase. Again, the opposite trends occur in the basins. Although these events have been presented sequentially to empha- size causal relationships, all occur simultaneously. The result is a slow transition from relatively homogeneous distribution of soil properties and biological activity in drained lake basins to the marked spatial heter- ogeneity seen in polygonal terrain. According to the nutrient-transport hypothesis, spatial differences in biological activity are largely a result of different availability of nutrients. The pattern of nutrient availability is imposed by activities of lemmings. Unfortunately, we do not have sufficient data to test the nutrient- transport hypothesis. The trends in soil properties and biological activity associated with polygonal terrain have been used to construct the hy- pothesis and cannot be used to test it. Ultimately, the causal links will need to be tested by field observations and experiments. In the absence of The Herbivore-Based Trophic System 407 TABLE 10-13 Nutrients (kg ha~^) Accumulated During a Standard Lemming Cycle (3 Years) as Calculated by a Simulation Model of Nutrient Transport by Lemmings Habitats same Hab itats different N. P Ca N P Ca High plant nutrients Troughs 0.25 0.05 -0.07 -0.17 -0.33 -0.58 Rims -1.03 -0.24 -0.28 -0.51 0.07 0.11 Basins -0.32 -0.07 -0.08 -0.39 0.05 0.04 Carcasses 1.40 0.34 0.54 1.40 0.34 0.54 Low plant nutrients Troughs -0.32 -0.08 -0.04 -0.31 -0.22 -0.61 Rims -0.58 -0.13 -0.17 -0.43 -0.02 0.14 Basins -0.16 -0.04 -0.05 -0.32 -0.02 0.04 Carcasses 1.40 0.34 0.54 1.40 0.34 0.54 Note: Four cases were run: with nutrient levels in forage high or low and with nutrient levels in forage the same or different in the three microhabitats at the beginning of the run. such data, we have tried to determine the feasibiHty of the hypothesis by using computer models of lemming population dynamics and nutrition to calculate deposition of urinary and fecal nutrients in the various mi- crotopographic units. The activity of lemmings in the various microtopo- graphic units was distributed according to time spent in nests and per- centage of shoots clipped in each unit. We assumed that excreta are formed and deposited continually. Thus, if lemmings spend 50% of their time in their nests, 50% of their excretions would be deposited in troughs, and the remaining amount would be distributed according to the amount of time spent foraging in each microenvironment. If the fig- ures are adjusted for the relative area of each unit, the removal and depo- sition of nutrients can be put on an areal basis. The results of these simulations indicate that net transportation of phosphorus from rims and basins to troughs only occurred when plant nutrient concentrations were high and the same in all habitats at the be- ginning of the run (Table 10-13). This would be the situation as poly- gonal development began in a recently drained lake basin. In all cases the concentration of nutrients in lemming carcasses was sufficient to alter the results, depending on where those carcasses were deposited. Most carcasses found on tundra are in the troughs, near or in 408 G. O. Batzli et al. winter nests. This would increase nutrient deposition in troughs. Many, perhaps most, carcasses are taken by predators. Weasels deposit their scats near the winter nests of lemmings, but avian predators regurgitate pellets on higher ground, the rims of low-centered polygons and the cen- ters of high-centered polygons. Deposition of pellets would therefore add nutrients to rims, and this would counteract the trend of greater nutrient removal from rims than from basins (Table 10-13). The total movement of phosphorus to troughs, assuming half of the carcasses decayed in troughs, would be 0.2 kg ha"' (20 mg m"^) over a three-year period, five times the average amount of soluble phosphorus found there at present. While this simulation does not prove that con- sumers move nutrients to polygon troughs, it does show that the nutrient- transport hypothesis is feasible. We do not know how effective predators are as nutrient-transport systems, but snowy owls spend long periods of time on favorite centers of high-centered polygons (owl mounds). Extremely high levels of phos- phorus accumulate in the soils of these mounds, and grasses such as Arc- tagrostis, Calamagrostis and Poa dominate the vegetation there as no- where else at Barrow. Jaegers deposit their pellets on lower mounds, such as the rims of low-centered polygons, as well. The deposition of these pellets affects the foraging patterns of shorebirds. MacLean (1974b) has shown that female shorebirds must consume lemming bones when laying eggs to obtain enough calcium for their eggshells. Shorebirds, which nor- mally forage in lower areas, search on mounds for lemming bones. Thus, nutrient transport by consumers can affect other consumers as well as producers and decomposers. More observations, calculations and experiments are required to de- termine the effectiveness of consumers as nutrient-transport systems in tundra. If confirmed, these systems will be an important example of the major impact consumers can have on ecosystem structure and dynamics, much greater than that predicted by simple measurement of their biomass. Consideration of changes in tundra vegetation and soils after re- moval of lemming grazing provides additional insight into the effects of grazers on coastal tundra. Batzli (1975b) sampled exclosures near Barrow that had been in place for 15 years (Schultz 1964) and 25 years (Thomp- son 1955c). Vegetation had changed little in low, wet sites that had stand- ing water most of the summer. In mesic and dry sites, however, the net production of graminoid stems and leaves was almost twice as high in grazed areas as in the exclosures. Standing dead material and detritus were greater within the exclosures, suggesting that nutrient cycling had diminished in the absence of grazing. Reduced phosphorus in the soil solution under exclosed areas corroborates the hypothesis of decreased nutrient availability in the absence of grazing (Barel, pers. comm.). The Herbivore-Based Trophic System 409 BatzH (1978) reviewed evidence for similar effects of herbivores in other ecosystems. Although the effects may not be as dramatic, herbivory in grasslands and forests may also increase the rate of nutrient cycling and change the distribution of nutrients. SUMMARY A single species, the brown lemming {Lemmus sibericus), dominates the herbivore community at Barrow. The number of trappable animals per hectare increases to a peak of 150 to 250 every three to six years and may drop to less than 1 in the years between. A simulation model shows that a dramatic increase in the popula- tion can be produced by a slight improvement in the survival rate of adult females and their young. The population increases occur during those winters when the structure of the snowpack allows access to food, when forage quality is high, and when predatory mammals are scarce. The high reproductive potential of lemmings then allows the population to in- crease greatly before the snow melts. Only catastrophic mortality can explain the radical declines. Current evidence suggests that the mortality, at least early in the summer, is caused by overgrazing accompanied by a rise in the number of predatory birds. However, the effects of increased social interactions on the disper- sal and genetics of high populations have not been studied sufficiently. Examination of the nutrient-recovery hypothesis as an explanation for cyclic fluctuations in lemming density leads to the conclusion that it requires modification. Although changes in the nutritional quality of the vegetation may affect lemming populations, lemming activity does not appear to produce the short-term effects required to alter nutrient con- centrations in soil and plants as proposed. The herbivore community in the Prudhoe Bay region is more diverse and more stable than that at Barrow. Caribou (Rangifer tarandus) pro- vide the greatest herbivore biomass. The brown lemming and the collared lemming {Dicrostonyx torquatus) exist in about equal numbers, but their density is an order of magnitude less than at Barrow. Even so, lemmings may consume three to six times as much vegetation as caribou because their metabolic rates are higher and their forage is less digestible. Ground squirrels {Spermophilus parryii) are also important herbivores in more restricted habitats. Comparison of ungulate and microtine grazers reveals two very dif- ferent, but equally successful, suites of adaptation to herbivory on tun- dra. The short development times and large litters of the microtines give them a high population growth rate. In order to fuel its high metabolic rate a microtine must fill its gut often. But because it eats faster and has a 410 G. O. Batzli et al. faster turnover rate of gut contents than an ungulate, it requires less time for foraging. Caribou, on the other hand, being large and mobile, are less vulnerable to predators and can therefore spend more time foraging. During the summer they must make up for the undernutrition they suffer during winter. Their grazing patterns can be interpreted as an attempt to maximize their intake of high quality forage. But on some days harass- ment by mosquitoes and warble flies prevents them from obtaining ade- quate nutrients. Owing to their periodic abundance lemmings can strongly affect vegetational composition and production on the coastal tundra at Barrow. There is little evidence of this in the Prudhoe Bay region, however, and the effects of grazing are not conspicuous there. Both lemmings and cari- bou influence soil characteristics by burrowing, trampling and manur- ing. A proposed nutrient-transport hypothesis ascribes the uneven distri- bution of soil nutrients and biological activity in polygonized terrain to the redistribution of nutrients by animals. 11 The Detritus-Based Trophic System S. F. MacLean, Jr. INTRODUCTION The detritus-based trophic system is composed of animals that use energy only after it has passed from living components through the pool of dead organic matter. The system includes animals feeding directly upon dead organic matter (detritivores), upon microbial tissue (microbi- vores), or upon other animals (carnivores) (Figure 10-1). This chapter considers the abundance, energetics, and ecological function of animals in the detritus-based trophic system, and the contribution that they make to the decomposition of organic matter and cycling of mineral nutrients in the coastal tundra at Barrow. Even in years with high lemming populations, assimilation of energy by herbivores amounts to only about 6% of net primary production (Chapter 10). Another 13% of net primary production is returned to the tundra as feces, while about 80% passes directly to the dead organic mat- ter pool when unconsumed vegetation (including moss and vascular plant roots) senesces and dies. Thus, each year, 93 to 99% of the annual pri- mary production enters the pool of dead organic matter and becomes available to microorganisms and invertebrate detritivores, which form the first link in the detritus-based trophic system. The rate at which soil and litter organisms use this energy source is limited by the quality of the organic matter, the length and temperature of the period of biological activity, and other factors such as soil mois- ture and aeration. These influence both the population density and the rate of activity of individual organisms. The cumulative effect of these rate-limiting factors is seen in the large accumulation of organic matter in the soil which indicates that, by and large, the processes involved in the decomposition of organic matter have been more limited by arctic conditions than have the processes involved in the synthesis of organic 411 412 S. F. MacLean, Jr. matter by green plants. As a result, soil invertebrates live in an environ- ment that is energy-rich, with up to 80% of the soil dry weight consisting of organic matter (> 1675 J cm"') in the top 5 cm, where most of the ani- mals are found. As in other functional units of the coastal tundra at Barrow, the di- versity of animals in the detritus-based trophic system is low compared to that found in more temperate and tropical ecosystems; however, the di- versity of soil invertebrates is not as limited as is that of herbivores. Some taxa that are important soil organisms in other ecosystems are missing altogether (earthworms, isopods, millipedes, ants, termites) or are poorly represented (beetles) in the fauna of the coastal tundra at Barrow. Other taxa, for example mites (Acari), springtails (Collembola), flies (Diptera), and enchytraeid worms (Enchytraeidae), show only a modest reduction in diversity compared with temperate ecosystems, and commonly are quite abundant in the coastal tundra ecosystem. Invertebrate carnivores are few; the most conspicuous are the preda- tory cranefly larvae, Pedicia hannai, and beetles of the families Cara- bidae and Staphylinidae. Only two families of spiders, Linyphiidae and Lycosidae, are found, the latter with only a single species. During the summer months soil invertebrate communities support an abundant and diverse group of breeding birds, especially shorebirds or waders. ABUNDANCE AND BIOMASS OF SOIL INVERTEBRATES The array of microtopographic units that compose the coastal tun- dra was described in Chapter 1 , Sample plots for the study of soil inverte- brates were established in representative units of polygonal and meadow terrain (Table 11-1). The characteristics of the study areas and more de- tailed reports of the data are given by Douce (1976), Douce and Crossley (1977) and MacLean et al.(1977). This discussion draws heavily from these data, and emphasizes higher taxonomic categories. Community or- ganization at the level of species is considered by Douce (1976), Douce and Crossley (1977), and MacLean et al. (1978). Energy budgets are cal- culated based upon abundance and biomass estimated in 1972, for which data were most complete. Large differences among microtopographic units are apparent in mean abundance and biomass of the major faunal groups during the pe- riod of biological activity (Table 11-1). A two-way analysis of variance, using density as the dependent variable and plot and sample date as inde- pendent variables, was performed for free-living and plant-parasitic nematodes, Enchytraeidae, Collembola, and three major suborders of soil Acari (Prostigmata, Mesostigmata, and Oribatei). Location made § o O <4j •«: ■*■*» "*^. O to ■fci ••i* s: :::) a: o •*^ •»^ 53 -•«. 5U (SO .^ :i^ i •■^ "S: ^ 53 2« "S^ 0 C" ^^ ■^« 1 0 .^ 1^ ^ s s 0 ^ 'C J3 :^ 0 C3 5^ s: •5G s s § ^ 0 V. ?3 0 s 0 0 ^ ^ '^ ■§ 0 o« to s: to Q !3 !^ o> !^ 0 ^ "^ S t3 8 5^ ■I s 1 = ^(2 sa < oc E ns ^^ ■0 7 0 ■0 c5 c E oil 0^ a. vO ^ «^ ^ tjN rvj r*-) oC 0 (N — r^ >»■ r- oc >C t NO (N rsj I^ — r*-i — (^ *N '^ Vi O r- — \C OC sC rsj rg Tf — — r^ T q (N — ■* r^ *y~t — »/~i d V-. i d g m — o — — E o E o OS ■ - 3 o S o ■a -a E - c ■^ c o 'J QJ 00 a 00 00 > s c =1 >- =1 >> >■ ^ ^ o ~ o o ,0 Q a. Q a. a. U '.CO _o c o C 0 D. 3 •a '7, r-- re r^ 3^ i_ re '■' >i >. 0 ■a 3 ;/; 1/5 t/: 0 Urn CI '0 > '7. 0 ■0 CI 5 s CJ CI d 3 E 0 It 0 Q 5 >. 5 C 'c X Cl 0 '-» ^ •" ra ■a re >< a> c 0 0 0 IE c. ,ti >, 0 1) * ■a 3 ^ CI T3 I/: CI 00 Q CI c J= re JZ E E 0 ^ ■„_ 3 t/5 t/5 0 ^ M 3 *re >. c c M p ? 0 E "00 0 ^. 0 ~ re I/; re XI CI CI E E 0 TJ ^— OJ '*- ^ X ^ '■0 ^ CI 3 0 ^ -^ re 'J CO ^ .3 -o E > L/i c 0 5: CI ra 15 0 0 ■a 1 ■■^ 0 CI c ^ -0 3 ra CI Q b. SI U -a ;/5 1» c C 3 ^ re CJ c^ ^ CQ CI > Lrt (— L_ 1- w. (Lf 11 Cl 00 X3 XI T3 03 E E 0 Z Z h- < 413 414 S. F. MacLean, Jr. the greatest contribution to total variation in five of the seven cases. The exceptions were mesostigmatid mites and Collembola, in which within- sample variation was greatest. Sampling date made a relatively small contribution to total variation, indicating that spatial variation is more important than temporal variation in determining the abundance of tun- dra soil invertebrates. A similar pattern is seen in the soil microflora (Chapter 8). Mean annual density of nematodes, estimated by soil sieving fol- lowed by concentration through sugar solution centrifugation, varied be- tween about 50,000 individuals m"^ in the basins of low-centered polygons and 724,000 m'^ in the mesic meadow. These values are very low, but have been confirmed by the use of three different extraction procedures. In a review of data on nematode density and biomass in a va- riety of terrestrial ecosystems, Sohlenius (1980) found that only in deserts were mean values below one million individuals m ^ with coniferous forest, deciduous forest, and temperate grasslands averaging over three, six, and nine million individuals m"% respectively. Procter (1977) re- ported densities in the range of one to almost five million individuals m"^ in the tundra of Devon Island, N.W.T., Canada. Thus, the low densities found at Barrow are not characteristic of tundra. As at Barrow, the wet meadow at Devon Island supported the lowest density of nematodes. Trophic function of nematodes was determined by examination of mouthparts. Although abundance varied by a factor of 15 among sample points, trophic structure of the population was remarkably constant. The free-living nematodes, which are largely bacterial and algal feeders, were most abundant overall. The plant-parasitic nematodes became relatively more abundant on the drier units, and were dominant on the mesic mea- dow. Predatory nematodes made up a small proportion of the popula- tion in all cases. The abundance of all three nematode groups was in- versely correlated with soil moisture; the Spearmann rank correlation be- tween total nematode abundance and moisture was -0.65 {p < 0.05). Drier units contain more dicotyledonous plants, many of which are mycorrhizal (Miller and Laursen 1978). This is particularly so for Salix on the mesic meadow. The much greater abundance of plant-parasitic nematodes here may indicate that the root systems of dicotyledonous plants are more susceptible to the attack of nematodes than are roots of the graminoids that dominate wetter areas, or that the nematodes are feeding directly upon mycorrhizal fungi. Biomass of nematodes was not directly determined. Based upon studies of nematodes at a variety of other locations (Sohlenius 1980), a mean biomass estimate of 0.1 ^g dry weight per individual was used to approximate population biomass (Table 11-1). More elaborate functions relating density to biomass, for example based upon differences in trophic function, might be used, but hardly seem justified in light of the The Detritus-Based Trophic System 415 low abundance. These results indicate that nematodes are relatively un- important in the coastal tundra at Barrow compared with other ecosystems. The northern Coastal Plain near Barrow is totally lacking in earth- worms (Annelida: Lumbricidae), but is rich in smaller worms of the fam- ily Enchytraeidae. These worms are largely aquatic, living in the soil in- terstitial water, and are less abundant in the drier areas. Mean densities ranged from 1 1 ,000 and 13,000 individuals m"^ on the well-drained poly- gon top and rim to 94,000 worms m"^ in the moist polygonal trough. The Spearmann ranic correlation of enchytraeid abundance and soil moisture across the nine microtopographic units (Table 11-1) was positive and significant (r = +0.75; p < 0.05). Biomass of Enchytraeidae was determined from the distribution of body lengths of each species, using the geometric equations of Abraham- son (1973). Biomass is dominated by a large species of the genus Mesen- chytraeus, which had a mean individual biomass of 65 ^g dry wt, and mean population biomass of 1500 mg m"^ in the polygon trough. The biomass of this one species exceeded the sum of nematode, mite, and col- lembolan biomass in this habitat, and the total enchytraeid biomass ex- ceeded the sum of all other animal biomass in eight of the nine micro- topographic units sampled. Mean enchytraeid biomass across all units was 2100 mg dry wt m"'. Thus, Enchytraeidae achieve considerable abundance and biomass in the coastal tundra at Barrow and strongly dominate the soil fauna. Mean density of Acari ranged from less than 10,000 individuals m"^ in the basins of low-centered polygons to 83, (XX) m"^ in the mesic mea- dow. The rank correlation with moisture (r = -0.62; 0.10 >/? > 0.05) in- dicated greatest density in drier areas. This trend occurred in all three of the major mite suborders: Prostigmata, Mesostigmata, and Oribatei. The Oribatei were particularly lacking from the basins of low-centered polygons. In most ecosystems the Oribatei are the dominant group of Acari; however, the small prostigmatid mites are relatively abundant in tundra ecosystems (Behan 1978). Prostigmata comprised 16 of 37 species and 43% of all individuals in our samples (Table 11-2). The average indivi- dual weighed only 1 ptg dry weight (Douce 1976); hence, the numerically dominant Prostigmata contributed relatively little to biomass. The most abundant oribatid species, Liochthonius scalaris Forsslund (= L. sellnicki S. thor) is also very small, with adults weighing only 0.5 yig; however, the mean weight of oribatid mites was 4.0 Mg- Overall, the Ori- batei composed 48<^o of the density but 66% of the biomass of mites. The predaceous Mesostigmata are the largest of the mites, averaging about 6.8 Mg, and thus made a much larger contribution to biomass (18%) than to density (8%). Density and, especially, biomass of mites are generally low 416 S. F. MacLean, Jr. TABLE 11-2 Composition of the Mite (Acari) Fauna in the Coastal Tundra at Barrow Taxonomic Species Density Weight Biomass group (no.) (%) (ind m"-) (%) (Mg ind') (mg m') (%) Prostigmata 16 43 17.800 44 1.0 18 16 Mesostigmata 9 24 3.200 8 6.8 22 18 Oribatei 11 30 19.400 48 4.0 78 66 Astigmata 1 3 compared with temperate forest and grassland ecosystems. Density of Collembola varied between 24,400 individuals m"- in the basins of low-centered polygons and over 150,000 m"^ in the polygon troughs and moist meadows, and was greater than that of mites in all nine microtopographic units sampled, although the difference was small in the three drier units. Collembolan density showed no relationship with soil moisture. Biomass of Collembola was not determined directly. Population bio- mass was estimated using a standard value of 4 ^g per individual, taken from Petersen's (1975) observations of adult Folsomia quadrioculata and from Fjellberg's (1975) observations of F. diplophthahna in arctic- alpine Norway. These two species made up 70% of the total collembolan density in our samples (MacLean et al. 1977). Using this value, estimated biomass of Collembola exceeded 0.5 gdw m"^ in the most favorable areas. The Diptera strongly dominated the numbers, biomass, and diver- sity of tundra macroarthropods (Table 11-3). Three cranefly species (Tipulidae) are particularly prominent: Prionocera gracilistyla, a large species with larvae in the wettest habitats; Tipula carinifrons, a large spe- cies of mesic and dry areas; and Pedicia hannai, a smaller species with carnivorous larvae that are most abundant in wet meadows and the polygon troughs. Larvae of other species of lower Diptera (Nematocera), including especially midges (Chironomidae) and fungus gnats (Myceto- philidae and Sciaridae), were recorded at densities of up to 685 individu- als m"^ This must be considered a minimum estimate due to uncertain- ties of samphng and extraction efficiency (Healey and Russell-Smith 1970). A large proportion of the smaller larvae may have been over- looked, which would lead to a serious underestimate of density, but not of biomass. Rotifers, which are more commonly associated with freshwater hab- itats, occurred in all microtopographic units, with a mean annual density in the wettest meadows of 40,000 ind m"-. Tardigrades occurred at densi- tl %l s: I I O — •" - — .-^ ,„^ E s« C ■■ v ^ 96 9. E 3 S — ^. " ^ ""^ t ^_^ ** V r^ ^ ol E r- £ -T r~ ". O ". "• oc '"' -r-d r*^, r^ r4 -t 54 — — --, -T V. — r- «^. r^ "T •''. **■- rj:. ri 'Jl r i r i ■C — — -T d — r) to .■a ■II?! 2i£l a, I rl n r~ X ■y o- >/■, ■y -. ^ f* ft C-'' S M, t. M. ; M. a: 417 418 S. F. MacLean, Jr. ties up to 15,000 ind m"^ in the wet meadows; very few were found in the drier areas. In both cases these must be regarded, cautiously, as mini- mum estimates since no special effort was made to census these animals. Microtopographic units differ markedly in the abundance and com- position of their invertebrate faunas, even using the gross taxonomic units considered here. Total invertebrate biomass differed by a factor of about three between the most productive polygon troughs and the least productive basins and rims of low-centered polygons and tops of high- centered polygons, although these units are separated by only a few meters distance and about 20 cm of relief. We must be very cautious in referring to a "mean" or "average" unit and its soil fauna in the very heterogeneous coastal tundra at Barrow. Rather, the tundra is a repeat- ing mosaic of polygon troughs, rims, tops and meadows that are quite distinct habitats for soil invertebrates and microflora (Chapter 8). Even within microtopographic units, populations of all major soil arthropod groups were significantly clumped or aggregated in their pat- terns of dispersion. There was a significant tendency for coincidence in the aggregations of total mites and Collembola (MacLean et al. 1977), plant-parasitic and free-Hving nematodes, and prostigmatid and oribatid mites, indicating that these groups respond similarly to microhabitat suitabiUty or richness. A very consistent feature of the tundra soil fauna is the concentra- tion of animals in the near-surface horizons of the litter and soil (Figure 11-1). In both Acari and Collembola over 90% of the individuals Q. a> Q Percent of Population "■ "I s- K.J -c 5 J I L 100 0 lO-J Acarina 0 100 0 I Collembola 0 100 1 1 1 1 J Ped cia hannai '3 ^ 10 Q 15-U 100 0 -u — I 1 0 Free-living Nematodes 10 15-'-' 100 I J I I Plant- parasitic Nematodes 100 J I . I 1 '0| Enchytraeidoe FIGURE 11-1. The annual mean depth distribution of representative groups of soil invertebrates. % The Detritus-Based Trophic System 419 occurred in the top 2.5 cm, and over 98*^0 in the top 5 cm of the soil. Sixty-six percent of the total Enchytraeidae occurred in the top 2.5 cm, and 86% in the top 5 cm. The Nematoda showed the greatest tendency to occur at depth. Approximately 85% of the free-living and predaceous and 55% of the plant-parasitic nematodes were found in the top 5 cm. The occurrence of plant-parasitic nematodes below 5 cm seems related to the presence of dicotyledonous plants. Plant-parasitic nematodes were the only animals to occur in significant numbers in the mineral soil beneath the peat. Their small size may allow them to exist within the small pores of mineral soil. Depth distribution of Enchytraeidae showed a marked seasonal pat- tern (Figure 11-2); worms were concentrated near the surface at snow- melt, but moved to deeper layers by mid-season. After the first sample of the season, the three dominant enchytraeid species were segregated by depth. Individuals of the smallest species, Cernosvitoviella atrata, were found even below 15 cm in the tundra and, when sampling was con- cluded in late August, had shown no tendency to return to the surface. Over 50% of the populations of Mesenchytraeus sp. and Henlea perpu- silla were found in the top 2.5 cm in all but the late July sampling. Concentration of animals in the surface layer was greatest in the basins of low-centered polygons where well over 90% of the animal bio- mass, including 99% of the Collembola, 97% of the mites, and 89% of the Enchytraeidae, occurred in the top 2.5 cm. C. atrata, the deep- dwelling enchytraeid species, was entirely absent. Although soil invertebrate densities are conventionally presented as number per unit area, the number per unit volume more accurately 1 — ' — \ r- Henlea perpusilla Cernosvifoviella atrata • •^ •- J 1 I L. 10 20 Jul 10 20 Aug FIGURE 11-2. The seasonal variation in depth distribu- tion (percentage of the population in the top 2.5 cm) of the three dominant species of Enchytraeidae. I 420 S. F. MacLean, Jr. expresses conditions encountered by an animal living in a three- dimensional environment. For example, an average cubic centimeter within the top 2.5 cm of the polygon trough contained 4.8 nematodes, 2.2 enchytraeid worms, 0.85 mite, and 5.8 Collembola. Although the densities of soil invertebrates in some other ecosystems may equal or ex- ceed those of the coastal tundra at Barrow, it is unlikely that many have consistently greater concentrations of animals. The abundance and vertical distribution of soil invertebrates can be compared with the vertical profiles of soil temperature and moisture, plant biomass and production, microbial biomass and production, and dead organic matter. The seasonal course of soil temperature in several microtopographic units was presented in Chapter 2. Temperatures signif- icantly above air temperatures occur in the top few centimeters of the soil, but soil temperature declines rapidly with depth, and below 10 cm rises only slightly above 0°C even at mid-season. Low temperatures below 5 cm may contribute to the surface concentration of many soil in- vertebrate species. Conversely, winter season temperatures are lowest near the tundra surface. In temperate regions many soil invertebrates descend into deeper layers of the soil to avoid the frozen soil and cold. In tundra, this movement is prevented by permafrost. Soil animals could gain some protection by descending to lower depths, but to do so would result in a later onset of activity in the following spring. It appears that the advantage of the longer and warmer near-surface growing season outweighs the increased risk of mortality from winter cold, and soil ani- mals remain near the surface during the winter. In fact, Enchytraeidae must move toward the surface late in the summer season (Figure 11-2). Abundance of Nematoda, Enchytraeidae, and Acari all relate signif- icantly to the moisture ranking of the microtopographic units, suggesting that moisture or some correlated factor is important in determining their distribution. The Enchytraeidae and Nematoda are basically aquatic or- ganisms living in the soil interstitial water. The Enchytraeidae are rela- tively large, thus requiring larger water-filled pores and cavities in the soil. Increased soil moisture apparently increases the amount of habitat available to them, thus increasing population abundance. The Nematoda are much smaller, and can use the thin film of water that surrounds soil particles even in relatively dry soils. Since their abundance is negatively correlated with soil moisture some other factor must be involved. At high soil moisture levels the soil pore volume is filled with water, and anaerobic conditions may develop. The commonly anaerobic condi- tions occur in the 5- to 15-cm depth interval (Chapter 7). Shortage of oxygen could well contribute to the observed depth distribution of Bar- row soil invertebrates. The drier soils of high-centered polygons and rims of low-centered polygons probably have sufficient oxygen throughout the period of biological activity, and a larger proportion of soil inverte- The Detritus-Based Trophic System 421 brate populations are found below 5 cm in these than in other microtopo- graphic units. Soils of the meadows, troughs, and basins of low-centered polygons may become strongly anaerobic in early summer when satu- rated by snow meltwater, but less so, or even oxidizing, as the surface soil dries out in mid-season. The soil may again become anaerobic fol- lowing late summer rains. Seasonal changes in depth distribution of En- chytraeidae follow the same pattern (Figure 11-2), suggesting that Enchy- traeidae may use the resources occurring at depth only when the aeration of the deeper levels allows. Springett et al. (1970) showed that surface drying of peat soils in a British moorland resulted in downward move- ment of the enchytraeids Cernosvitoviella briganta and Cognettia sphag- netorum. In the present case, the fact that seasonal changes occur in even the wettest areas suggests that access to the deeper strata, determined by temperature and aeration, is responsible for population movements rather than exclusion from the surface layers by desiccation. The basins of low-centered polygons are characterized by very low densities and high surface concentration of invertebrates. This could be a result of anaerobiosis developing when the basins are flooded in spring by meltwater confined within the surrounding rims. Over 9Q^o of the in- vertebrate biomass in the polygon basin consisted of Enchytraeidae, which are known to be more tolerant of anaerobiosis than other inverte- brates, as indicated by the very high densities achieved in sewage beds. Little is known of the feeding habits of soil invertebrates, making it very difficult to discern the relationship of the animals to their food sup- ply. Progress will probably require detailed examination on a spatial scale much smaller than is reported here, and careful experimental and manipulative study. We can, however, attempt to relate the abundance of soil fauna in major habitat units to estimates of microbial biomass and productivity. From direct counts (Table 8-1), it appears that abundance of bacter- ia changes in the order rims » meadows > basins > troughs. Free-living nematodes are largely bacterial feeders, and their abundance (Table 11-1) shows a similar pattern, rims » meadows = troughs » basins, with the exception of the very low abundance in the basins. Fungal biomass is greatest in dry habitats. The abundance of soil Acari corresponds; however, the Enchytraeidae, which are probably the major soil fungivores, are most abundant in wet habitats. Fungal bio- mass is higher in the basins than in the troughs, but low in both; how- ever, fungal productivity, estimated as the sum of positive biomass incre- ments between sampling occasions (Chapter 8), is highest in the troughs. This could account for the abundance of soil invertebrates found in troughs. Soil algae may also help support the abundance of invertebrates near the surface of the wet meadows and troughs. In all microtopograph- ic units, abundance of soil invertebrates drops off much more rapidly 422 S. F. MacLean, Jr. E 6000 E u in O E o m □ 01 4000- o _ a. _ o o o o 6000 E in S 4000 o CD ^ 20001- ° -I o 2000 0> k. 0) '0 20 40 60 « 20 30 40 50 Current Year's Vascular Growth Accumulated Orgonic Motter b. o o o o o o ° o ■ g m"^ I 60 o 3 kg m" o E o 0) >- 40- c a> - O C. o _ o o o 0 20 30 40 50 Accumulated Organic Matter kg m' ,-2 FIGURE 11-3. The relationship of invertebrate biomass to current year's vascular plant growth (a) and total organic matter to 20 cm (b), and of vascular plant growth to total or- ganic matter to 20 cm (c). (After MacLean 1974a.) with depth than does biomass of microorganisms. Invertebrate abundance and biomass can also be compared to the amount and distribution of net primary production, which represents in- put of fresh substrate for heterotroph activity. Peak season aboveground vascular plant biomass was used as an index of annual input into the var- ious microtopographic units. This measure is available for five of the study plots in which invertebrate populations were sampled. Total inver- tebrate biomass shows a strong positive correlation with this index of pri- mary production on these five plots (Figure 11 -3a). In many ecosystems concentration of invertebrates near the soil sur- face maximizes their access to fresh substrate in the form of litter falling from a plant canopy to the ground surface. In tundra plants, however, the larger part of the annual primary production is invested in roots and rhizomes (Chapter 3). Billings et al. (1978) estimated the annual root turnover of the Carex-Oncophorus meadow as between 60 to 65 g m'^ yr"' and 90 g m'^ yr"'. Even the minimum estimate is in excess of annual The Detritus-Based Trophic System 423 aboveground production. In Eriophorum angustifolium, which makes the largest contribution to this total because of its annual root system, new roots occurred throughout the soil profile to a depth of 30 cm. New roots of Carex were concentrated between 10 and 20 cm depth, and roots of Dupontia were concentrated between 5 and 15 cm. Dennis (1977) found 62% of the belowground live plant biomass, including rhizomes, and 38% of the dead biomass in the top 5 cm of the soil. The depth distribution of plant-parasitic nematodes resembles the distribution of live plant biomass below the ground. All other inverte- brates show greater confinement to the surface layers than either biomass or production of belowground plant parts. Clearly, a significant part of the annual net primary production appears as growth below 10 cm in the soil and is very little used by soil invertebrates. Since soil invertebrates constitute a detritus-based trophic system, a positive relationship might be expected between the quantity of soil or- ganic matter at the base of the food chain and the abundance of organ- isms supported by this base, in much the same way that a rich plant bio- mass may support an abundance of herbivores. In fact the total biomass of soil fauna is inversely related to accumulated soil organic matter to a depth of 20 cm (Figure 11 -3b) and the hypothesis that a large detritus base leads to an abundance of animals in the detritus-based trophic sys- tem must be rejected. Most of the organic matter lies below the layers of abundant fauna. The soil animals are concentrated in the near-surface organic-rich horizons, but they have access to only a small part of the total pool of organic matter. All of the habitats sampled are highly or- ganic in the near-surface horizons, and no relationship between faunal density and organic matter between 0 and 2.5 cm or between 0 and 5 cm is evident. Annual net primary production in the coastal tundra at Barrow also shows a strong inverse correlation with accumulated organic matter (Fig- ure 1 l-3c). The most productive plots are characterized by little accumu- lation, hence rapid turnover of organic matter and recycling of nutrients. In the less productive habitats the annual production of organic matter and uptake of nutrients are small relative to the amounts tied up in ac- cumulation. Rate of organic matter turnover may be taken as an index of total microbial activity. The fact that it correlates poorly with microbial bio- mass (Chapter 8) indicates that much of the microbial biomass is inactive at any one time. Total invertebrate abundance, then, correlates positively with the input and turnover of dead organic matter and with microbial activity and productivity. Much of the annual input of dead organic matter and microbial pro- duction lies below the depth of significant invertebrate density; thus, po- tential food goes uneaten. This does not necessarily indicate that food is 424 S. F. MacLean, Jr. unimportant in determining the distribution and abundance of tundra soil invertebrates. Rather, it appears that physical factors, among which temperature, moisture, and aeration are prominent, determine access to the habitat and the food that it contains. Invertebrates reach their greatest abundance in the polygon troughs and meadows, where they are part of a syndrome that involves relatively little organic matter accumulation, rapid nutrient recycHng, and high rates of primary production. Low invertebrate abundance is associated with more organic matter accumulation and low rates of primary pro- duction, suggesting a system that is constrained by a low rate of nutrient release and recycling. This analysis does not show causation. A very large number of cause-and-effect relationships are doubtless included in these broad pat- terns. It would be of little value to debate whether an impoverished soil fauna is cause or effect of low decomposition rate; surely feedback rela- tionships make both points, in part, true. The same can be said for soil nutrient concentration and decomposition rate. These data do suggest that the activity of soil invertebrates is interwoven with the pattern of production and decomposition that characterizes the coastal tundra ecosystem. LIFE CYCLES OF TUNDRA SOIL INVERTEBRATES The abundance, seasonal dynamics, and energy requirements of in- vertebrates derive, in large part, from the life cycle characteristics of the species involved. Life cycle characteristics have been studied in the crane- fly species Pedicia hannai (MacLean 1973) and Tipula carinifrons (Cle- ment 1975), and in the three dominant enchytraeid species, Mesenchy- traeus sp., Henlea perpusilla, and Cernosvitoviella atrata. Life cycles of these groups differ in a fundamental way. The Diptera pass through four larval instars in the soil and then undergo a complete metamorphosis in a pupal stage to the adult form. The adults leave the soil to swarm over the tundra surface, where reproduction is quickly accomplished. In Enchy- traeidae development is gradual, and reproduction may occur over a con- siderable part of the total life cycle. There is no marked change in habitat or Hfe form associated with the onset of reproduction. Tipulidae The cranefly species require at least four years to complete larval de- velopment. The early instars are completed relatively quickly, while the final, fourth instar lasts about two years. As a result, the population at The Detritus-Based Trophic System 425 any time contains a high proportion of large, fourth-instar larvae. Aver- age biomass is high, akhough the ratio of productivity to biomass is low. The slowing of development that leads to multi-annual life cycles with overlapping larval generations is an important contributor to the high density and biomass of Diptera larvae in the tundra. In both cranefly species, respiration rate of larvae increases with temperature over the entire range observed: 0.5° to 20°C. Qio values cal- culated over this range are virtually identical: 2.34 for P. hannai and 2.35 for T. carinifrons, comparing fourth instar larvae. In contrast, the growth responses to increasing temperature differ. The growth rate of P. hannai increases with increasing temperature over the range of tempera- tures observed in the field. Growth was fastest in mid-season, when tem- peratures are highest. Thus, in P. hannai assimilation of energy must in- crease by a factor greater than the increase in respiration rate as tempera- ture increases; that is, the Q,o of assimilation is greater than 2.34. In con- trast, Clement (1975) found a distinct growth optimum for T. carinifrons at 4° to 5°C. Growth rate was reduced at temperatures above or below this optimum. Assimilation of energy does not increase to compensate for the increase in respiration at temperatures above 5°C. It follows from the differing growth responses to temperature that Tipula carinifrons is an obligate arctic species that is unable to complete development in warmer climates, while Pedicia hannai is a facultative arctic resident that might also occur in warmer climates. In fact, P. han- nai is known from a number of locations in northern Alaska, including Anaktuvuk Pass in the Brooks Range, Umiat in the northern Foothills (Weber 1950b), Meade River, and the Prudhoe Bay region (MacLean 1975b), areas varying widely in summer-season length and climate. T. carinifrons occurs extensively in the Soviet far north (Chernov and Sav- chenko 1965), apparently always in arctic coastal localities. In Alaska it is known only from the coastal tundra at Barrow and Cape Thompson (Watson et al. 1966), and from coastal tundra in the Yukon-Kuskokwim River delta. Following the long period of larval development, the adult life span is completed very quickly. Pupation begins in late June, and most adults emerge quite synchronously around mid-July (MacLean and Pitelka 1971). In both P. hannai and T. carinifrons females have very small, non-functional wings. In P. hannai the mouthparts, antennae, eyes, and legs of females are poorly developed. Males of both species are winged, but the wings are used only for feeble fluttering along the surface and they are incapable of sustained flight. The morphological reduction of females may lead to greater fertility (Byers 1969); the limited use of wings by males indicates that wings would probably be of small advantage to females anyway. Females of the third cranefly found in the coastal tun- dra at Barrow, Prionocera gracilistyla, do retain wings and are capable 426 S. F. MacLean, Jr. of flight on warm days. This species has the most restricted larval habi- tat, and thus may require more searching for oviposition sites. Adults do not feed. Copulation may occur almost immediately after emergence, and egg-laying commences soon thereafter. Clement (1975) hypothesized the release of a sex pheromone by nearly emerged (or emerging) females to attract males. Given favorable conditions, males are quite active. The weight-specific respiration rate of adult male T. car- inifrons is 3 to 4 times that of larvae at the same temperature, and the metabolic response to temperature (Q,o) is greater. All of this increases the likelihood of successful reproduction under the variable and unpre- dictable weather conditions of the arctic summer, a likelihood that is fa- vored by emergence into a high-density population. The advantage of synchronous emergence is further enhanced by predation. The abundant avian predators feed upon insect larvae early and late in the season, but switch almost entirely to adult Diptera, especi- ally craneflies, when they are available. During the main period of emer- gence the density of adult flies far exceeds consumption by birds, and the impact of predation is relatively low. As discussed below, predation is more intense upon individuals emerging into low density populations, early and late in the emergence period. The result is selection for syn- chrony of emergence. During the midsummer emergence period photoperiodic cues are weak, particularly for soil-dwelling pupae. It appears that the timing of emergence is controlled entirely by temperature as it affects rate of pupal development and ecdysis. For instance, MacLean (1975b) documented emergence at Prudhoe Bay in two seasons (1971-1972) differing by about one week in the time of snowmelt and the onset of activity, and found that emergence differed by a like amount in the two seasons. Because of the length of the life cycle, differences between areas or years in the density of emerging adults may reflect differences in the lar- val population density or in the relative abundance of the cohorts com- posing the population. The large difference in emergence of adult P. hannai on the Carex-Oncophorus meadow in 1970 and 1971 (Table 11-4) can be attributed to differences in cohort size rather than total popula- tion; the population was actually much larger in June 1970 than in June 1971, as can be easily surmised, since the 1970 population also included the large cohort giving rise to adults in 1971. The emergence of adult P. /jcrrtwa/ increased from 1970to 1971 on the Carex-Oncophorus meadow, declined on the Dupontia meadow, and re- mained essentially stable in the polygon trough (Table 11-4); thus, differ- ences in emergence between years do not simply reflect weather patterns. In some years, however, cold weather before or, especially, during the emergence period can delay or even inhibit emergence, as was recorded in the very cold July of 1969 (MacLean and Pitelka 1971, MacLean 1973). -5: I a J3 "a •9- *«. ^ I s: On SQ < -a 1^ ca E -o " E S-5 OJ u c o 00 a. -J 5 § i — -^ — > O 00 O o a -4: Cl o 427 428 S. F. MacLean, Jr. Taken over a number of years and cohorts, mean emergence of adults must reflect population size. P. hannai are clearly most abundant in wet meadows and polygon troughs (Table 11-4). Tipula carinifrons adults emerged at greatest density from mesic meadows. The overall den- sity of emerging T. carinifrons (3.0 individuals m"^) was less than that of P. hannai (11.5 individuals m"^); however, because of the difference in size of adults, the biomass of emerging T. carinifrons (35.3 mg m"^) was greater than that of P. hannai (19.6 mg m'^). Enchytraeidae The life cycles of the three dominant enchytraeid species last from one to two years, and may include one or two distinct periods of recruit- ment each season (Figure 11-4). In both Cernosvitoviella atrata and Mesenchytraeus sp. early season recruits come from eggs that were de- posited in cocoons the prior season, while the late season recruits hatch from eggs deposited in the same season. Thus, Enchytraeidae can over- winter successfully as eggs, immature worms, or mature worms. Growth rates of these species are reflected in the ratio of production to average biomass, P/B. Mesenchytraeus sp. grows to the largest size, over 300 \xg dry weight over the two-year life cycle, and the P/B (annual production divided by average biomass, both in mg m"^) ratio is quite high, 3.22. C atrata is much smaller with a maximum size of about 20 pig, but growth is accomplished in one year, and P/B is 2.89. The two- year growth period and modest size (maximum = 65 ^g) of Henlea per- pusilla gives rise to a P/B of 1 .49. Although the average biomass of C atrata (131 mg m"^) is half that of H. perpusilla (260 mg m"^), the estimated annual production of the two species is nearly equal. This em- phasizes the danger of basing estimates of ecological importance upon density and biomass data alone. The relative growth rate of the three a> 0 5 0 D^ Cernosvitoviella atrata j_ _L nz Henlea perpusilla X ^ Ih a: EirE Mesenchytraeus sp A 1 — I 1 1 Summer Winter , 2 Summer Winter 3 Summer Winter 4 Summer FIGURE 11-4. A schematic view of the life cycles of the three dominant species of Enchytraeidae. .am ,4^' The Detritus-Based Trophic System 429 enchytraeid species was well above those of the two cranefly species studied, and this is reflected in much higher P/B ratios in the En- chytraeidae (Table 11-6). Thus, the difference in energetic activity of En- chytraeidae and Diptera is even greater than the difference in biomass. Evolution of Life Cycles Life cycles lasting more than one year occur in many arctic inverte- brates (Chernov 1978, MacLean 1975a). Given the short growing season and low temperatures of the Arctic, few species may be able to complete growth and development in a single season. Those species unable to over- winter and renew growth in the following season will be eliminated from the arctic fauna. An invertebrate species might be able to complete devel- opment in many, even most seasons; however, an obligate annual life cycle demands successful development and reproduction every season for maintenance of the population. Thus, a sequence of severe summers could eliminate annual species from the fauna. Many herbivorous insect species have obligate annual life cycles that are closely tied to the phen- ology of the plants upon which they feed. This may contribute to the shortage of foliage-dwelling insect herbivores in the coastal tundra at Barrow. In the Arctic, life cycle length is determined by both temperature and length of the active season, that is, by growth rate and duration of the growth period. Thus, were the season lengthened with no change in mean temperature as, for instance, occurs in the subantarctic islands (Rosswall and Heal 1975), Tipula carinifrons and Pedicia hannai might achieve annual life cycles. Relative growth rate of both Tipulidae and Enchytraeidae declines in larger individuals. This observation is not unique to arctic inverte- brates. Consider a boreal and an arctic species characterized by the growth rate functions g, and gi, respectively, with ^2 < g\ due to lower temperatures in the arctic regions (Figure 11 -5a). Such a growth rate function results in the pattern of growth shown in Figure 1 l-5b. Let W„ be the weight at maturity. The slower growth rate (^2) of the arctic popu- lation requires a prolonging of the development period to reach W„. This increases the period of exposure to mortality, and may lead to a smaller population, M (Figure 11 -5c), at maturity. Thus, many species that might be able to grow and complete the life cycle in the Arctic are unable to maintain a population due to the total mortality during the long devel- opment period. This may be one factor contributing to the reduced diver- sity of northern ecosystems. Reduction of development time and de- creased generation mortality would help to explain the steep increase in species diversity found along climatic gradients away from the immedi- 430 S. F. MacLean, Jr. Growth (g.g-l. Rate day"') ""^^ 9^^ ^--- Weight a. Growth rate functions relating growth rate to body weight for a subarctic (gj and an arctic (gi) invertebrate. log of Weight Time b. Growth curves resulting from growth rate functions g, and gz. Different development periods A, and U) are required to reach the size fWJ necessary for pupation. log of N Number of Survivors Time c. Survivorship curve showing the number of adults (Nt and Nj^ produced after develop- ment periods t, and tz. Slower growth (gz) leads to a longer development time (U) and fewer individuals (N2) surviving to become reproducing adults. FIGURE 11-5. Effect of growth rate on sur- vivorship. ate arctic coastal tundra (MacLean 1975b, MacLean and Hodkinson 1980). In Diptera, larval development continues until some point {W„) is reached. At this time, there is a complete change (pupation) from a grow- ing larva to a reproducing adult. The onset of sexual maturity in Enchy- I The Detritus-Based Trophic System 431 So W "m ' '■ Log Wg of Weight Time, years FIGURE 11-6. Factors influencing the evolution of multi- annual life cycles of Diptera. S = summer, W = winter. traeidae is a much less radical change. The form of the animal remains much as before, and growth continues. In fact, most of the biomass is added after the onset of sexual maturity. In Diptera, successful completion of the life cycle is influenced by constraints upon adult biology. The highly synchronous emergence of adult Diptera suggests that these constraints are rigorously imposed. On- ly flies emerging within a narrow time span successfully complete the life cycle. The short period each summer during which successful emergence may occur is indicated by the areas 5,, Si, Sy (Figure 11-6). In a particular climate larvae growing according to the growth function g, reach W^ and complete the life cycle in two years. In a somewhat more severe climate growth is slowed to that described by growth function ^2. In this case W„ is reached at 6 which falls between S2 and Si, however emergence at time ti is disadvantageous. Some individuals in the population might retain the two-year life cycle and pupate in 52, but at a smaller size {W2) than in the less severe climate. Since fecundity in insects is related to body size, this carries with it a cost of reduced fecundity. Other individuals in the population may extend the life cycle and pupate in S3 at a size equal to or even larger than IV,„. In this case full fecundity is maintained, but at the cost of increased period of exposure to mortality. The strategy that maxi- mizes expected reproduction (probability of survival x fecundity) should prevail in the population. Body size and fecundity can change continuously with climatic se- verity and growth rate. Life cycle length, however, changes discontinu- ously, a year at a time. Thus it is likely that small changes in climatic se- verity will result in changes in body size and fecundity (Figure 11-7). This would explain the well-known decline in insect body size along elevational 432 S. F. MacLean, Jr. Life Cycle Length Annual 2 Years 3 Years Weight of Adults Climatic Severity Growth Rate FIGURE 11-7. The resulting changes in body size and life cycle length of Diptera along a continuous gradient of increasing cli- matic severity and decreasing growth rate. gradients (Mani 1962, 1968, Houston 1971). Similar changes, apparent- ly, occur over latitudinal gradients (Hemmingsen and Jensen 1957). Body size and fecundity should continue to fall until the disadvan- tage of reduced fecundity balances the disadvantage of increased mortal- ity accompanying a lengthening of the life cycle. At this point the life cy- cle will change discontinuously. The unit lengthening of the life cycle should be associated with a discontinuous increase in adult size (Figure 1 1-7). Thus, we expect a saw-tooth pattern of size and life-cycle length in species that are widely distributed along gradients of environmental se- verity. Data are not available to test this prediction. ENERGETICS OF SOIL INVERTEBRATES Estimates of production, respiration, assimilation, and consump- tion of energy were made for each of the major invertebrate groups (Table 11-5). The source and reliability of the estimates varied with amount of information available. Information was most complete for the cranefly species T. carinifrons and P. hannai; production was esti- mated from changes in the size distribution of larval populations in the field, and respiration was estimated from laboratory measurements of respiration as a function of temperature and size of larvae, extrapolated to field temperature and size distribution (MacLean 1973, Clement 1975). A similar technique was used to estimate production of Enchy- traeidae. Respiration of Nematoda, Enchytraeidae, Acari, and Collem- bola was estimated using equations or parameters derived from the literature. Calculation of energy budgets for these groups was then com- pleted using bioenergetic ratios (e.g. production/respiration, assimila- tion/consumption; Table 11-6) derived from published values (summar- The Detritus-Based Trophic System 433 TABLE 11-5 Estimates of Energetic Function (J m'^ yr'J of Major Invertebrate Groups in the Coastal Tundra Ecosystem Production Respiration Assimilation Consumption Nematoda 970 1,940 2,910 7,265 Enchytraeidae 120,525 97,100 217,625 544,050 Acari Prostigmata 810 1,010 1,820 2,420 Mesostigmata 865 1,085 1,950 2,600 Oribatei 1,250 1,780 3,030 7,560 Collembola 13,770 13,770 27,540 68,850 Diptera P. hannai 3,810 4,310 8,120 10,150 T. carinifrons 5,020 3,850 8,870 22,180 P. gracilistyla 2,510 1,925 4,435 11,090 Other Nematocera 4,270 4,270 8,540 21,340 Muscidae 335 335 670 1,675 ized in Heal and MacLean 1975). Thus, estimates of consumption, as- similation, production, and respiration are not independent. The results (Table 11-5) emphasize the importance of Enchytraeidae in this coastal tundra ecosystem. The energetic role of enchytraeids far exceeds the sum of all other invertebrates, and is exceeded by lemmings only in the year of a population high (Chapter 10). Even then, the enchy- traeid value is achieved in an active period of about 100 days, while lem- mings are active year-round. Enchytraeidae are similarly important in tundra-like moorland of the British Isles, where they accounted for 46 and 55% of the assimilation of energy by animals in two habitats on peat soils (Coulson and Whittaker 1978). On high arctic tundra on Devon Island, N.W.T., Canada, nematodes were the major invertebrate con- sumers on a dry cushion plant-Hchen community, while Enchytraeidae were the dominant consumers in a sedge-moss meadow community (Ryan 1977). The estimated annual production of the two abundant cranefly spe- cies, Tipula carinifrons and Pedicia hannai, was compared with the an- nual emergence of aduh flies estimated earlier (Table 11-4), assuming an energy value of 22.6 J mg"' of tissue produced. The emergence values (35.3 and 19.6 mg m"^ yr"' respectively) represent 15.9 and 11.6% of the estimated annual production of larvae. Since larvae lose approximately 434 S. F. MacLean, Jr. TABLE 11-6 Bioenergetic Parameters of Various Taxa of Soil Invertebrates R (Mf O2 mg-' hr- ') P/R' P/B^ A/C Nematoda 1.6 0.50 1.34 0.4 Enchytraeidae 1.2 1.24 2.5 0.4 Acari Prostigmata 1.4 0.8 1.71 0.75 Mesostigmata 1.3 0.8 2.26 0.75 Oribatei 0.7 0.7 0.99 0.4 Collembola 1.0 1.0 1.43 0.4 Diptera P. hannai 1.1 0.88 0.70 0.8 T. carinifrons 0.7 1.30 1.61 0.4 P. gracilistyla 0.5 1.30 1.07 0.4 Other Nematocera 0.8 1.0 1.36 0.4 Muscidae 0.7 1.0 1.23 0.4 P — production; R — respiration; B — biomass; A — assimilation; C — consumption. ' joule joule' or cal cal"'. ^ mg mg"'; production may be changed from milligrams to joules assuming 22 J mg'. 60% of their mass in pupation, the emergence actually represents ap- proximately 88 and 49 mg of larval production. The remainder repre- sents mortality of larvae prior to the emergence of adult flies: 134 mg m■^ or 60% of annual production, for Tipu/a carinifrons, and 120 mg m'\ or 71% of production, for Pedicia hannai. In order to relate these calculations to ecosystem function, the ener- getic estimates must be partitioned according to the trophic role of the animals. Only the Nematode data were collected and reported according to trophic-functional categories. The literature abounds with observa- tions of gut contents or feeding preferences of particular invertebrate species; however, generalizations are few and tenuous. Literature values and our field data were used to partition the activity of taxonomic cate- gories according to the scheme shown in Table 11-7. In view of the great importance of Enchytraeidae, it would be par- ticularly valuable to know their feeding habitats. Unfortunately, direct observations are lacking. Enchytraeid guts commonly contain plant de- tritus in various stages of decomposition, microorganisms, and, where ■:m The Detritus-Based Trophic System 435 TABLE 11-7 Division of Energetic Activity According to Trophic Function (% of activity) Assigned to the Various Invertebrate Groups. Microbi vore Bacteria Herbivore Saprovore and algae Fungi Carnivore Nematodes Free-living 0 20 60 20 0 Predaceous 0 0 0 0 100 Plant-parasitic 100 0 0 0 0 Enchytraeidae 0 20 20 60 0 Acarina Prostigmata 0 0 20 10 70 Mesostigmata 0 0 0 20 80 Oribatei 0 50 10 40 0 Collembola 5 50 15 25 5 Diptera P. hannai 0 0 0 0 100 T. carinifrons 25 50 10 15 0 P. gracilistyla 25 50 10 15 0 Other Nematocera 0 70 10 20 0 Muscidae 0 70 10 20 0 Coleoptera Carabidae 0 0 0 0 100 Staphylinidae 0 0 0 0 100 Chrysomelidae 100 0 0 0 0 available, mineral matter. O'Connor (1967) found fungi in greater pro- portion in the gut than in the available substrate for two of three enchy- traeid species examined, while the third species showed no selectivity. Dash and Cragg (1972) found that boreal woodland Enchytraeidae were attracted to fungal baits placed on the soil surface. Nielsen (1962) exam- ined the digestive enzymes of four species of Enchytraeidae, including one species of Mesenchytraeus, and concluded that they are unable to break down the complex structural polysaccharides of higher plants. Thus, it appears that Enchytraeidae are primarily microbivores, selec- tively ingesting fungi and also feeding upon bacteria and algae ingested with dead organic matter. Most investigators emphasize the importance of fungal hyphae in the diet of Collembola (Peterson 1971); however, evidence from gut con- tents (Bodvarsson 1970), feeding preference and growth rate (Addison 1977), and digestive enzymes (Zinkler 1969) of Collembola of the genus Folsomia indicate that they feed directly upon dead organic matter. Since Folsomia quadrioculata and F. diplophthalma, together, constitute 70% 436 S. F, MacLean, Jr. of the Collembola found in the coastal tundra at Barrow, the feeding ac- tivity of this group is biased toward saprophagy. The litter and soil-dwelling prostigmatid and mesostigmatid mites are mainly predatory (Wallwork 1967); many Prostigmata feed upon the eggs and juvenile stages of Collembola. The Oribatei include species which feed upon microbial tissue, dead plant litter, and combinations of these (Luxton 1972, Behan and Hill 1978). The Diptera include a variety of trophic types. Larvae of Pedicia hannai are predatory and have been observed preying upon Enchytrae- idae in culture (MacLean 1973). Larvae of aquatic Tipulidae are com- monly indiscriminate detritus feeders (e.g. Hall and Pritchard 1975). In British moorland blanket bog larvae of Tipula subnodicornis feed upon Hverworts, and thus are herbivores (Coulson and Whittaker 1978). Smir- nov (1958, 1961) examined gut contents of invertebrates in a Sphagnum bog and found large quantities of Sphagnum leaves only in Tipula lar- vae. In the coastal tundra at Barrow, Prionocera gracilistyla is restricted to mossy depressions and Tipula carinifrons is commonly found in dry moss hummocks. Although many of the invertebrates found living in moss do not actually consume living moss (Smirnov 1958, 1961), more than the estimated 25% of the energy consumed by these craneflies may come from living plants. The majority of the remaining Diptera larvae are probably sapro- phagous (Raw 1967, Healey and Russell-Smith 1970), although microbial tissue is undoubtedly digested as it is consumed along with plant Utter and humus. ENERGY STRUCTURE OF THE DETRITUS-BASED TROPHIC SYSTEM Estimates of trophic function can be applied to biomass and ener- getic estimates to infer the trophic structure of the invertebrate fauna of this coastal tundra ecosystem. The result (Table 11-8, Figure 11-8) is strongly determined by the division of trophic function assigned to En- chytraeidae, which makes the largest contribution to three of the five tro- phic categories: saprovores, bacterial and algal feeders, and fungivores. Canopy-dwelling herbivores are virtually absent from the fauna, the only exception being sawfly (Tenthredinidae) larvae, which feed upon the prostrate willows that occur in drier habitats. Even with a fraction of the biomass and activity of the large and abundant cranefly larvae as- signed to herbivory, invertebrate herbivores are of minor importance in this ecosystem. Consumption of 15.1 kJ m"^ yr"' (Figure 11-8) amounts to less than 1 gdw m"^ yr"', a negUgible part of annual net primary production. The Detritus-Based Trophic System 437 TABLE 11-8 Biomass of Invertebrates (mg m'^Jin the Coastal Tundra at Barrow Partitioned According to Trophic Function Microbivore Bacteria Total Herbivore Saprovore and algae Fungi Carnivore Nematodes Free-living 16 0 3 10 3 0 Predaceous 1 0 0 0 0 1 Plant-parasitic 15 15 0 0 0 0 Enchytraeidae 2119 0 424 424 1271 0 Acarina Prostigmata 25 0 0 5 3 18 Mesostigmata 21 0 0 0 4 17 Oribatei 72 0 36 7 20 0 CoUembola 336 18 183 55 92 18 Diptera P. hannai 164 0 0 0 0 164 T. carinifrons 206 52 103 21 31 0 P. graciiistyla 104 26 52 10 16 0 Other Nematocera 139 0 97 14 28 0 Muscidae 12 0 8 1 2 0 Coleoptera Carabidae and 14 0 0 0 0 14 Staphylinidae Chrysomelidae 2 2 0 0 0 0 Total 3277 113 909 547 1481 227 The microbial feeders (bacteria and algae grazers, and fungivores) make up the majority of the biomass of the soil fauna. This contrasts with many forest ecosystems where large populations of saprovores such as earthworms, millipedes and isopods may dominate the litter and soil fauna. Consumption by saprovores amounts to approximately 9 gdw m'\ about 4.5% of the total annual input to the detritus-based trophic system. In contrast, fungal-feeding invertebrates consume over 18 gdw m"% an amount in excess of the mean annual standing crop of fungi. Animals are influenced by both availability and quality of their food. The food of microbivores is less abundant but of higher quality (higher in soluble carbohydrates and nutrients and more easily digested) than that of herbivores and saprovores. It may be that animals using a low-quality diet are at an extra disadvantage under the severe conditions of the Arctic. Schramm (1972) investigated the interaction of tempera- ture and food quality as determinants of growth rate of herbivorous in- sects. Diets low in protein produced poor growth at all temperatures. Diets lacking in low molecular weight carbohydrates (starch and sugar) 438 S. F. MacLean, Jr. Harvested by next trophic level TUnharvested Production J. Respiration Bio- mass (gm )1 I Recycled r through ►J D.O.M. Feces Consumption (kJm-^ yr-') Vertebrate Carnivores B = 0.003 T 0.03 6.7 145.2 23.4 Tnv. Carn. J. 5.9 6.3 17.6 -5.9 — 6.3 — 123.0 28.0 Bacteria & Algae Grazers 0.55 64.9 38.5 76.6 TT28.1 i78.7 Fungivores ^ -^ 1.48 ""^-^ 2.9 2.9 tLLi9.2 15.1 Plants 33.5 Saprovores 0.91 3222.5 108.4 12.1 J 357.8 214.3 76.6 1309.9 Microorganisms ca. 18.0 824.4 108.4 180.4 4532 3640 1364.3 Dead Organic Matter (22,200-45,000 to 15 cm) 288.8 — rf (Accumulation) FIGURE 11-8. The bioenergetic structure of the detritus-based trophic system in the coastal tundra at Barrow. The Detritus-Based Trophic System 439 produced poor or no growth only at low temperature (10 °C). Plant de- tritus, the food source of saprovores, contains little protein and consists largely of carbohydrate in the form of cellulose and other long-chain polysaccharides. The quality of the diet may also provide an explanation for the rela- tively great abundance of prostigmatid compared with oribatid mites. Luxton's (1972) review of data for Oribatei indicates that approximately 25% are wholly (macrophytophages) and 50<7o are partially (panphyto- phages) dependent upon low-quality dead organic matter for food. In contrast, Prostigmata feed on microorganisms and on other animals and their eggs, a higher quality diet. Larvae of the cranefly Pedicia hannai are the dominant soil carni- vores in the coastal tundra at Barrow, followed by the predatory mites. Predatory beetles, Carabidae and Staphylinidae, were abundant only on the drier areas. Spiders were not accurately sampled and are not included in this analysis. The spider fauna is poorly developed, consisting almost entirely of small web-spinners of the family Linyphiidae that probably contribute minimally to energy flow. The total consumption by inverte- brate carnivores of about 16.7 kJ m"^ yr'' amounts to about 12% of the productivity and 27''7o of the average biomass of their prey, figures that indicate a modest level of predation. The total consumption by invertebrates is about 700 kJ m"^ yr'. Consumption by carnivores represents energy consumed at least twice by animals. Subtracting this, the equivalent of about 35 g m"^ of input to the ecosystem passes through invertebrate animals each year, 34 g of this in the detritus-based trophic system. This is approximately 19% of the an- nual input, assuming an input of 190 g to the detritus-based system. The fractions of consumption that appear as feces and as produc- tion remain within the detritus-based system. Feces and unharvested pro- duction are recycled through the dead organic matter pool (Heal and MacLean 1975). Through respiration, animals in the detritus-based tro- phic system are directly responsible for the loss of 128.5 kJ m~^ yr"', which is the equivalent of less than 7 g of input. Assuming that annual accumulation of organic matter is insignificant, 3.5% of the total annual input of 190 g m'^ is dissipated by respiration of invertebrates, and the remaining 96.5% by microbial respiration. Allowing for accumulation of up to 15 g organic matter m"' yr"' changes these figures only slightly, to 3.8% for animal respiration and 96.2% for microbial respiration. Inver- tebrate respiration varied between 65.3 kJ m"^ yr"' in the low polygon rim and 242.3 kJ m"^ yr"' in the adjacent trough, with the difference due pri- marily to Enchytraeidae; however, because of the relationship of inverte- brate biomass to primary production (Figure 1 l-3a), it is unlikely that the proportionate contribution of invertebrates to total community respira- tion changed considerably between habitats. 440 S. F. MacLean, Jr. TABLE 11-9 Invertebrate Respiration and Primary Production at a High Arctic Tundra and North Temperate Moorland Site Invertebrate Primary respiration production IR/PP (kJ m-^ yr') (kJ m-^ yr') (%) Devon Island, N.W.T., Canada Hummocky sedge-moss 26.8 3,549 0.8 meadow Cushion plant-lichen 27.6 448 6.1 ridge Moor House, U.K. Peat soils Blanket bog 352 12,351 2.8 Juncus squarrosus 829 14,962 5.5 Mineral soils Alluvial grassland 2263 9,790 23.1 Limestone grassland 1963 9,280 21.2 Source: Devon Island, Whitfield (1977); Moor House, Coulson and Whittaker (1978). Comparable data are available for high arctic tundra on Devon Island, Canada, and for the tundra-like peat moorlands of the British Isles (Table 11-9). Values for invertebrate respiration in two habitats at Devon Island are about one-fifth the value reported here for the coastal tundra. Primary production in the sedge-moss meadow at Devon Island is about equivalent to that of the Carex-Oncophorus meadow of the Bar- row area; thus, the estimated role of invertebrates in this habitat is much smaller than we find in meadows of the coastal tundra of northern Alaska. In the dry cushion plant-lichen community on Devon Island pri- mary production is lower and invertebrates apparently play a greater di- rect role in community energetics. Expressed as a proportion of primary production, invertebrate res- piration in peat soils of the British moorland site is about the same as in the arctic coastal tundra near Barrow. The higher invertebrate activity in mineral soils results from the large populations of earthworms (Lumbri- cidae), which are scarce in peat soils. The difference between peat and mineral soils is attributed to the low nutrient status of the vegetation and resulting litter of peat soils (Coulson and Whittaker 1978), a situation that may be shared with the coastal tundra ecosystem. In Chapter 9 a theoretical estimate of annual production by micro- The Detritus-Based Trophic System 441 organisms, based upon decomposition of the entire net primary produc- tion, was given as 75 g m"^ yr''. Using this estimate, the consumption by microbivores of 25 g m'^ yr"' accounts for 33% of the annual production of microorganisms. If we hypothesize an annual accumulation of 10 g m'^ yr'' and reduce the estimate of microbial production accordingly, estimated consumption by microbivores increases only slightly, to 35%. Alternatively, observed rates of decay indicated a maximum value of 90 g m'^ yr"' for microbial production, which sets the level of consumption by microbivores at 28%. Animal biomass and activity are strongly confined to the near- surface layers whereas microorganisms are more evenly distributed, at least through the organic layer. Overall, 71 % of the invertebrate biomass occurs in the top 2.5 cm. Assuming 25% of the microbial biomass occurs there (Figure 8-4), the effect of invertebrate feeding activity is magnified nearly three-fold in the top 2.5 cm compared with estimates integrating over all depths. Thus, animals might consume an amount approximately equal to the annual microbial production in the top 2.5 cm. It is clear that, at least in the surface layers of the tundra, animal activity may exert a considerable influence upon microbial function, and hence upon the decomposition process. Below 5 cm depth the impact of animals is prob- ably minimal. THE ROLE OF SOIL INVERTEBRATES IN NUTRIENT CYCLES Invertebrate production represents nutrients temporarily withdrawn from cycling and unavailable to plants, and biomass represents the amount of nutrients immobilized at any time. Since invertebrate biomass contains a number of important nutrients such as phosphorus and nitro- gen in concentration much greater than either living plant tissue or detri- tus, these nutrients may be immobilized in amounts larger than dry mat- ter or energy content might suggest. Can this amount be significant in ecosystem dynamics? The density and surface concentration of soil invertebrates have pre- viously been combined to express number of animals per unit volume of the soil. The maximum concentration, about 125 f^g of biomass cm"', oc- curred in the top 2.5 cm of the polygon trough. The organic matter den- sity (bulk density x percent organic matter) at the same place is 68 mg cm"'. This value includes live plant parts, microorganisms, invertebrates, and dead organic matter. Thus, living invertebrates at their greatest aver- age concentration represent less than 0.2% of the organic matter of the system. Even allowing for the approximately five-fold concentration of nitrogen and a three-fold concentration of phosphorus in animals rela- 442 S. F. MacLean, Jr. tive to plant tissue (Coulson and Whittaker 1978), invertebrate biomass contains less than 1% of the nutrients of the system. The high nutrient concentration of invertebrate biomass may be sig- nificant as a source of nutrients for microorganisms following death of the animal. Such a nutrient source might be sufficient to stimulate mi- crobial decomposition of surrounding energy-rich but nutrient-poor or- ganic matter. That is, local concentrations in an otherwise nutrient-poor environment may lead to a higher overall rate of decomposition. For a brief period each summer the coastal tundra at Barrow is aswarm with adult insects, mainly Diptera, that have emerged from the soil to complete the life cycle. The emergence of the craneflies P. hannai and T. carinifrons, alone, represents 55 mg m"^ (Table 11-4); the total emergence of Diptera is, perhaps, twice this amount. Thus, each summer about 100 mg m ^ including 1 mg of phosphorus and 10 mg of nitrogen, leaves the soil and becomes mobile over the tundra surface. This provides a considerable potential for nutrient transport. It is likely that insect death and oviposition in most microtopo- graphic units approximates emergence, resulting in no net movement. However, where insects develop in areas that are relatively uncommon and disperse in search of other similar areas, there is Hkely a net move- ment away from the preferred unit. This movement is increased when predators intercept dispersing individuals. Thus, it seems certain that there is a net movement of nutrients from pond sediments to surrounding tundra caused by the emergence of adult chironomids (midges), which are heavily preyed upon by terrestrial birds. This process would tend to reverse the movement of dissolved nutrients and detritus from the land surface into ponds with spring snowmelt each year. The net movement in any year is undoubtedly small, but accumulated over many years this me- chanism may contribute to current patterns of nutrient distribution. The action of soil saprovores in reducing the average particle size of litter, thereby increasing surface area available for attack by microorgan- isms, is frequently cited as an important factor in the decomposition pro- cess (van der Drift 1959, Crossley 1977, but see also Webb 1977). Given the estimated rate of consumption of detritus by saprovores, and even adding an additional amount for consumption of detritus by microbi- vores, it appears that less than 10<^o of the annual input of detritus is con- sumed by invertebrates each year. In contrast, where large earthworms dominate the fauna, as on the mineral soils at Moor House, consumption may approach the annual input of detritus (Satchell 1971). In the coastal tundra at Barrow, the most important interaction of invertebrates in the decomposition process is their consumption of mi- croorganisms and consequent effect upon the composition, biomass and activity of the microbial community. Direct evidence is scanty, but recent research suggests that the activity of soil invertebrates can significantly i The Detritus-Based Trophic System 443 influence microbial decomposition, and that the magnitude of this effect is poorly represented by measures of direct energetic involvement (Chew 1974, Crossley 1977, Kitchell et al. 1979). Coleman et al. (1977) estab- lished soil microcosms containing bacteria as decomposers with and without amoebae and nematodes as grazers of the bacteria. Release of CO2 and mineralization of N and P occurred more rapidly in the micro- cosms that included the grazers. Parkinson et al. (1977) showed that selective grazing by Collembola influenced the growth and colonizing ability of competing fungal species, and that this effect was as marked at low experimental densities of Col- lembola as it was at higher densities. Addison and Parkinson (1978) found that addition of Collembola stimulated the release of carbon diox- ide from tundra cores that had been sterilized, then inoculated with fresh litter and microorganisms. Addition of the saprovore species Folsomia regularis had a greater stimulatory effect than addition of the fungivore Hypogastrura tullbergi or of a mixture of the two Collembola species. In this regard, it may be significant that Folsomia species dominate the Col- lembola fauna of the coastal tundra ecosystem at Barrow. Standen (1978) used litter bags to study the effect of soil fauna on decomposition of litter from a British peat moorland. Bags containing litter with either enchytraeid worms or tipulid larvae lost weight faster and showed a higher rate of oxygen uptake than did bags containing only the litter without animals. At Barrow, Douce (1976) compared the rate of weight loss of litter on control and chemically defaunated plots in polygon rim, basin, and trough habitats. Weight loss from litter was reduced on the defaunated plots. Similar results from a variety of other ecosystems were reviewed by Chew (1974) and Crossley (1977). These data indicate that invertebrate activity stimulates the decom- position of organic matter. This is consistent with the correlation be- tween invertebrate biomass and organic matter turnover rate. Because of the limited vertical distribution of invertebrates, any effect of their activ- ity occurs only near the surface. The rate of microbial activity is also highest at the surface and drops off with depth because of the combined effects of lower temperature, poorer aeration, and reduced substrate quality. This produces an important interaction. A high rate of decom- position near the surface (to which invertebrates contribute) limits the proportion of the annual organic increment that reaches the lower depths, where its decomposition rate would be reduced. Anything that inhibits surface activity, such as reduced invertebrate populations, may allow material to reach the deeper layers, and thus will contribute to ac- cumulation of organic matter and reduction of rates of nutrient cycling. The interaction of invertebrates and microorganisms in the near-surface layers may have an importance for overall ecosystem function that is be- yond the proportion suggested by the amount of energy that is actually respired by animals. 444 S. F. MacLean, Jr. ABUNDANCE, PRODUCTIVITY, AND ENERGETICS OF AVIAN INSECTIVORES Insectivorous birds are the top carnivores in the detritus-based tro- phic system. The avifauna of the coastal tundra at Barrow includes a large number of accidental or occasional breeding species; this probably results from the geography of northern Alaska (Pitelka 1974). Barrow Hes at the apex of a triangle of land that concentrates birds that have made an error in navigation. The list of bird species recorded in the Bar- row region includes 151 species, but only 22 of these are regarded as reg- ular breeders (Table 11-10). This includes five species of waterfowl (loons and ducks), nine species of waders (plovers, sandpipers and phala- TABLE 11-10 Species of Birds Breeding Regularly in the Coastql Plain Tundra near Barrow Graviiformes Arctic loon Red-throated loon Anseriformes Pintail Oldsquaw Steller eider Charadriiformes Golden plover Ruddy turnstone Pectoral sandpiper White-rumped sandpiper Baird's sandpiper Dunlin Semipalmated sandpiper Western sandpiper Red phalarope Pomarine jaeger Parasitic jaeger Sabine gull Arctic tern Strigiformes Snowy owl Passeri formes Redpoll Lapland longspur Snow bunting Gavia arctica Gavia stellata Anas acuta Clangula hyemalis Polysticta sielleri Pluvialis dominica Arenaria inlerpres Calidris melanotos Calidris fuscicollis Calidris bairdii Calidris alpina Calidris pusilla Calidris mauri Phalaropus fulicarius Stercorarius pomarinus Stercorarius parasiticus Xema sabini Sterna paradisaea Nyctea scandiaca Carduelis flammea Calcarius lapponicus Plectrophenax nivalis Source: After Pitelka (1974). The Detritus-Based Trophic System 445 ropes), four gulls (including two jaegers and the arctic tern), the snowy owl, and three passerine species. Excluding the aquatic-feeding species, the terrestrial bird fauna consists of a group of carnivores (pomarine and parasitic jaegers, and snowy owl) that are conspicuous only in years when lemmings are abundant, and alarge group of "insectivorous" birds. The paucity of passerine species and abundance of waders contrasts sharply with temperate avifaunas. Of the three passerine species, the snow bunting is limited by lack of natural nesting cavities to areas around present or past human settlements, and the redpoll, a seed-eating finch, occurs sporadically in both time and space. Thus, the Lapland longspur is the only generally distributed passerine bird species of undis- turbed tundra in the Barrow region. The dominance of waders over passerines is limited to the northern Coastal Plain. Many more passerine species breed in the Foothills and the Brooks Range. This argues for the importance of ecological rather than biogeographic factors in limiting the diversity and composition of the breeding bird fauna of the arctic coastal tundra. Studies on the birds of the Barrow area conducted for many years by F. A. Pitelka and his associates provide a firm foundation of information on natural history and breeding ecology. The present discussion will focus upon six species: the dunlin, pectoral sandpiper, Baird's sandpiper, semipalmated sandpiper, red phalarope and Lapland longspur. These six species are the major consumers of arthropods in the coastal tundra eco- system. They form a guild of avian consumers linked to the detritus- based trophic system. Phenology The birds arrive on the tundra in early June, as the tundra is just be- ginning to emerge from the winter snow cover, and daily mean tempera- tures are still well below freezing. Establishment and defense of territor- ies occurs as the snow melts; and courtship and nesting follow shortly. The median date of clutch completion falls on or before 15 June in both dunlin and Lapland longspurs (Table 11-11). On this date the tundra is normally about SO'Vo snow covered (Chapter 2), and much of the exposed habitat is unavailable to feeding birds because of ponded meltwater. Breeding of other species follows shortly, with median dates of clutch completion falling within a two-week period. Except for the phalaropes, which are semi-aquatic, the early onset of egg-laying concentrates early season activities in areas of upland tundra. Nesting synchrony within species is high. Custer and Pitelka (1977) found that the median date of egg-laying fell between 7 and 14 June, and followed the first by an average of only seven days in Lapland longspurs o .<3 CO s: OQ t: I ^ (J cq PQ < C K - It ac T3 3 00 I - 00 c 'ob ■a 00 c _3 U c '^ 1 ?! = il -a CO T3 C c — ra - '^ •= (/5 C — 1> t» i/; T3 00 CO > 7: s D. CO d d •a u jr 'J 0 0 0 # j: 0 »/"l KTi >/-i 0 CnI 5 3 Tt 0^ ^_^ r- 0^ (Nl r*-, — »o vO — ■* II vO II s •0 II 5 ^ II =^ II 00 II c 3 _ o :x o O d S 9 o d O \C o o o r- r^ o o O vO O o o V^ 0 0 Vi 00 »/-i r«-l v^ rs| rsj bo bo bo bo bo bo 3 a ■a a CG Q. -a c ts (/) O Q. T3 C CIS CO E a o « D. ■o c (L> E (U w 3 CO OJ 03 h ■a '■*_ 0 0 ir d) L. 0 « c . c/5 n OJ 0 j= (U c ■ ' tTi X5 r IJ lU i/^ o ^ Of) E c •— CO !•: OS tu 446 The Detritus-Based Trophic System 447 studied over seven seasons. Seastedt and MacLean (1979) found that no new male longspurs were successful in establishing a territory and obtain- ing a mate after 12 June, only eight days after the first arrival on the study area. Incubation lasts from 12 days in Lapland longspurs to 22 days in dunlin (Table 11-11); thus, the peak of hatching occurs in late June in longspurs and in the first half of July in the wader species. The altricial longspur young remain in the nest for about eight days, during which they are fed and brooded by both adults. Young waders are very preco- cial. They usually abandon the nest within hours of hatching of the last egg, and they gather all of their own food; however, they are metabolic- ally unable to maintain their own body temperature (Norton 1974) and require frequent brooding by adults (50 to 83% of the time during the first week for semipalmated sandpipers; Ashkenazie and Safriel 1979a). The young of both longspurs and waders appear on the tundra before the warmest weather of the season. A large proportion of the birds have left the tundra by the second week in August, when average temperatures are only slightly below the mid-July peak. Thus, avian activities are strongly skewed toward the early season and are not coincident with the warmest weather. Neither temperature nor length of the snow-free season can be considered major factors in the evolution of breeding phenology. It is more likely that the timing of breeding is determined by the emergence of adult Diptera. Prior to the mass emergence of adult flies, adult birds feed heavily on dipteran larvae, which they capture by inserting the bill into the tun- dra. Young waders have growing, incompletely ossified bills. They are unable to probe into the tundra, and thus require surface-active prey for the first three to four weeks of life. The appearance of wader young closely follows the appearance of their prey. Longspur hatching precedes the emergence of adult Diptera, and larvae and pupae are the major prey fed to nestlings. The young leave the nest in early July, just as adult Dip- tera become abundant. Similarly, the abrupt decline in emergence of adult Diptera, which occurs sometime after mid-July, may limit the peri- od in which newly hatched birds can forage successfully, and thus may be responsible for the synchrony of reproduction. The phenological relationship of avian reproduction and insect emergence may help to explain an apparent paradox in incubation peri- ods (Norton 1974). In dunlin both adults incubate, providing almost con- tinuous attention to the eggs; incubation lasts about 22 days. In pectoral sandpipers, the eggs are slightly larger than those of dunlin; the female incubates alone and is present about 86% of the time, yet the incubation period is 19 to 20 days. Nest initiation is later in pectoral sandpipers than in dunlin, presumably because of the later snowmelt of the lowland mea- dows in which they feed. The shorter incubation period of pectoral sand- 448 S. F. MacLean, Jr. pipers may have evolved to allow them to exploit the abundance of prey in lowland meadows while maintaining hatching at the optimal time of the season. Use of Habitats and Food Habits The bird species differ in their distribution over major habitat units (Figure 11-9) and their use of microhabitats within these (MacLean 1969, Custer 1974); however, there is considerable overlap in both habitat use and prey selection throughout the breeding season (Holmes and Pitelka 1968). Early in the season the birds make greatest use of upland habitats, with only pectoral sandpipers and red phalaropes making significant use of wet meadow and pond habitats. Longspurs feed mainly on seeds dur- ing the first ten days of June, and thus differ from the waders, but switch to larval Diptera once breeding commences (Custer and Pitelka 1978). All birds seem to feed preferentially along the edge of retreating snow- fields, suggesting that insects just exposed are more easily located or cap- tured than those that have achieved full activity in thawed tundra. Territorial defense diminishes during incubation, and defense of the breeding territory ceases altogether when the young leave the nest in early July. The adults commonly lead the young from upland nesting habitats, which provide relatively little cover, to low-lying meadows, where vege- tation provides greater cover from predators. It is unlikely that abun- dance of food is a serious factor influencing habitat choice at this time. Since emergence of adult insects is sensitive to weather, periods of food Boird's sandpiper Loplond longspur dunlin semipalmated sandpiper Well-drained Ter- ♦ ^ races and High-centered pectoral sandpiper^ Stream Banks Polygons Beoch Ridges red pholarope Mesic Meadows Low-centered y^g^ Polygons Meodows Ponds 3,5,6 1.2 /. Carex-Oncophorus meadow. 2. Wet Dupontia meadow. 3. Polygon trough. 4. Moist Dupontia meadow. 5. Basin of low-centered polygon. 6. Rim of low-centered polygon. 7. Carex-Poa mesic meadow. 8. High-centered polygon. FIGURE 11-9. The distribution of preferred breeding hab- itat of birds along a mesotopographic gradient. n The Detritus-Based Trophic System 449 scarcity may occur even during mid-July; however, such periods prob- ably influence all habitats alike. Striking year-to-year differences in growth rates and survival of wader young appear to be closely related to weather conditions during this mid-summer period (Norton 1973, Myers and Pitelka 1979). Interspecific overlap in diet is greatest during this peri- od of the season (Holmes and Pitelka 1968). In August, after the period of adult insect abundance, dunlin and longspurs move back to upland tundra. Longspurs take large numbers of sawfly (Tenthredinidae) larvae and seeds, both items that are little used by the wader species. Dunlin feed on Tipula larvae and, if drying and ex- posure of pond margin sediments permit, on midge (Chironomidae) lar- vae. Holmes (1966) noted a segregation between adult and immature dunlin, the immatures making greater use of coastal, brackish lagoons. In semipalmated and Baird's sandpipers the two adults share in in cubation; however, females (occasionally males in Calidris bairdii) depart at or soon after the time of hatching, leaving only one adult to ac- company the young. Adults, and then immatures, move to coastal la- goons as soon as the young become independent, and their southward migration begins soon thereafter. In the polygamous pectoral sandpiper, females incubate alone; males form flocks in lowland marshes in late June and early July, and leave the tundra before the eggs hatch. Thus, male pectoral sandpipers that have migrated to arctic Alaska from southern South America remain on the tundra for a period of less than 30 days. The females remain with the young in July, but leave soon after the young become independent around the first of August. Flocks of immatures remain on the tundra throughout August, feeding mainly on pond-margin chironomid larvae. In some years large numbers of immature pectoral sandpipers, clear- ly representing far more than local production of young, appear on the tundra in the Barrow area in August. Similarly, in some years large flocks of immature long-billed dowitchers (Limnodromus scolopaceus) may be found in August, although dowitchers breed only occasionally and sparsely in the immediate Barrow area. There is thus a premigratory coastward movement of shorebirds from inland breeding areas that con- tributes to the use of resources of the coastal tundra in August. In the sex-reversed polyandrous red phalarope, females form flocks and depart soon after all nests are completed, and males incubate the eggs alone (Schamel and Tracy 1977). Coastal lagoons and even the coast of the open ocean are used as premigratory staging areas by red phala- ropes, and large flocks may remain throughout August and even into September. Although a variety of tundra arthropods appear in the diets of these birds, the majority of breeding activities are supported by larval and adult Diptera, especially of the three cranefly species, Tipula carinifrons. 450 S. F. MacLean, Jr. Prionocera gracilistyla, and Pedicia hannai (Holmes and Pitelka 1968, Custer and Pitelka 1978). T. carinifrons, alone, makes up between 40 and 72% of the diet of adult longspurs between 10 June and 20 July (Custer and Pitelka 1978), and over 50*^0 of the food fed to nestlings (Seastedt and MacLean 1979). Overall, about 40% of the longspur diet consists of T. carinifrons. Dunlin are almost entirely dependent upon T. carinifrons (Holmes 1966), and both Baird's and pectoral sandpipers feed largely upon larvae of this cranefly in June. Pectoral sandpipers feed more in lowland meadows, and consequently take more larvae of Pedicia hannai and Prionocera gracilistyla than do the other sandpiper species. Semipalmated sandpipers differ from the other sandpipers in specializing upon the smaller larvae of Chironomidae throughout the season. Overlap in diet between species is greatest in early June, when feeding sites are limited, and in mid- to late-July, when food is maxi- mally available (Holmes and Pitelka 1968, Custer and Pitelka 1978). Density and Reproductive Success Both density and breeding success of these birds differ between years and between areas of the coastal tundra, with at least part of the variation due to differences in food supply. Average nesting density of the five wader species varied from 0.09 nest ha"' in the pectoral sandpiper to 0.18 nest ha"' in Baird's sandpiper (Table 11-11); however, these aver- age values obscure both yearly and spatial variation. Baird's and semi- palmated sandpipers occurred abundantly in census plots adjacent to coastal lagoons; dunlin and pectoral sandpipers were scarce or absent from these areas. Pectoral sandpipers are particularly variable in nesting density from year to year. Holmes (1966) found from 0 to 27 nests in a 40-ha study plot censused each season from 1960 to 1963. Average den- sity varied between 0.02 and 0.20 nest ha"' in the four years (1969-1972) included in this analysis. In contrast, nesting density of dunlin is more stable, varying from 0.07 to 0.14 nest ha'' between 1968 and 1972 (Nor- ton 1974). Pitelka et al. (1974) argued that some of the sandpipers, not- ably the pectoral sandpiper, have an "opportunistic" social system that allows maximum reproduction in favorable years and places; others, such as the dunlin, use a more "conservative" social system that pro- vides for a modest level of reproduction in all years. The waders are determinate layers. The vast majority of nests con- tain four eggs, with some tendency for clutches laid very late in the sea- son (usually replacement clutches) to be smaller. Hatching success varied considerably between species, from 49% in Baird's sandpiper to 75% in dunlin (Table 11-11). Most egg loss was caused by predation from jaegers and least weasels. Baird's sandpipers, which suffer the greatest losses. The Detritus-Based Trophic System 451 nest in the most exposed sites. Given this strong selection pressure, the continued use of exposed nesting sites by Baird's sandpipers at Barrow seems paradoxical. The solution to the paradox may lie in the lack of ten- acity in this species, indicated by the paucity of return sightings of the many breeding birds and juveniles banded in the Barrow area. Soon after hatching, adults may lead the chicks a distance of up to 2 to 3 km from the nest site (Ashkenazie and Safriel 1979a); thus, broods are difficult to follow, and data on survival of young are rarely collected. Safriel (1975), however, was able to follow the fate of 39 broods of semi- palmated sandpipers, and found a mean of 1.74 young fledged per brood, for a success rate of 44<^o. As with eggs, most losses of juveniles were attributed to predation. The average density of a population of longspurs studied over a seven-year period (1967-1973) was 0.47 nest ha"' (Custer and Pitelka 1977), a value well above the wader species. Density was highest (0.82 and 0.88 nest ha') in 1967 and 1968, but dropped steadily to a low of 0.12 nest ha"' in 1972. The population showed some recovery in 1973 and 1975 (Seastedt and MacLean 1979). In longspurs, year-to-year variation in productivity of nesting habitat cannot be detected at the time of terri- tory establishment because most of the ground is still covered by snow; hence, territory size and nesting density are related to average or "ex- pected" productivity of the habitat (Seastedt and MacLean 1979). Nest- ing success should be more responsive than nesting density to year-to- year variation in habitat productivity. The decline in nesting density of longspurs recorded between 1968 and 1972, then, should reflect changes in the size of the potential breeding population, as influenced, in part, by breeding success in preceding years (Figure 11-10). Since the study area used by Custer and Pitelka was placed in optimal longspur habitat, a slightly more conservative estimate of 0.30 nest ha"' is used for the area as a whole (Table 11-11). Lapland longspurs are indeterminate layers, and clutches of four, five, and six eggs are common. The modal clutch size, five eggs, occurred in 45% of the nests examined. Mean clutch size was 5.06 eggs, and varied between only 4.76 and 5.50 between years (Custer and Pitelka 1977). Over this period as a whole, 64''7o of the eggs hatched and 68^/0 of the chicks survived to fledging, about eight days after hatching. Thus, the average female produced about 2.2 fledged young. The occurrence of starved nestlings in longspur nests after some of the young have fledged indicates that food supply can influence repro- ductive success; however, only 3.1% of all nestlings observed over the seven-year period died of starvation. Loss to starvation was greatest (7.1%) in the very cold summer of 1969. By far the major source of reproductive failure was predation; 22.3% of all eggs were taken by predators prior to hatching, and 22.7% 452 S. F. MacLean, Jr. 2 3 Young Fledged per Nest in Previous Year FIGURE 11-10. Relationship of nesting density and breeding success during previous year. (Data from Custer and Pitelka 1977.) of chicks were lost before fledging. Predation accounted for 73% and 72% of losses of eggs and nestlings, respectively, and for most of the var- iance in reproductive success (Figure 11-10). The decline in nesting den- sity from 1968 to 1972 was attributed to a sequence of years with heavy predation. As with the waders, the major predators upon eggs and nest- lings are pomarine and parasitic jaegers and least weasels. Since the den- sities of these predators are primarily determined by density of lemmings (Chapter 10), the reproductive success of insectivorous birds may be de- termined primarily by events in the herbivore-based trophic system. Energetics and Impact Upon Prey Populations Norton (1974, West and Norton 1975) used gas exchange techniques to study the bioenergetics of sandpipers breeding at Barrow. An alter- native time-energy budget approach was used by Custer (1974) on Lapland longspurs and by Ashkenazie and Safriel (1979b) on semipal- mated sandpipers. Energy budgets calculated for semipalmated sand- pipers using these aUernative approaches yielded virtually identical estimates. Norton found that the temperatures ordinarily encountered by the birds during the summer season are below thermoneutrality, so that The Detritus-Based Trophic System 453 energy must be expended for thermoregulation throughout the season, and particularly in early June. Various energy-demanding processes and activities — territorial defense and display, egg formation, incubation, care and brooding of chicks, molt (in dunlin and longspurs), and pre- migratory fat deposition — are distributed through the season. This led Norton (1974) to suggest that the daily energy requirement for mainten- ance plus productive activities remains near the maximum metabolic rate throughout the period of residence on the tundra. Differences among species in the energy requirement for reproduc- tion stem primarily from differences in body size and in duration of resi- dence on the tundra (Table 11-12). Energy required per nesting attempt (Figure 11-11) is greatest in pectoral sandpipers and dunlin, and least in semipalmated sandpipers; however, when differences in nesting density are included to estimate total energy removed from the tundra (Table Chironomid P hannai T. carinifrons 10' IQl c E E V o c o a: a. o o ^ 10° Q. w m or 10 I MUM _L I I 10^ 10' Prey Size, mg d.w. 10' FIGURE 11-11. The requisite rate of prey capture (items per minute of foraging) to satisfy daily energy require- ments for birds with different energy requirements and different proportions of the day available for foraging. o p -s: to ?3 s s: fs CQ < OO ^-, o f- o -^ 2? i3 „ OJ ^ V", r~- vC vO s c ca (N \o ^i O 4> ^. ■* c> OO o OO H- c vC o Tf m t/5 f^ m oo rvj o t/5 9£ K C r=: c 2 00 'ob c 1> ■o 00 3 1/5 •y5 O O n a o oo — - a> i, -^ c 2 '-' o f^ aj o =* ^ ^ ^ '-" Tt r«-j o '-J ^~> ^ -^ '5 '_» "*^ o - - 03 a- i 00 i> ■o < §i = g ■a T3 >> c ,-K alt «J "" — 1* F >. 3 ii -o u •4* „ 3 '—i OJ rr J£ 1> u a !3^ O o 00 OO «. vO vO ■^ rn 00 NO o 5 _ OO — r- OO r- >o § Ov v~i — M O 00 r-~ OO oo r- 00 3 vO 00 rr-t r*-) o r- o o rsi — rsl rsl to •* o rsl O ^ vO v, v^ 1/-1 1--. f> r*"j <~4 rsj — <^J sC o r- -^ oc ON «o r- r- OO W-1 «/^ so OO O A 73 r^ C r-- CO O j£ *-' ■* 1^ o vO o -a Z c ^ 5 'J i _o ra vO 5 3 454 The Detritus-Based Trophic System 455 11-12), the average value for longspurs is almost twice that of the next species, the dunlin, and accounts for 35% of the energy removed from the tundra by the complex of avian insectivores. The total food requirement for these birds (4.35 kJ m'^ yr~') amounts to only 3% of the productivity of invertebrates in the detritus-based food chain, which suggests a modest intensity of predation; however, a large proportion of invertebrate productivity is accounted for by taxa, particu- larly Enchytraeidae, that are not eaten by the birds. It appears that about 40% of the total consumption by avian insectivores is supported by the cranefly species Tipula carinifrons, and about 60% by the three cranefly species. Thus, about 35% of the annual production of T. carinifrons, and 23% of total cranefly production, is taken by avian predators. The annual emergence of adult craneflies amounts to about 35 mg m'^ in T. carinifrons and 20 mg m"^ in P. hannai, and is largely confined to a three-week period. MacLean and Pitelka (1971) recorded the median 67% of the total captures within a period of 5.3 to 11. 6 days for T. car- inifrons, and within 3.8 to 6.3 days for P. hannai. Thus, during this peak period emergence of each species is on the order of 3 mg m"^ day"'. During this period 66% of the diet of adult dunlin and 79% of the diet of juvenile dunlin consists of adult tipulids (Holmes 1966). In order to satisfy the gross energy requirement of two adults plus three juvenile dunlin (3.95 eggs clutch "' x 75% hatching success) at these proportions, 52 g of adult craneflies would be required each day, the equivalent of the total emergence of adult Tipula carinifrons and Pedicia hannai from 9600 m^ At 0.11 dunlin nest ha'', even at the peak of emergence dunlin alone consume about 11% of the daily emergence of craneflies. Each longspur family (two adults plus an average of 3.25 chicks hatched per nest) requires about 55 grams of prey each day. In early July, adult craneflies compose about 70% of the diet (Custer and Pitelka 1978). This removes another 11.6 g ha'' day' or 19% of the peak emergence. The addition of other avian insectivore species may raise the daily intake above 40% of the peak emergence of adult craneflies. It is easy to appre- ciate the impact that these predators must have, particularly upon adult insects that emerge into less than peak populations, when loss to preda- tion must approach 100%. Earlier, I estimated the mortality of cranefly larvae prior to pupa- tion at 134 mg m"^ for T. carinifrons and 120 mg m'^ for P. hannai. Over the season, 76% of the diet of adult dunlin consists of larvae of T. carini- frons. This results in a consumption of 202 g of T. carinifrons larvae ha'', or 15% of estimated larval mortality. Addition of other avian spe- cies might double this value. Predation by birds is heavily concentrated on fourth instar larvae; mortality at earlier stages of development must be due to other causes. Varying amounts of time are available to the birds for foraging, de- 456 S. F. MacLean, Jr. pending upon the time demands of other activities, notably incubation. Female pectoral sandpipers incubate their eggs 85% of the time (Norton 1972), leaving no more than 15% of the day for foraging. They must ob- tain food at a rate of about 100 mg dry wt min"' during foraging bouts. This might be satisfied by taking about 100 chironomid larvae, 20 P. hannai larvae, or 4 larvae of T. carinifrons per minute (Figure 11-10). The advantage of feeding on the large cranefly larvae is clear. Female dunlin have a similar energy demand, but by sharing incubation with the males they have much more time available for foraging, and their requi- site rate of prey capture is about one-fourth that of the female pectoral sandpiper during incubation. Shortly after arrival female longspurs forage about 80% of the day (Custer 1974), and must find food at a rate of about 10 mg dry wt min"'. During this period longspurs feed on seeds, Collembola, and small chiro- nomid larvae, items too small to be used by other species with higher energy requirements, or by longspurs later in the season when time avail- able for foraging is reduced. Thus, as is so often the case in studying animal populations, the im- pact of avian predators stated in relation to the total energy budget of the ecosystem appears small (Figure 11-8); however, because of the concen- tration of predation upon Diptera and particularly upon the craneflies, avian predation may have a large influence as a force in the evolution of life cycles and in the reproduction and population dynamics of Diptera populations. SUMMARY The coastal tundra ecosystem supports abundant populations of En- chytraeidae, Collembola, and Diptera, modest populations of Acari, and small populations of Nematoda. Soil invertebrates are concentrated near the surface of the tundra, where the number of individuals per cubic cen- timeter can be quite high. Large differences in abundance are found among the various microtopographic units that compose the coastal tun- dra. These differences are related to soil moisture, aeration, and annual input of detritus. Both invertebrate abundance and biomass and plant production are inversely correlated with accumulated soil organic matter. Tundra soil invertebrates have long life cycles, often extending over several seasons. Craneflies require four years to complete larval develop- ment and their ratio of annual production to average biomass is conse- quently small. The energetics of the detritus-based trophic system in the coastal ecosystem is dominated by Enchytraeidae. The fauna is lacking in large, I The Detritus-Based Trophic System 457 abundant saprovores, and microbivory is the most important trophic function of soil invertebrates. Invertebrate respiration accounts for about 130 kJ m"^ yr'', which is 3.5% of the annual input of detritus. Consumption of microorganisms is greater than average biomass and ac- counts for 33 to 37% of estimated microbial production. Near the tundra surface the entire annual production of microorganisms may be con- sumed by invertebrate microbivores. Evidence suggests that grazing upon detritus and microorganisms by soil invertebrates stimulates the de- composition of organic matter and accelerates the turnover of energy and cycling of mineral nutrients in the ecosystem. Soil invertebrate populations, especially Diptera, support an abun- dant and diverse community of breeding birds; four sandpiper species, the red phalarope, and the Lapland longspur are the most important of these. Bird breeding is timed so that the young can feed on the adult Dip- tera that emerge in early and mid-July. In June and August dipteran lar- vae, especially those of craneflies, are the most important prey. Energy requirements are determined by body size and duration of residence on the tundra. When breeding density is also considered, longspurs are the most important consumers of tundra arthropods. Birds may consume 35% of the annual production of the cranefly Tipula carinifrons, and 50% of the peak emergence of adult craneflies. This level of predation must influence the evolution of life cycles and contemporary population dynamics of tundra Diptera. 12 Carbon and Nutrient Budgets and Their Control in Coastal Tundra F. S. Chapin III, P. C. Miller, W. D. Billings, and P. I. Coyne INTRODUCTION Arctic tundra ecosystems are characterized by low productivity, slow energy flow, slow nutrient cycling, and, in many cases, peat accum- ulation, despite a diversity of parent material and variable species com- position (Billings and Mooney 1968, Bliss et al. 1973, Rosswall and Heal 1975, Dowding et al. 1981). These common features derive in some fashion from the low annual solar irradiance and consequent low tem- peratures that prevail at high latitudes. Earlier chapters have shown that most organisms inhabiting the wet meadow tundra have adaptations that minimize the effect of low temperature upon vital processes. How then do low irradiance and associated low temperature help to generate the unique characteristics of coastal tundra, and to what extent are other fac- tors responsible? In this chapter we summarize, from information pre- sented in earlier chapters, pool sizes and average annual fluxes of carbon and selected inorganic nutrients. Through comparisons with other eco- systems we attempt to identify those aspects of energy flow and nutrient cycling that are peculiar to the Arctic and consider the nature of causal links with climate. STANDING CROPS Two sets of standing crop data are presented for Barrow. Pool sizes and fluxes at the intensive study site in a Carex-Oncophorus meadow were measured in considerable detail in 1970 and 1971(Figures 12-1, 12-2 and 12-3). Biomass was also measured in 1972 in all major microtopo- graphic units and then weighted by relative abundance to give average 458 Carbon and Nutrient Budgets 459 pool sizes (Chapter 3, Table 12-2). The Carex-Oncophorus meadow values are based upon more thorough study and may thus be more accu- rate, whereas the data for the mosaic of wet meadow vegetation types are more representative of the coastal tundra of northern Alaska. Unless otherwise specified (e.g. Table 12-2) data refer to the Carex-Oncophorus meadow. Although relatively little carbon is fixed by tundra vegetation in any one year, up to 20 kg m"^ of carbon has accumulated in the top 20 cm of tundra (Table 12-1). This is similar to the carbon content of other wet tundras (17 to 32 kg m'^), tropical rain forests (17 to 34 kg m"^) and a red alder shrub stand (20 kg m"-). It is greater than the carbon accumulation in grassland (3 kg m"^), chaparral (6 kg m"^), and a Douglas fir forest (12 kg m"^) (Table 12-1). Many tundra communities, including those near Barrow, have buried organic horizons preserved in the permafrost. Tem- perate communities with their deeper soil profiles often contain consider- able carbon which has been leached from upper horizons and thus also have more total carbon than is shown in Table 12-1 . Because tundra con- stitutes a significant fraction (5%) of the total terrestrial landscape (Whittaker 1975) and because wet and moist tundra (including coastal and tussock tundra) make up a substantial proportion of all tundra com- munities, a major alteration of the carbon balance of tundra could sig- nificantly modify the global carbon balance. Tundra differs from most other ecosystems in that the bulk of its carbon is contained in soil, rather than in live biomass (Table 12-1; Schlesinger 1977). At Barrow over 96% of the organic carbon is bound in dead organic matter or peat, and only 1 .7% or less is in living organisms. The remainder is dead plant parts. In contrast, 50 to 75% of the organic carbon in forests and 10% of the carbon in grasslands is in living organ- isms (Table 12-1). This implicates decomposer organisms as a major bottleneck for carbon and energy flow at Barrow and other wet tundra sites. Like the tundra, mid-latitude grasslands contain a substantial pro- portion of carbon in dead organic matter. But in grasslands, roots pene- trate 1 to 2 m so that carbon from dead roots and associated microorgan- isms is distributed throughout the soil (Weaver 1958, Clark 1977). In contrast, tundra soils typically exhibit a distinct surface horizon 10 to 20 cm thick, in which the percentage of organic matter is 90 to 96%. Such high concentrations of organic matter are associated with low pH and consequently with reduced nutrient availability, as discussed in Chapters 7 and 8. Most of the carbon in living organisms is in plants, and most of this is below ground in roots and rhizomes (Figure 12-1, Table 12-2). The average aboveground vascular standing crop for the coastal tundra at Barrow is 24 g C m'^ (Table 12-2), half that of the intensive study site (Figure 12-1). The greater vascular aboveground standing crop in the « (^ to <^ )< a, 5 o s: C5 0 *• 13 ■§ 0 -*»-* ?3 S s^ 0 0 u ■s: g- 0 ;!:: f^ Q, -$ ■^ t*J ■§ •\ 53 s: til 00 S. 0 0 ^ i^ ^ •s s: s: (5j 0 ^ 53 « 0 ^ ^ •^ 5 s: 53 s: JO 53 i<» to ^^ =s ?3 to ■^ 0 ,0 s h^ t*J < u c Urn o a. u c * >> (^ '^ E c 03 Ofl 3 ^^ CT E _ OJ C3 wn 0 >. H UD u T3 3 O UJ ON 2 13 2 ^ ^^ :e[:: T3 !^ ns — X) q U c > T3 C 03 ON ^ 03 > J= S oa c/5 H T3 > I/) Q OS s -f; O (-0 u^ — ; — Tt o 00 (Ov rn O '^ — O -^ Tt O rsi /-i Ov r^ ro r<^ d 00 NO r^ — Tt O 00 1/-1 Tt -"^ NO r^ rvj NO — ON r«-i /-i NO ■^ ON 00 NO NO r- TT NO ON 00 03 T3 -O c c ^ -^ T3 ^ T3 - 0! C C O 03 1^ 00 -o -a OS .^ u ON c o oo 00 ON ON 03 T3 C 03 ON 00 C o Q 00 c o Q ON .t; 03 ,_ c/5 u- o o ON .Si o 00 r^ NO r^ 00 >/-, ON d 1 1 00 i O NO !/5 On no on o — — • 00 "o 0 t^ r^ r^ I — NO NO ^ C/D O NO CO (/5 o o ...♦. .p* ^ 1^ ^ "T ^ .— ■ T3 ■o C c r^ — 3 cd _« J= 0 '0 «-» ^ t/5 (/5 0 C/5 S 0 03 C 3 0 00 0 C/5 03 0! "c5 3 0 'E "O T3 T3 13 ■o X) -0 00 iJ ;o « ca E 1) E C 3 0 03 E a 5 c 0 0 00 0 C '^ !> u. 00 0 i> iS <1> 0 V5 (/5 a> t/5 OS 5 03 C/5 CS X "c? ^ ^ 0 a. ^ 00 ii 3 ^ cS 3 3 CO 1 460 06 c 0. c * >. '■l ™ E c CO OX) 3 ^-^ CT E QJ C3 on o >. H ;/: u c75 u CQ < o 'J tu r- x: c >> S C -^ On -a re c O -3 c > 5 - 03 O CQ c c « CO > > ^..^ r- ^.^ V3 r- ON r~ i« J,^-^ r- r^ 15 r~- av ^— ^ Ov Ov ^ lo - -a o ON CO On >, CO OJ CO r- — It c u ;/: C IJ S p o i( C o SJ c ^ o 1 o U 1 1 O 1 O U o 25.2 Joh 16.7 Odi q — NO oo ON — TT ON nO \0 o ON ON ON c^ U X o ro O vO r^ v-i r-) NO Tf ■* •<1' m ' rA so m '^ ^ c CO •o o o c o S o O CO a 3 O 3 •a 3 o 3 T3 o CNl t-~- NO NO o NO NO CnI q — r^ nO o (N Tt NO oo NO 00 On rsf no' C o I' ™ •P E o iT o -o C/5 "2 ^ 2 >- 0-; CO CO -P CO •> 3 S ^ IS _ 0" U O UJ NO o _ O I/) O (- 5 00 CO o CNl T3 O a: CO D. CO ■o T3 o u« O L. ^ u. ^« 1) -o o 00 c 'cO lU w^ OJ 2, ^ t/5 5 D. CO D. c/) 1/5 OO >> lo a o E ^ E ^ 3 O rn CO H H Q o 1- CO ao .S o 461 462 F. S. Chapin III et al. Atmosphere Rr = 40 Pw = 202 Rw = 38 C= 40 PNw= 164 RC = 0.05 Carnivore (0.001), . I I I I Herbivore (0.09)f- vV" I ' ' PRw = 31 I Rh= 1.2 Live vascular aboveground >f — 0 Dead vascular aboveground 251 PRo = 44 Live vascular belowground I I I I I I P = 12 M Rfui - 2 PNm = 10 Rn = 21 18 Live moss PRm=16 8 347 Dead vascular belowground Invertebrate Eli I ganisms -L. [p Microorganisms Dead moss 1 X 30 Litter Ui 1 1 i 1 E-7 A= ' Peat (18992) • COo transfers . ^—Organic C transfers FIGURE 12-1. Carbon budget of the wet meadow vegetation type at the intensive study site to a depth of 20 cm. The number in each box is the pool size of carbon for that compartment expressed in g C m'^ to a depth of 20 cm. We assume that carbon constitutes 44% of organic material, the average measured at another Alaskan tundra site (Chapin et al. 1979). The area of each box is propor- tional to its compartment size. Values next to arrows indicate the annual carbon fluxes between respective compartments (g C m~^ yr'J. Pv and Pm - net daytime photosynthesis of vascular plants and moss, respectively; PNv and PN^ = net an- nual carbon exchange between atmosphere and aboveground vascular plants and mosses, respectively; Rb, Rd, Rh, Rc - respiration of vascular belowground, de- composers, herbivores and carnivores, respectively; PRg, PRv, PRm - net annual production of vascular belowground, vascular aboveground and mosses, respec- tively; C - net carbon flux from atmosphere to community; A = net annual ac- cumulation; E = export in runoff. Animal compartment sizes assume peak lem- ming abundance to indicate maximal animal role. All other data were collected in 1970 and 1971, years of low lemming abundance. All values shown were obtained independently by direct field measurements, extrapolated to an annual basis, and corrected for light and temperature. Values for vascular plant and litter carbon are calculated from Tieszen (1972b) and Dennis (1977), moss carbon from Oechel and Sveinbjornsson (1978), and soil carbon from Flint and Gersper (1974). Refer- ences for transfer values are given in Table 12-4. ',^ Carbon and Nutrient Budgets 463 TABLK 12-2 Standing Crops of Carbon in the Coastal Tundra at Barrow g C m - Chapter Carnivores 0.001 12 Herbivores 0.12 (max) 10 Standing dead 16 3 Litter 40 3 Moss 51 3 Algae 0.15 8 Lichens 0.12 3 Aboveground vascu lar 24 3 Belowground vascu lar 374 3 Nematodes 0.01 11 Enchytraeids 0.95 11 Acarina 0.051 11 Collemboia 0.16 11 Fungi 2.0 8 Bacteria 6.0 8 Total live plant 461 3 Soil organic matter 18992 12 (Fig. 12-1) TOTAL 19518 Note: The moist meadow is a mosaic of vegetation types. Standing crops of each vegetation type are weighted by relative area (Chapter 3). Localized variations in these values are discussed in the text. intensive study site (Figure 12-1). The greater vascular aboveground standing crop in the intensive site in 1971 (Figure 12-1) than in a variety of other Carex-Oncophorus meadows in 1972 (Table 3-2) suggests sub- stantial yearly and/or microtopographic differences in production. Esti- mates of belowground standing crop range from 207 to 574 g C m'\ de- pending upon the microtopographic unit (Chapters 3 and 6). The above- to belowground ratio of live vascular plants is about 1:10 (Dennis and Johnson 1970, Dennis 1977), similar to that found in a shortgrass prairie (Clark 1977). The carbon in moss at Barrow (51 g C m'^) is twice that in vascular aboveground parts. Algae and lichens are less important com- ponents of biomass (Table 12-2). Standing dead and litter constitute a larger standing crop (56 g C m'^ than the live vascular material above ground and retard nutrient cycling by altering the radiation regime with- in the canopy (Chapter 3) and by directly immobilizing nutrients (Chap- ters 8 and 9). Bacteria constitute the largest standing crop of decomposer organ- isms (6.0 g C m"^), three times that of fungi, although the balance be- tween fungal and bacterial biomass varies strikingly with habitat and soil depth (Chapter 8). Soil invertebrates account for 1.8 g C m"^ (Chapter 464 F. S. Chapin III et al. 11). The standing crop of lemmings, the only major herbivore, ranges from 0.00002 to 0.12 g C m "^ depending upon the stage in the lemming cycle. Carnivorous birds attain a maximum standing crop of only 1.4x10* g C m"^ The small size of the herbivore-based relative to the saprovore-based trophic system emphasizes the importance of below- ground interactions in the Barrow region. The amounts of nitrogen and phosphorus in the coastal tundra at Barrow are comparable to those in other ecosystems (Table 12-1). In fact, tundra communities generally have more accumulated nitrogen (960 g N m'^ at Barrow) than do temperate systems, including grasslands (270 g N m-^), alder shrub (700 g N m"^), Douglas fir forest (250 g N m"'), and oak forest (250 g N m"^). The parent materials of tundra soils at Barrow are not unusually phosphorus-deficient, because they are derived largely from marine sediments (Chapter 1). Limitation of primary production by nitrogen and phosphorus (Chapter 5) is thus not a result of small quantities in the system but rather of their slow rate of cycling. It should be noted that in tundra permafrost limits the quantity of nutrients avail- able for exploitation whereas deeper soil horizons often play an impor- tant role in long-term replenishment in temperate and tropical ecosys- tems. As with carbon, the proportion of nutrients in the living compo- nents of the tundra system is quite small, approximately 1% in the case of nitrogen and phosphorus. This is typical of wet tundra (e.g. Babb and Whitfield 1977, Dowding et al. 1981) and differs from forests, where the biota constitute a significant reservoir of nutrients (Table 12-1). Tundra thus appears to represent an end-point on the latitudinal spectrum by having the smallest proportion of the system's nutrient capi- tal tied up in live biomass. Ovington (1968), Marks and Bormann (1972), and Whittaker et al. (1979) have pointed out the importance of vegeta- tion as a nutrient reservoir that retains nutrients within the system by ab- sorption from soil and internal recycling. In the tundra it is primarily the dead soil organic matter that serves this function by structurally binding a large proportion of the nutrients, by providing exchange sites for ca- tions that otherwise would move through the soil during runoff, and by physically preventing thermokarst (thawing and subsidence) and erosion of the underlying mineral soils. The distribution of nitrogen and phosphorus among biomass com- partments follows a pattern similar to that of carbon. However, living soil organisms, microbes plus invertebrates, contain 8% of the nitrogen and 18% of the phosphorus, but only 2.5% of the carbon in living mate- rial. The live belowground vegetation contains almost 80% of the carbon but only 68% of the nitrogen and less than 60% of the phosphorus in liv- ing material. Thus, soil organisms are more concentrated sources of nu- trients than are plants or plant-derived detritus, and form the major avenue of nutrient and energy flow through the saprovore-based trophic Carbon and Nutrient Budgets 465 TABLE 12-3 Distribution of Carbon and Nutrient Pools in Vascular Plants on 4 August, Time of Maximum Aboveground Biomass in the Intensive Study Site part Nutrient (g m-^) Plant C N P K Ca Fe Shoot 44.7 1.79 0.12 0.90 0.17 0.01 Stem base 37.4 1.18 0.10 0.08 0.29 0.41 Rhizome 38.7 0.90 0.10 0.06 0.24 0.82 Root 174.5 4.71 0.40 0.11 1.92 6.12 Total 295.3 8.58 0.71 1.15 2.65 7.36 Percent belowground 85 79 84 22 92 100 system (Chapter 11). The bulk of the nutrient pools of each trophic level occur below ground. Except for potassium, 80% or more of all vascular plant nutrient pools also are localized below ground (Table 12-3). Ani- mal nutrient pools are concentrated below ground and are associated with turnover of organic matter. Even during lemming population highs, 3 to 10 times more nutrients are present in soil invertebrates than in lem- mings (Figure 12-2). AVERAGE ANNUAL FLUXES Carbon Budget During the growing season, the long period of daylight compensates for low sun angle and consequent low light intensities, so that during July the total daily input of photosynthetically active radiation is similar in arctic and in temperate ecosystems (Billings and Mooney 1968). The vegetation is seldom light-saturated and captures more than 1% of the available energy in July (Chapter 4), an efficiency comparable to or higher than that of most other natural communities (Ricklefs 1973). Aboveground plant parts maintain a positive carbon balance 24 hours a day during most of the growing season (Chapter 4), and 81 % of the pho- tosynthate is converted to aboveground biomass or translocated below ground. This aboveground net production efficiency is higher than that of most temperate communities (Ricklefs 1973). Because of the minor amount of dark respiration and the high photosynthetic and net produc- tion efficiencies, relative production rates (g g"' day') for the coastal 466 F. S. Chapin III et al. tundra at Barrow are similar to rates observed at much higher tempera- tures in the shortgrass prairie (Chapter 5). However, at Barrow the grow- ing season is short, and 90% of the aboveground vascular biomass senes- ces each fall. Consequently, the tundra exhibits a small standing crop of photosynthetic tissue. The net quantity of carbon fixed by photosynthe- sis at Barrow (174 g m"' season"'; Figure 12-1) is about half the quantity fixed by a shortgrass prairie. This corresponds to a Barrow growing sea- son which is half as long as that in the prairie (Coleman et al. 1976). It thus appears that the annual carbon input into the coastal tundra ecosys- tem at Barrow is limited not so much by light and temperature effects upon photosynthesis as by the shortness of the growing season, which in turn limits the standing crop of photosynthetic tissue (Miller et al. 1976). Carbon flux through the tundra at Barrow is slow. Less than 1% of the ecosystem carbon pool turns over annually. This contrasts sharply with tropical systems, where 40% of the organic carbon in a rain forest is fixed and respired each year (calculated from Odum 1970). Radiocarbon dating of surface and buried organic matter from soils in the Barrow re- gion yields ages of as much as 10,000 years (Brown 1965). These suggest that there are large pools of soil organic matter with very slow turnover rates and other pools that turn over much more rapidly than the ecosys- tem average, as demonstrated in temperate and tropical soils (Jenkinson and Rayner 1977, Jenkinson and Ayanaba 1977). The partitioning of photosynthetic carbon and its subsequent loss in respiration emphasize the belowground nature of the system. Of the 214 g m"^ of total carbon fixed annually (total net daytime photosynthesis), only 19% is lost as aboveground dark respiration, in part because of the long periods of daylight during summer (Figure 12-1). Another 22% is converted to aboveground biomass, and the remaining 59% is translo- cated below ground. Of the carbon translocated below ground, approxi- mately half is converted to new tissue and half is lost in respiration. This contrasts with the shortgrass prairie (Coleman et al. 1976), where 34% of the total carbon fixed is respired above ground, 9% is converted to aboveground production, and of the 57% translocated below ground, 85% is converted to new biomass. Apparently tundra plants produce shoots efficiently because of the long period of daylight, but much of the belowground carbon is used in maintenance respiration for the large standing crop of roots and rhizomes. The large proportion of litter re- leased below ground, where decomposition rates are low, may be one factor leading to organic accumulation in tundra. However, in grass- lands, where a substantial proportion of the litter is also shed below ground, soil conditions are more favorable for decomposition, and there is less accumulation of soil organic matter. The relative importance of plants and soil organisms as sources of soil CO2 is unclear. The proportion of soil respiration accounted for by Carbon and Nutrient Budgets 467 plant roots and rhizomes has been variously estimated as 6897o (Billings et al. 1977 using gas exchange), 30 to 70*^0 (Bunnell and Scoullar 1975 using computer simulations), and 31% (Chapter 9 from litter bag weight loss). Belowground plant respiration may constitute a proportionately larger CO2 source in tundra than in ecosystems of warmer climates (Coyne and Kelley 1975, 1978, Billings et al. 1977). For example, roots contribute 17 to 30*^0 of soil respiration in a tallgrass prairie (Herman 1977, Redmann and Abouguendia 1978) and 35% in a temperate forest (Edwards and Sollins 1973). Root respiration is estimated to be 10% of total soil respi- ration in a tundra-like heath ecosystem (Chapman 1979). In tundra the period of active decomposition is only slightly longer than the period of primary production, whereas in many temperate ecosystems decomposi- tion continues throughout the year, and primary production exhibits a more restricted season. Thus, the short duration of the arctic summer may affect decomposition even more than primary production when compared with temperate ecosystems. The saprovore-based food web is clearly the predominant pathway of energy flow in the coastal tundra ecosystem at Barrow (Figure 12-1; Chapter 11), as it is at Devon Island (Whitfield 1977) and in grassland systems (Woodmansee et al. 1978). Even during a lemming high at Bar- row, herbivores consume only 20% of the net primary production. In the Barrow region, 60 to 80% of the annual ecosystem respiration occurs when plants are photosynthetically active (Coyne and Kelley 1975). Because photosynthesis occurs 24 hours a day, much of the respir- atory CO2 is fixed immediately by photosynthesis so that the net CO2 flux between atmosphere and canopy is small. The CO2 concentration within the canopy is seldom reduced more than 5% below free atmos- pheric values (Coyne and Kelley 1975). In mid-latitude forest and grass- land ecosystems a larger proportion of the respiratory CO2 is released at night and during the non-photosynthetic season, so that the bulk atmos- phere plays a greater role as net CO2 source and sink than in arctic tundra (Coyne and Kelley 1975). During the winter, particularly in April and May, there is a net CO2 flux from the ground to the atmosphere (Kelley et al. 1968, Coyne and Kelley 1974). Plant forms differ considerably in the role they play in energy flow in the coastal tundra at Barrow. Although mosses constitute 40 to 70% of the maximum standing crop of aboveground plant biomass, they are responsible for only 6% of the carbon fixed by wet meadow vegetation. Clearly mosses do not play a major role in carbon flux in the Barrow sys- tem in the short term. Estimates of the total carbon flux for the Barrow intensive site sug- gest that during the period of study this ecosystem was not in steady state but fixed three times as much carbon annually as it lost in respiration of nonphotosynthetic organs and organisms (Figure 12-1): 468 F. S. Chapin III et al. TABLE 12-4 Barrow Carbon Budget Calculated by Three Methods Method Compartment Aerodynamic* Turfchambert KOH absorption** Total C fixed GCt/' =210 PM+P/V„ + /?r ' =202 PN,+PN» + R,+R» = 1U Ecosystem respiration R, = \10 Rs + Rt' + Ro' = »2 Rs' +Rv+R» + Ro' = \55 Net ecosystem C gain C= 40 C=120 C= 59 Subcompartments of ecosystem respiration /?«= 40 Rb'= 93 Rr'= 29 /?v+/?M = 38 + 2= 40 Ro'= 13 /?„= 22 Note: All values are direct measurements unless indicated as calculated by prime notation. Fluxes given as g C m'^ yr"'. GCU= gross community CO? uptake Rt= total dark respiration above ground Rb = total carbon efflux from eco- system in dark Rd = decomposer plus saprovore respiration PNv and PN^ = net photosynthesis, vas- cular and nonvascular, respectively Rv and /?M = dark respiration, above- ground vascular and moss, respectively ^B = belowground vascular respiration C = net carbon flux from atmosphere *Coyne and Kelley (1975). t PNy (Tieszen 1978b); PN» (Oechel and Sveinbjornsson 1978); /?«, {Rg + Ro), and C (Billings et al. 1978). **PNv and PN^, (as above); R, (Tieszen 1978b); R^ (Oechel and Sveinbjornsson 1978), calculated by extrapolating photosynthesis/light curves to 0 light intensity; Ru (field) (Benoit pers. comm.); Rg ' (Miller 1979) calculated from theoretical respiration costs of growth and maintenance. _^_^_ {PNv + PN^)-iRB + Ro + RH + Rc) = 173-53 = 120 g C m-^ Terms are defined in Figure 12-1. With a carbon budget constructed for only one year, it is unwise to extrapolate and assume that this large imbalance is responsible for the observed organic accumulation at Barrow. Such an imbalance would generate the observed organic accumulation in 160 years, whereas radio- carbon dating indicates that accumulation has occurred for thousands of years (Chapter 1). August was 5°C cooler in 1971 than the long-term average and may have reduced respiration below the norm and partially accounted for the net carbon gain measured in that year. June and July, when most of the growth and photosynthesis occur, were close to the Carbon and Nutrient Budgets 469 long-term temperature average. Furthermore, for the years 1972-1977 litter bag data suggest that, on the average, decomposition exceeded pro- duction at the same site (Chapter 9). Because of the many possibilities for error, we used various independent methods and combinations of meth- ods to calculate the carbon budget of wet meadow tundra. These meth- ods differ in the calculated magnitude of carbon imbalance in 1971, but all agree that there was a net gain of 40 to 120 g C m~^ in that year (Table 12-4). Therefore, although methodology may be responsible for part of the observed discrepancy between total ecosystem carbon fixation and ecosystem respiration, this carbon imbalance was probably real. The car- bon budget in Figure 12-1 is based on actual field measurements, where possible, rather than calculated values. Although the latter show greater self-consistency, they require steady state assumptions. The extent of agreement between field methods indicates that the carbon balance of coastal tundra and other terrestrial ecosystems may deviate substantially from steady state in any given year. Similarly, Woodmansee et al. (1978) suggest that steady state conditions are an unrealistic assumption in eco- system nutrient budgets. Inorganic Nutrients Nutrient Input The presence of permafrost within tens of centimeters of the ground surface limits the amount of thawed soil that is available for weathering. The remainder of the frozen nutrient capital could be tapped only in the event of disturbance or climatic fluctuations that result in deeper thaw. Judging from clay mineralogy, the rate of chemical weathering of miner- als in thawed soil is negligible, due primarily to low temperature (Hill and Tedrow 1961). Hence, the coastal tundra at Barrow apparently de- pends largely upon the atmosphere for nutrient input. This contrasts strongly with most temperate systems where weathering represents the primary source of nutrient supply (Cole et al. 1967, Ovington 1968, Likens and Bormann 1972). The arctic climate and the general global atmospheric circulation patterns severely Hmit atmospheric nutrient input. Arctic tundra regions receive little precipitation. The presence of sea ice minimizes the amount of sea spray that could carry nutrients inland from the Beaufort Sea. Snow cover, wet surfaces and distance from agricultural and urban cen- ters render dry nutrient fallout negligible. The annual atmospheric inputs of nitrogen and phosphorus (Figure 12-2) are an order of magnitude less than those characteristic of temperate systems (Ovington 1968). The resupply of phosphorus and cations to the ecosystem must occur not S I c u c c O " Z Exchang. inorg. 3406 1 C * f- « ■D H l»l T) (D a. o A es I a; -C — (Q o) a. ;; - « ^ e E * c m 3 t o> re — O 0) O) c i ^ - I - ^ S a, O ° S S 05 ir o - ; 1 =h u ^ c a; re t- o> to 5 E ^ i: s: s: so o ^ o . ■«: at to to O <3 or ^ ^^ ^ § ^ § to o s: :^ 53 1^ ao 5 7 1^ 53 r-x .00 oo to Si 53 ^ -Si lij lij -s: o -s: ^ ^ 3 ^ =^ ^^ 53 to 5>0 ^ t^ H ^ to 5- .O Si 53 - to i^ Co "w k: s: <3 53 to £- "^ tr ^ o S ^ 'o 53 5j <^ oo '^ >i: 5 S f: t3 3 to 53 470 8 I U tf ^ y ^ § Si C3 ^ !3 s: o ;j to 53 :^ -5: •. (J ^ i: s: ^ 3 ^ to •^ Q 5 (/2 g > .s: O 0\ ^ 53 s: oo" to ^ .to s: s: 00 >:! ^^ — »^ s: s: 53 s: 53 3 TT ti 53 s: Si 5U 53 0<3 -^ s: hvl 53 o 5j 53 ^ -Si 5 3 53 -.J to 53 QQ o to s: 53 to • ^ s: o ex, ^ t^ OQ 5*. 53 .2 be ^ 471 472 F. S. Chapin III et al. through regular annual fluxes but through the sea spray from rare sum- mer or autumn storms and through geologic and successional processes such as the cyclic occurrence of frost action and draining and coloniza- tion of phosphorus-rich lake basins. In the absence of such renewing in- puts, the rates of decomposition, nutrient cycling, and primary produc- tion decline, and the standing crops of all trophic groups diminish. This situation is most clearly seen in the basins of low-centered polygons on the coastal tundra (Chapter 3). Nitrogen accrues at Barrow primarily through nitrogen fixation, as in most ecosystems (Figure 12-2b). Blue-green algae associated with mosses account for the bulk of nitrogen fixation (Alexander and Schell 1973). Even in the Arctic, nitrogen fixation is strongly temperature- dependent, so that the total annual nitrogen input to coastal tundra is 10-fold smaller than precipitation inputs alone in temperate latitudes (Barsdate and Alexander 1975). In fact, the total annual nitrogen input is only 5% of that which annually cycles through the vegetation in coastal tundra as compared with an estimated 21% in the shortgrass prairie (Woodmansee et al. 1978). The annual input of carbon is l.O^^o of the total amount in the sys- tem. The annual input of nitrogen is only about 0.01 '^o of the total eco- system nitrogen content, and the input of phosphorus is about 0.0014% of the total phosphorus content. Thus, at the current input rates, it would take 10,000 years to regenerate the present standing crop of nitro- gen and over 70,000 years to regenerate the present standing crop of phosphorus. By contrast, at Hubbard Brook (Likens et al. 1977) the an- nual inputs of nitrogen and phosphorus are about 1% and 0.01% of the total standing crop respectively, about 100- and 7-fold higher than in coastal tundra. This comparison emphasizes the importance of nutrient conservation within the tundra system. Nutrient Loss In the Arctic, where chmate dictates that nutrient input must be small, the system can remain in steady state only if it has characteristics that lead to low rates of nutrient loss. Some of these characteristics are associated with climate and landform, others with development of the system during succession. Low precipitation and flat terrain reduce the amount of runoff, so that nutrient loss from the coastal tundra is small. Ninety-five percent of summer precipitation normally evaporates (Brown et al. 1970). Only during the ten days of snowmelt is runoff from the Bar- row tundra appreciable, and at this time the organic mat readily absorbs available nutrients such as ammonium and phosphate (Chapin et al. 1978). Permafrost prevents downward leaching of nutrients. Carbon and Nutrient Budgets 473 Nitrogen is lost from the system primarily in spring runoff, but these losses are less than the annual gain through precipitation and N fixation. The large accumulation of nitrogen in tundra systems (Table 12-1) sug- gests that nitrification and/or denitrification are restricted more severely than nitrogen fixation under tundra conditions. Denitrification rates are very low, even though facultative denitrifying organisms are abundant in the tundra at Barrow (Chapter 7). Low phosphorus availability is one factor limiting denitrification rates (Barsdate and Alexander 1975), and phosphorus thereby plays a role in the accumulation of organic nitrogen. Assuming that organic carbon accumulates in parallel with organic nitro- gen, the continued accumulation of organic matter will further reduce phosphorus availability (Chapter 7) and hence denitrification rate. This positive feedback loop may continue to operate in the absence of large external inputs, as discussed above. Transfer Within the Ecosystem Because of the low nutrient input to the coastal tundra the function- ing of this ecosystem depends greatly upon internal recycling of the exist- ing nutrient capital, more so than do temperate systems. Yet the low tem- perature regime of tundra restricts these rates of nutrient cycling, both directly through temperature effects upon biological processes and wea- thering and indirectly through the occurrence of permafrost, which re- stricts drainage and results in poorly oxygenated soil. Fungi are important in the breakdown of nutrient-containing com- pounds in litter and in better drained soils (Chapter 9). However, in waterlogged, low-oxygen soils, fungal metabolism is depressed and bac- teria dominate (Chapter 8). In contrast to fungi, bacteria at Barrow lack the capacity to break down complex substrates at low temperature (Chapter 9). Thus, any nitrogen and phosphorus contained in complex organic molecules might tend to accumulate in the anaerobic zone be- cause of the absence of fungi and the inability of bacteria to attack such substrates at low temperature and low oxygen. Such nutrients would be largely removed from active cycling within the ecosystem. Low tempera- ture also has direct and indirect effects upon microbial growth (Chapter 9), such that microbial biomass is an order of magnitude less than that characteristic of temperate grasslands (Chapter 8). As soil organic matter accumulates, the soil becomes more acid, restricting the kinds of bacteria and further constraining decomposition. In short, the arctic climate re- stricts decomposition, but many of the temperature effects are indirect and complex as consequences of permafrost, low oxygen and acidity. The cycling of some nutrients is more directly dependent upon de- composition than the cycling of others. Eighty-two percent of the potas- 474 F. S. Chapin III et al. Lemming peak year (0.01} ■-r: 0.21 0.19 Standing dead 10 36) 0 19 N(0 38l 0.04 Vase, shoots jMOIQ) 0.21 Exchangeable calcium 295 Rate during average year ■ Rate during lemming high a. Calcium. Runoff? Exchangeable potassium 13.8 b. Potassium. Carbon and Nutrient Budgets 475 Input^ . Runoff? Soil organic Mg (0,561 Soluble magnesium Exchangeable magnesium 93.9 c. Magnesium. FIGURE 12-3. Annual budgets for calcium, potassium and magnesium in the Barrow wet meadow tundra intensive study sites to a depth of 20 cm. The area of each box is proportional to its compartment size, which is indicated in g m'\ Values next to arrows indicate annual fluxes in g m~^ yr'\ Assumptions for budgets and sources of data are presented in Fig- ures 12-1 and 12-2. Soil nutrient pools are calculated from Gersper (unpubi). slum content of litter is leached during snowmelt, so that decomposition plays only a minor role in the cycling of this element (Figure 12-3). In contrast, almost no calcium is leached from litter, so calcium must recy- cle exclusively through the decomposition process. However, this ele- ment is relatively abundant in the marine sediments from which soils of the Coastal Plain are derived. The soil contains a large exchangeable pool (Figure 12-3), suggesting that calcium does not strongly limit pri- mary productivity. Calcium addition does not stimulate production in cottongrass tundra at Atkasook, 100 km south-southwest of Barrow (McKendrick et al. 1980). Sixty percent of the phosphorus and 80% of the nitrogen contained in litter must be recycled through the decomposi- tion process (Figure 12-2), and both may occur in compounds that are not readily broken down. Both of these nutrients have been found to limit primary productivity in the coastal tundra at Barrow and at other 476 F. S. Chapin III et al. TABLE 12-5 Nutrient Turnover Rate in Various Compart- ments of the Moist Meadow Vegetation Type Net turnover rate (% of pool yr"')* Compartment N P K Ca Mg Dissolved soil inorganic 6,089 21,919 371 96 54 Soil organic 0.21 0.41 t • * ** Live vascular 23 22 36 22 17 Dead vascular 15 21 85 11 18 Live moss 55 55 55 55 55 Lemmings (peak year) 4,760 1,733 28,600 571 4,545 Soil organisms 243 82 ** 2,015 249 ♦Assumes a one-way transfer of material. The actual turnover rate will be much greater in pools where there are multiple paths of nutrient flow through the compartment, as in soil organisms. tPool nonexistent. **Not measured. tundra sites (Chapter 5). Hence, the role of decomposition and nutrient mineraUzation is extremely important in the cycling of these elements and in the functioning of the system in general. Nutrient turnover rates of soil organic matter are 5 orders of magni- tude lower than those of the soluble soil pools (Table 12-5). The turnover times ( = total/input) for carbon, nitrogen and phosphorus averaged for the top 20 cm of soil organic matter are 220 years for carbon, 480 years for nitrogen and 240 years for phosphorus. This compares with an aver- age world carbon turnover time of 40 years (Schlesinger 1977). Such turnover estimates mask complexities associated with different soil or- ganic fractions, each with distinct turnover times (Jenkinson and Rayner 1977). The slow turnover of organic nutrient pools in the soil seems re- sponsible for the slow overall cycling of nutrients in the coastal tundra at Barrow. Slow nutrient cycling characterizes high latitude ecosystems gen- erally (Jordan and Kline 1972, Babb and Whitfield 1977, Dowding et al. 1981). Soluble inorganic phosphorus and nitrogen pools are extremely small and turn over rapidly (Table 12-5), contributing to microorganisms and vascular plants while receiving input from a variety of sources. The soluble nitrogen must be replenished at least 60 times in the course of the growing season to supply the quantity absorbed by plants. Soluble phos- phorus must be replenished 220 times a season, an average of three times a day. The soluble pools for mineral cations turn over less rapidly (Table 12-5). Although the dissolved nutrient pools are small, they are presumably in equilibrium with larger exchangeable pools. The input into the dis- Carbon and Nutrient Budgets 477 solved inorganic nutrient pools from vegetation, microflora, animal wastes, etc., is strongly seasonal. The exchangeable inorganic nutrient pools may perform an important buffering function by diminishing the size of soluble pools when concentrations are high and replenishing them as concentrations decrease. Although the net annual flux into and out of the exchangeable pool is small, the flux at any moment may be large. The labile phosphorus pool, which replenishes phosphorus removed by plants, is smaller in coastal tundra and may provide less seasonal buffer- ing of soil solution concentration than in most soils (Brewster et al. 1975, Barel and Barsdate 1978). The absorption of phosphorus (and presumably of other nutrients) by vascular plants is much more strongly limited by the soluble nutrient concentration in the soil than by temperature (Chapin and Bloom 1976). Because the soluble nutrient pools are small relative to the annual plant requirement, particularly for nitrogen and phosphorus, uptake by the vegetation must depend upon simultaneous nutrient release by decompo- sition or chemical exchange processes. Nutrient release by decomposition and nutrient absorption by the vegetation are thus apparently closely coupled. Nutrient release and nutrient uptake do not occur at constant rates through the season. Microbial populations are characterized by several population increases and crashes each growing season, due to a variety of factors such as changing soil moisture and grazing by invertebrates. One microbial population crash releases enough phosphorus to supply 90% of the annual vascular plant requirement. Such population crashes re- duce the biomass of microbes that might otherwise effectively compete with vascular plants for nutrients. Thus, conditions that cause crashes in microbial populations may be essential for nutrient uptake by vascular plants. Further evidence for this hypothesis is presented elsewhere (Cha- pin et al. 1978). Vascular plants of the coastal tundra are relatively conservative with nutrients and replenish only about 20% of their nutrient capital each year (Table 12-5). This estimate ignores losses from leaching, which may be considerable for elements such as potassium. Much of the plant nutri- ent capital invested in leaves is retranslocated to rhizomes during the lat- ter half of the growing season (Chapter 5). Mosses play an important role in nutrient cycling as well, although this role has not been documented for the coastal tundra at Barrow. Mosses appear to derive their nutrients from plant leachates as well as from the soil and snowmelt water and may effectively filter these nutri- ents before they become available to vascular plants or microorganisms (Tamm 1964). ,The nutrient concentration of brown moss tissue is very similar to that of green tissue (Rastorfer 1978), and moss decomposition rates are low. Therefore, mosses represent an important avenue by which 478 F. S. Chapin III et al. nutrients become bound in peat (Moore and Bellamy 1974). However, slow decomposition of mosses is probably due more to low nutrient con- tent and tundra environmental conditions than to any inherent resistance of mosses to decay (Heal et al. 1978, Coulson and Butterfield 1978). Therefore, the idea that mosses are a major path of nutrients to soil or- ganic matter deserves further critical examination. The importance of animals in the coastal tundra at Barrow is much greater than is indicated by their biomass, owing to their ability to recycle nutrients, as shown by high turnover rates (Table 12-5). Lemmings recy- cle 80 to 90% of the ingested plant phosphorus directly back to the sol- uble inorganic pool, whereas most phosphorus contained in standing dead material must recycle through the soil organic pool at a much lower rate. Thus herbivores can short-circuit the decomposition process, just as cattle do in the shortgrass prairie (Dean et al. 1975). The annual phos- phorus turnover in vascular plants more than doubles in years of peak lemming abundance. The annual absorption of phosphorus by vascular plants that would be necessary to balance this loss to herbivores exceeds the calculated annual uptake rate by 25%. In contrast, most of the calci- um recycled through lemming feces is contained in undigested cell walls and must still cycle largely through the decomposition process before be- coming available again to plants. Soil invertebrates also apparently in- crease the rate of nutrient release from organic matter, perhaps by graz- ing upon microbes (Chapter 11). A given group of organisms may play very different roles in cycling of different elements. The roles of mosses and vascular plants in cycling of nitrogen and of carbon are quite different (Figures 12-1 and 12-2). In the case of carbon, mosses and vascular plants function in parallel, both fixing carbon photosynthetically and then passing it on to herbivore- or saprovore-based trophic systems. In the case of nitrogen, mosses act in series with vascular plants. Mosses probably receive a large proportion of their nitrogen as leachate from vascular plant leaves and are thus an in- termediate step in the movement of nutrients from vascular plants to the soil. Moreover, because blue-green algae associated with mosses are re- sponsible for the bulk of the nitrogen fixation at Barrow, mosses repre- sent a major point of entry of nitrogen into the ecosystem. Mosses un- doubtedly retain some elements more effectively than others, e.g. dival- ent cations more than monovalent cations. LONG-TERM CHANGES IN COASTAL TUNDRA Lakes and ponds, which constitute 30% or more of the surface area of arctic coastal tundra, are continually forming, enlarging and draining in a cyclic process which encompasses thousands of years (Chapter 1). i Carbon and Nutrient Budgets 479 Virtually the entire coastal landscape has at some time been part of this thaw lake cycle. Nutrient cycles and nutrient budgets of terrestrial sys- tems and their yearly variations must be viewed in this cyclic successional context. Furthermore, the extensive interdigitation of aquatic and terres- trial systems has important implications for the function of coastal tun- dra as a landscape unit. Our concept of the exchanges between terrestrial and aquatic nutrient pools and changes in nutrient availability during the thaw lake cycle is shown in Figure 12-4. The annual carbon and nitrogen budgets of wet meadow tundra sug- gest a gradual accumulation of these elements through time, and this view is supported by the accumulation of peat. Some of this accumulation may have occurred in the past. In contrast, tundra ponds exhibit negative carbon balances and virtually no peat accumulation in the sediments (Stanley and Daley 1976). The more effective recycling of organic carbon in ponds than in wet meadow may result from higher temperatures, bet- ter aeration, more intensive aquatic grazing, a more elaborate food web, or some combination of these factors (Barsdate et al. 1974, Hobbie 1980). Considerable organic matter of terrestrial origin decomposes in tundra ponds, including dissolved and particulate organic matter that flows into the ponds during snowmeh and some of the accumulated soil organic matter at the pond margin. Small tundra ponds may enlarge, although others may be invaded by emergent macrophytes. Similarly, large tundra thaw lakes enlarge by erosion and eventually drain. Nitrogen accumulates slowly in both terrestrial and aquatic systems. On land, nitrogen accumulates primarily in peat. When organic nitrogen enters the aquatic system, either through runoff or erosion, much of the carbon portion of the organic matter is respired, but the nitrogen cycles through the aquatic system and is eventually deposited in the sediments as ammonium. In contrast to carbon and nitrogen, phosphorus may exhibit a net loss from wet meadow tundra but accumulates in sediments of small ponds as iron hydroxy-phosphate compounds (Prentki 1976). The phos- phorus in the sediments of the lakes and ponds is insoluble and is not recycled within the aquatic system to any significant extent. Lakes and ponds eventually drain, either through the process of en- largement or when captured by a headward eroding stream (Britton 1957; Chapter 6). When lake sediments are exposed, they are characterized by low organic accumulation and presumably by relatively high phosphorus and nitrogen availability. In spite of low rates of seed production in coastal tundra, complete cover of drained lake basins can be attained within 20 to 25 years by seedling establishment and subsequent vegetative spread (Dennis 1968, Peterson 1978, Webber 1978). Such early succes- sional communities are relatively productive, presumably because of high nutrient availability in the lake sediments. 480 F. S. Chapin III et al. ^ Aquattc ' Lake / Drainage / Succession Phosphorus FIGURE 12-4. Comparison of nutrient pools in terres- trial and aquatic systems on tundra during the course of thaw lake cycles. The relative sizes of the circles on the left indicate differences between terrestrial and aquatic habitats in pool sizes of organic carbon or available N or P. Solid arrows represent annual net transfers by fixation (F), respiration (R), erosion (E) and transport by runoff at thaw (T). Dashed arrows indicate conversion of habitat from aquatic to terres- trial or vice versa. Carbon and Nutrient Budgets 481 The movement of carbon and nutrients from terrestrial to aquatic systems plays a relatively minor role in the annual budgets of these sys- tems. Nonetheless, awareness of the slow alternation of the terrestrial and aquatic landforms across any point in the landscape is essential to an understanding of the overall functioning of coastal tundra, because this alternation prevents long-term nutrient accumulation in soil organic mat- ter. The impact of industrial activities on the whole biosphere and on the tundra itself, either directly or indirectly, is likely to speed up these cycles and processes. From the present work, we have gained considerable knowledge about the coastal tundra ecosystem. Needed now are quanti- tative data on rates of ecosystem change through time. Upon these, we can construct models of what this tundra ecosystem may be like in the fu- ture. The initiation of such long-term ecological research is required so that utilization of this coastal ecosystem can proceed in the most rational and scientific manner possible. SUMMARY AND CONCLUSIONS This study of energy flow and nutrient cycling examined the causal relationships linking low solar irradiance with the unique features of the wet coastal tundra at Barrow (Figure 12-5). Although the low light inten- sity limits photosynthetic rate directly, low annual solar irradiance exerts its influence most strongly by limiting the length of the season for most Short active season Permafrost Poor soil aeration Low solar irradiance Low temperature J^ilT Organic matter accumulation Low pH Slow deconn position I Slow N-fixation Low nutrient availability Negligible chemical weathering Low precipitation Low atmospheric nutrient input Low productivity FIGURE 12-5. Causal relationships between low solar irradi- ance and low primary productivity of arctic coastal tundra. Thickness of arrows indicates magnitude of effect. 482 F. S. Chapin III et al. biological activity and by causing a low ambient temperature during that season. Because they have evolved numerous characteristics that enhance activity at low temperature and lengthen the season of activity, tundra organisms are generally relatively insensitive to the direct effects of low temperature, while still strongly influenced by the indirect temperature effects upon other environmental factors, such as length of growing season, permafrost formation and soil aeration. Low temperature and precipitation and a short growing season limit nutrient input to the tundra ecosystem from precipitation, weathering and nitrogen fixation. Therefore, the system depends almost entirely upon recycling of organically bound nutrients. Decomposition, the main bottleneck in nutrient recycling, is ultimately limited by low temperature. The nature of this limitation is largely indirect: the presence of perma- frost, resulting from negative mean annual temperature, restricts drain- age and soil aeration and thereby decomposition. In lower soil horizons aerobic decomposition is severely restricted, so nutrients accumulate in organic matter and pH decHnes. Soil microorganisms, and therefore decomposition, are more severely restricted in their activity by tundra conditions than are other trophic groups, not because they are any less well adapted, but because they must bear the full brunt of interaction between low temperature and anaerobiosis. Moreover, tundra decom- posers have a season of activity only slightly longer than that of primary producers, whereas in temperate regions the season of decomposition may greatly exceed that of most primary producers. Nutrients slowly ac- cumulate in soil organic matter, where they are unavailable to plants un- til decomposed. As a resuh, primary productivity, and therefore energy flow, are strongly limited. Animals may increase nutrient release to plants by stimulating or by short-circuiting the slow decomposition process. Soil invertebrates con- sume microorganisms, and lemmings consume vegetation. Both release much of the nutrient content from their food into the soil in soluble form, thus bypassing the decomposition process. Lemmings also fell standing litter and live plant biomass, thereby improving the quality of the substrate and the temperature regime for decomposition. The gradual accumulation of nutrients in soil organic matter through succession reduces rates of primary production and nutrient cycling. The continued functioning of the tundra ecosystem may depend upon per- turbations in the steady-state nutrient cycles. Perturbations that assist long-term cycling of nutrients include grazing, thaw lake cycles, and frost action. Thus the present functioning of the tundra system must be viewed in the context of processes that operate on time scales of hun- dreds and thousands of years. References Abele, G., D.A. Walker, J. Brown, M.C. Brewer and D.M. Atwood (1978) Effects of low ground pressure vehicle traffic on tundra at Lonely, Alaska. U.S. Army Cold Regions Research and Engineering Laboratory Special Report 78-16, 63 pp. Abrahamsen, G. (1973) Studies on body volume, body-surface area, den- sity and live weight of Enchytraeidae (Oligochaeta). Pedobio/ogia, 13: 6-15. Addison, J. A. (1977) Population dynamics and biology of Collembola on Truelove Lowland. In Truelove Lowland, Devon Island, Can- ada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 363-382. Addison, J. A. and D. Parkinson (1978) Influence of Collembolan feeding activities on soil metabolism at a high arctic site. Oikos, 30: 529-538. Agricultural Research Council (1965) The nutrient requirements of farm livestock. In No. 2 Ruminants. Technical Reviews and Summaries. Section 2 (Technical Committee on the Nutrient Requirements of Farm Livestock, Agricultural Research Council, Ed.). London: Her Majesty's Stationery Office, pp. 71-85. Ahmadjian, V. (1970) Adaptations of Antarctic terrestrial plants. In Antarctic Ecology (M.W. Holdgate, Ed.). London: Academic Press, pp. 801-811. Akiyama, M. (1970) Some soil algae from the arctic Alaska, Canada and Greenland. Shimane University of Natural Science, Memoir of the Faculty of Education, 4: 53-75. Alexander, M (1961) Introduction to Soil Microbiology. New York: John Wiley, 472 pp. Alexander, M. (1971) Microbial Ecology. New York: John Wiley, 511 pp. Alexander, V. (1974) A synthesis of the IBP Tundra Biome circumpolar study of nitrogen fixation. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and In- vertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr., and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 109-121. Alexander, V. (1975) Nitrogen fixation by blue-green algae in polar and subpolar regions. In Nitrogen Fixation by Free-living Microorgan- isms (W.D.P. Stewart, Ed.). Cambridge: Cambridge University Press, pp. 175-188. 483 484 References Alexander, V. and D.M. Schell (1973) Seasonal and spatial variation in nitrogen fixation in the Barrow, Alaska, tundra. Arctic and Alpine Research, 5: 77-88. Alexander, V., M. Billington and D.M. Schell (1974) The influence of abiotic factors on nitrogen fixation rates in the Barrow, Alaska, arc- tic tundra. Reports from the Kevo Subarctic Research Station, 11: 3-11. Alexander, V. and R.J. Barsdate (1975) Studies of nitrogen cycle process- es in arctic tundra systems. In Proceedings of the Circumpolar Con- ference on Northern Ecology, Ottawa 1975. Ottawa: National Re- search Council of Canada, pp. III-53 to III-64. Alexander, V., M. Billington and D.M. Schell (1978) Nitrogen fixation in arctic and alpine tundra. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 539-558. Allden, W.G. and LA. McD. Whittaker (1970) The determinants of herbage intake by grazing sheep: The interrelationship of factors in- fluencing herbage intake and availability. Australian Journal of Agricultural Research, 21: 755-766. Allessio, M.L. and L.L. Tieszen (1975a) Patterns of carbon allocation in arctic tundra grass, Dupontia fisheri (Gramineae), at Barrow, Alaska. American Journal of Botany, 62: 797-807. Allessio, M.L. and L.L. Tieszen (1975b) Leaf age effect on translocation and distribution of '"C photoassimilate in Dupontia at Barrow, Alaska. Arctic and Alpine Research, 7: 3-12. Allessio, M.L. and L.L. Tieszen (1978) Translocation and allocation of '"C-photoassimilate by Dupontia fisheri. In Vegetation and Produc- tion Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 393-413. Anderson, E.A. and N.H. Crawford (1964) The synthesis of continuous snowmelt runoff hydrographs on a digital computer. Stanford Uni- versity Department of Civil Engineering, Technical Report No. 36. Anderson, J.E. and S.J. McNaughton (1973) Effects of low soil temp- erature on transpiration, photosynthesis, leaf relative water content, and growth among elevationally diverse plant populations. Ecology, 54: 1220-1233. Anderson, L.E. and P.P. Bourdeau (1955) Water relations of two species of terrestrial mosses. Ecology, 36: 206-212. Anderson, M.C. (1966) Stand structure and light penetration. II. A theo- retical analysis. Journal of Applied Ecology, 3: 41-54. Andrews, R.V., K. Ryan, R. Strohbehn and M. Ryan-Klein (1975) Physiological and demographic profiles of brown lemmings during their cycle of abundance. Physiological Zoology, 48: 64-83. Antibus, R.K. and A.E. Linkins (1978) Ectomycorrhizal fungi of Salix References 485 rotundifolia Trautv. I. Impact of surface applied Prudhoe Bay crude oil on mycorrhizal structure and composition. Arctic, 31: 366-380. Aristoskaya, T.V. and O.M. Parinkina (1972) Preliminary results of the IBP studies of soil microbiology in tundra. In Proceedings, IV Inter- national Meeting on the Biological Productivity of Tundra, Lenin- grad (F.E. Wielgolaski and Th. Rosswall, Eds.). Stockholm: Inter- national Biological Programme Tundra Biome Steering Committee, pp. 80-92. Arnborg, L., H.J. Walker and J. Peippo (1966) Water discharge in the Colville River, 1962. Geografiska Annaler, 48A: 195-210. Arnold, G.W. (1964) Factors within plant associations affecting the behavior and performance of grazing animals. In Grazing in Terres- trial and Marine Environments: A Symposium of the British Eco- logical Society, Bangor, 11-14 April 1962 (D.J. Crisp, Ed.). Oxford: Blackwell Scientific Publications, pp. 133-154. Arnold, G.W. and M.L. Dudzinski (1967) Studies on the diet of the graz- ing animal. III. The effect of pasture species and pasture structure on the herbage intake of sheep. Australian Journal of Agricultural Research, 18: 657-666. Ashkenazie, S. and U.N. Safriel (1979a) Breeding cycle and behavior of the semipalmated sandpiper at Barrow, Alaska. The Auk, 96:56-67. Ashkenazie, S. and U.N. Safriel (1979b) Time-energy budget of the semi- palmated sandpiper Calidris pusilla at Barrow, Alaska. Ecology, 60: 783-799. Atlas, R.M., E.A. Schofield, F.A. Morelh and R.E. Cameron (1976) Ef- fects of petroleum pollutants on arctic microbial populations. Envi- ronmental Pollution, 10: 35-43. Auclair, A.N.D., A. Bouchard and J. Pajazkowski (1976) Productivity relations in a Carex-dominated ecosystem. Oecologia, 26: 9-31. Babb, T.A. and L.C. Bliss (1974) Effects of physical disturbance on arc- tic vegetation in the Queen Elizabeth Islands. Journal of Applied Ecology, 11: 549-562. Babb, T.A. and D.W.A. Whitfield (1977) Mineral nutrient cycHng and limitation of plant growth in the Truelove Lowland ecosystem. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosys- tem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 589-606. Babiuk, L.A. and E.A. Paul (1970) The use of fluorescein isothiocyanate in the determination of the bacterial biomass of grassland soil. Can- adian Journal of Microbiology, 16: 57-62. Baile, C.A. and J. Mayer (1970) Hypothalamic centres: feedbacks and receptor sites in the short term control of feed intake. In Physiology of Digestion and Metabolism in the Ruminant (A.T. Phillipson, 486 References Ed.). Newcastle Upon Tyne: Oriel Press, pp. 254-263. Baker, J.H. (1970a) Quantitative study of yeasts and bacteria in a Signy Island peat. British Antarctic Survey Bulletin, 23: 51-55. Baker, J.H. (1970b) Yeast, moulds, and bacteria from an acid peat on Signy Island. In Antarctic Ecology (M.W. Holdgate, Ed.). New York: Academic Press, pp. 717-722. Banks, E.M., R.J. Brooks and J. Schnell (1975) Radiotracking study of home range and activity of the brown lemming {Lemmus trimucron- atus). Journal of Mammalogy, 56: 888-901. Barel, D. and R.J. Barsdate (1978) Phosphorus dynamics of wet coastal tundra soils near Barrow, Alaska. In Environmental Chemistry and Cycling Processes: Proceedings of Symposium, A ugusta, Georgia, April 28-May 1, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. Department of Energy, CONF-760429, pp. 516-537. Barghoorn, E.S. and R.L. Nichols (1961) Sulfate-reducing bacteria in pyritic sediments in Antarctica. Science, 134: 190. Barkley, S.A. (1976) The influence of mineral nutrients on the lemmings of Point Barrow, Alaska. M.S. thesis, San Diego State University, 110 pp. Barkley, S.A., D. Barel, W.A. Stoner and P.C. Miller (1978) Controls on decomposition and mineral release in wet meadow tundra — A simulation approach. In Environmental Chemistry and Cycling Pro- cesses: Proceedings of Symposium, Augusta, Georgia, April 28-May 1, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. De- partment of Energy, CONF-760429, pp. 754-778. Barkley, S.A., G.O. Batzli and B.D. Collier (1980) Nutritional ecology of microtine rodents: a simulation model of mineral nutrition for brown lemmings. Oikos, 34: 103-144. Barrow, N.J. (1975) Chemical form of phosphorus in sheep faeces. Australian Journal of Soil Research, 13: 63-67. Barry, R.G. (1967) Seasonal location of the Arctic Front in North America. Geographical Bulletin, 9: 79-95. Barry, R.G. and J.D, Ives (1974) Introduction. In Arctic and Alpine En- vironments (J.D. Ives and R.G. Barry, Eds.). London: Methuen, pp. 1-13. Barry, R.G., R.E. Moritz and J.C. Rogers (1976) Studies of climate and fast ice interaction during the decay season along the Beaufort Sea coast. In Science in Alaska, 1976: Proceedings Twenty-seventh Alaska Science Conference, Fairbanks, Alaska, August 4 to August 7, 1976, Vol. II, Resource Development — Processes and Problems. Fairbanks, Alaska: Alaska Division, AAAS, pp. 213-228. Barsdate, R.J., R.T. Prentki and T. Fenchel (1974) The phosphorus cy- cle of model ecosystems: Significance for decomposer food chains and effect of bacterial grazers. Oikos, 25: 239-251. References 487 Barsdate, R.J. and V. Alexander (1975) The nitrogen balance of arctic tundra: pathways, rates and environmental implications. Journal of Environmental Quality, 4: 111-117. BatzH, G.O. (1974) Production, assimilation, and accumulation of or- ganic matter in ecosystems. Journal of Theoretical Biology, 45: 205-217. Batzli, G.O. (1975a) The role of small animals in arctic ecosystems. In Small Mammals: Their Productivity and Population Dynamics (F.B. Golley, K. Petrusewicz and L. Ryszkowski, Eds.). Cambridge: Cambridge University Press, pp. 243-268. BatzH, G.O. (1975b) The influence of grazers on tundra vegetation and soils. In Proceedings of Circumpolar Conference on Northern Ecol- ogy, Ottawa, 1975. Ottawa: National Research Council of Canada, Scientific Committee on Problems of the Environment, pp. 1-215 to 1-225. Batzli, G.O. (1978) The role of herbivores in mineral cycling. In En- vironmental Chemistry and Cycling Processes: Proceedings of Sym- posium, Augusta, Georgia, April 28-May 1, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. Department of Energy, CONF- 760429, pp. 95-112. Batzli, G.O. (1981) Populations and energetics of small mammals in the tundra ecosystem. In Tundra Ecosystems: A Comparative Analysis (L.C. Bliss, O.W. Heal and J.J. Moore, Eds.). Cambridge: Cam- bridge University Press, pp. 377-396 (in press). Batzli, G.O., N.S. Stenseth and B.M. Fitzgerald (1974) Growth and sur- vival of suckling brown lemmings Lemmus trimucronatus. Journal of Mammalogy, 55: 828-831. Batzli, G.O. and F.R. Cole (1979) Nutritional ecology of microtine rodents: Digestibility of forage. Journal of Mammalogy, 60: 740- 750. Batzli, G.O. and H.G. Jung (1980) Nutritional ecology of microtine ro- dents: Resource utilization near Atkasook, Alaska. Arctic and Alpine Research, 12: 483-499. Batzli, G.O. and S.T. Sobaski (1980) Distribution, abundance, and for- aging patterns of ground squirrels near Atkasook, Alaska. Arctic and Alpine Research, 12: 501-510. Batzli, G.O., R.G. White and F.L. Bunnell (1981) Herbivory: a strategy of tundra consumers. In Tundra Ecosystems: A Comparative Anal- ysis (L.C. Bliss, O.W. Heal and J.J. Moore, Eds.). Cambridge: Cambridge University Press, pp. 359-375 (in press). Batzli, G.O., H.G. Jung and G. Gunterspergen (In press) Nutritional ecology of microtine rodents: Linear functional response curves for brown lemmings. Oikos. Baumgardt, B.R. (1970) Control of feed intake in the regulation of 488 References energy balance. In Physiology of Digestion and Metabolism in the Ruminant (A.T. Phillipson, Ed.). Newcastle Upon Tyne: Oriel Press, pp. 235-253. Beadle, N.C.W. (1962) Soil phosphate and the delimitation of plant communities in eastern Australia II. Ecology, 43: 281-288. Bee, J.W. and E.R. Hall (1956) Mammals of Northern Alaska on the Arctic Slope. University of Kansas Miscellaneous Publication of the Museum of Natural History 8, 309 pp. Behan, V.M. (1978) Diversity, distribution, and feeding habits of North American Arctic soil Acari. Ph.D. dissertation, MacDonald Col- lege, McGill University, 428 pp. Behan, V.M. and S.B. Hill (1978) Feeding habits and spore dispersal of Oribatid mites in the North American Arctic. Revue d'Ecologie et de Biologic du Sol, 15: 497-516. Benson, C.S. (1969) The seasonal snow cover of arctic Alaska. Arctic In- stitute of North America Research Paper 51, 47 pp. Benson, C, B. Holmgren, R. Timmer, G. Weller and S. Parrish (1975) Observations on the seasonal snow cover and radiation climate at Prudhoe Bay, Alaska, during 1972. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 13-50. Bergman, R.D., R.L. Howard, K.F. Abraham and M.W. Weller (1977) Water birds and their wetland resources in relation to oil develop- ment at Storkersen Point, Alaska. U.S. Department of Interior, Fish and Wildlife Service Resource Publication 129, 38 pp. Bernard, J.M. (1974) Seasonal changes in standing crop and primary production in sedge wetland and an adjacent dry oil-field in central Minnesota. Ecology, 55: 350-359. Beschel, R.E. (1970) The diversity of tundra vegetation. In Proceedings of the Conference on Productivity and Conservation in Northern Circumpolar Lands, Edmonton, Alberta, 15 to 17 October J 969 (W.A. Fuller and P.G. Kevan, Eds.). lUCN Publication N.S. 16. Morges, Switzerland: International Union for Conservation of Na- ture and Natural Resources, pp. 85-92. Biederbeck, V.O. and C.A. Campbell (1971) Influence of simulated fall and spring conditions on the soil system. I. Effects on soil micro- flora. So/7 Science Society of America, Proceedings, 35: 474-479. Bieleski, R.L. (1973) Phosphate pools, phosphate transport and phos- phate availability. Annual Review of Plant Physiology, 24: 225-252. Bilello, M.A. (1966) Survey of arctic and subarctic temperature inver- sions. U.S. Army Cold Regions Research and Engineering Labora- tory Technical Report 161, 35 pp. Billings, W.D. (1973) Arctic and alpine vegetations: Similarities, dif- References 489 ferences and susceptibility to disturbances. BioScience, 23: 697-704. Billings, W.D. (1974a) Arctic and alpine vegetation: Plant adaptations to cold summer climates. In Arctic and Alpine Environments (J.D. Ives and R.G. Barry, Eds.). London: Methuen, pp. 403-444. Billings, W.D. (1974b) Adaptations and origins of alpine plants. Arctic and A/pine Research, 6: 129-142. Billings, W.D. and R.J. Morris (1951) Reflection of visible and infrared radiation from leaves of different ecological groups. American Journal of Botany, 38: 327-331. Billings, W.D. and H.A. Mooney (1968) The ecology of arctic and alpine plants. Biological Reviews of the Cambridge Philosophical Society, 43: 481-529. Billings, W.D., P.J. Godfrey, B.F. Chabot and D.P. Bourque (1971) Metabolic acclimation to temperature in arctic and alpine ecotypes of Oxyria digyna. Arctic and Alpine Research, 3: 277-289. Billings, W.D., G.R. Shaver and A.W. Trent (1973) Temperature effects of growth and respiration of roots and rhizomes in tundra gramin- oids. In Primary Production and Production Processes: Proceed- ings of the Conference, Dublin, Ireland, April 1973 (L.C. Bliss and F.E. Wielgolaski, Eds.). Stockholm: International Biological Pro- gramme Tundra Biome Steering Committee, pp. 57-63. BilHngs, W.D., K.M. Peterson, G.R. Shaver and A.W. Trent (1977) Root growth, respiration, and carbon dioxide evolution in an arctic tundra soil. Arctic and Alpine Research, 9: 129-137. BilHngs, W.D., K.M. Peterson and G.R. Shaver (1978) Growth, turn- over, and respiration rates of roots and tillers in tundra graminoids. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 415-434. Billore, S.K. and L.P. Mall (1976) Nutrient composition and inventory in a tropical grassland. Plant and Soil, 45: 509-520. Black, R.F. (1954) Precipitation at Barrow, Alaska, greater than record- ed. Transactions, American Geophysical Union, 35: 203-206. Black, R.F. (1964) Gubik formation of Quaternary age in northern Alaska. U.S. Geological Survey Professional Paper 302-C, pp. 59-91. Blair, W.F. (1977) Big Biology: The US/IBP. Stroudsburg, Pa.: Dow- den, Hutchinson and Ross, 261 pp. Blaxter, K.L. (1962) The Energy Metabolism of Ruminants. Springfield, 111.: C.C. Thomas, 332 pp. Blaxter, K.L. (1967) The efficiency of energy transformation in rumi- nants. In Energy Metabolism of Farm Animals (K.L. Blaxter et al., Eds.). European Association of Animal Production Publication No. 12, pp. 21-28. Bliss, L.C. (1956) A comparison of plant development in microenviron- 490 References ments of arctic and alpine tundras. Ecological Monographs, 26: 303-337. Bliss, L.C. (1962a) Adaptations of arctic and alpine plants to en- vironmental conditions. Arctic, 15: 117-144. Bliss, L.C. (1962b) Caloric and lipid content in alpine tundra plants. Ecology, 43: 753-757. Bliss, L.C. (1963) Alpine plant communities of the Presidential Range, New Hampshire. Ecology, 44: 678-697. Bliss, L.C. (1966) Plant productivity in alpine microenvironments on Mt. Washington, New Hampshire. Ecological Monographs, 'i^: 125-155. Bliss, L.C. (1970) Oil and the ecology of the Arctic. Transactions of the Royal Society of Canada, 4th Series, VII: 1-12. Bliss, L.C. (1971) Arctic and alpine plant life cycles. Annual Review of Ecology and Systematics, 2: 405-438. Bliss, L.C. (Ed.) (1977) Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem. Edmonton: The University of Alberta Press, 714 pp. Bliss, L.C, W.O. Pruitt, Jr., R.A. Hemstock, F.K. Hare, J.D. Ives and G.A. McKay (1970) The tundra environment: a symposium of Sec- tion III of the Royal Society of Canada, Winnipeg. Transactions of the Royal Society of Canada, 4th Series, VH: 50 pp. Bliss, L.C. and R.W. Wein (1972) Plant community responses to disturb- ances in the western Canadian Arctic. Canadian Journal of Botany, 50: 1097-1109. Bliss, L.C, G.M. Courtin, D.L. Pattie, R.R. Riewe, D.W.A. Whitfield and P. Widden (1973) Arctic tundra ecosystems. Annual Review of Ecology and Systematics, 4: 359-399. Bliss, L.C, O.W. Heal and J.J. Moore (Eds.) (1981) Tundra Ecosystems: A Comparative Analysis. Cambridge: Cambridge University Press, 813 pp. (in press). Bodvarsson, H. (1970) Alimentary studies of seven common soil- inhabiting Collembola of Southern Sweden. Entomologica Scandin- avica, 1: 74-80. Bokhari, U.G. and J.S. Singh (1975) Standing state and cycle of nitrogen in soil-vegetation components of prairie ecosystems. Annals of Botany, N.S., 39: 273-285. Bormann, F.H. and G.E. Likens (1979) Pattern and Process in a For- ested Ecosystem. New York: Springer-Verlag, 253 pp. Bowen, E.J. (1931) Water conduction in Polytrichum commune. Annals of Botany, 45: 175-200. Bowen, E.J, (1933) The mechanism of water conduction in the Musci considered in relation to habitat. Annals of Botany, 47: 401-422, 635-661, 889-912. Bowen, G.D. and A.D. Rovira (1967) Phosphate uptake along attached References 491 and excised wheat roots measured by an automatic scanning meth- od. Australian Journal of Biological Science, 20: 369-378. Boyd, W.L. (1967) Ecology and physiology of soil microorganisms in polar regions. In Proceedings of the Symposium on Pacific-Antarc- tic Sciences. National Science Museum, Tokyo, J ARE Scientific Re- port, Special Issue 1, pp. 265-275. Boyd, W.L. and J.W. Boyd (1962) Presence oi Azotobacter species in polar regions. Journal of Bacteriology, 83: 429-430. Boyd, W.L., T.J. Staley and J.W. Boyd (1966) Ecology of soil micro- organisms in Antarctica. In Antarctic Soils and Soil Forming Pro- cesses (J.C.F. Tedrow, Ed.). Washington, D.C.: American Geo- physical Union, pp. 125-159. Breed, R.S., E.G.D. Murray and N.R. Smith (1957) Bergey's Manual of Determinative Bacteriology. 7th edition. London: Tindall and Cox, 1094 pp. Bremner, J.M. and K. Shaw (1958) Denitrification in soil. II. Factors af- fecting denitrification. Journal of Agricultural Science, 51: 40-52. Brewer, M.C. (1958) Some results of geothermal investigations of per- mafrost in northern Alaska. Transactions, American Geophysical Union, 39: 19-26. Brewster, J.L., K.S.S. Bhat and P.H. Nye (1975) The possibility of pre- dicting solute uptake and plant growth response from independently measured soil and plant characteristics. III. The growth and uptake of onions in a sod fertilized to different initial levels of phosphate and a comparison of the results with model predictions. Plant and Soil, 42: 197-226. Breymeyer, A.I. and G.M. Van Dyne (Eds.) (1980) Grasslands, Systems Analysis and Man. Cambridge: Cambridge University Press, 950 pp. Britton, M.E. (1957) Vegetation of the arctic tundra. In Arctic Biology: Eighteenth Annual Biology Colloquium (H.P. Hansen, Ed.). Cor- vallis: Oregon State University Press, pp. 26-61. Britton, M.E. (Ed.) (1973) Alaskan Arctic Tundra: Proceedings of the 25th Anniversary Celebration of the Naval Arctic Research Labora- tory. Arctic Institute of North America Technical Paper 25, 224 pp. Brody, S. (1945) Bioenergetics and Growth. New York: Reinhold, 1023 pp. Brouwer, R. (1962) Nutritive influences on the distribution of dry matter in the plant. Netherlands Journal of Agricultural Science, 10: 399- 408. Brouwer, R. (1965) Root growth of grasses and cereals. In The Growth of Cereals and Grasses: Proceedings of the Twelfth Easter School of Agricultural Science, University of Nottingham (F.L. Milthorpe and J.D. Ivins, Eds.). London: Butterworth, pp. 153-166. 492 References Brown, J. (1965) Radiocarbon dating, Barrow, Alaska. Arctic, 18: 36-48. Brown, J. (1969) Soil properties developed on the complex tundra relief of northern Alaska. Biuletyn Peryglacjalny, 18: 153-167. Brown, J. (1970) Tundra Biome applies a new look to ecological prob- lems in Alaska. Northern Engineer, 2: 9. Brown, J. (Ed.) (1975) Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska. Biological Papers of the Universi- ty of Alaska, Special Report 2, 215 pp. Brown, J., S.L. Dingman and R.I. Lewellen (1968) Hydrology of a small drainage basin on the Coastal Plain of northern Alaska. U.S. Army Cold Regions Research and Engineering Laboratory Research Re- port 240, 18 pp. Brown, J., W. Rickard and D. Victor (1969) The effect of disturbance on permafrost terrain. U.S. Army Cold Regions Research and Engi- neering Laboratory Special Report 138, 13 pp. Brown, J., H. Coulombe and F.A. Pitelka (1970) Structure and function of the tundra ecosystem at Barrow, Alaska. In Proceedings of the Conference on Productivity and Conservation in Northern Circum- polar Lands, Edmonton, Alberta, 15 to 17 October 1969 (W.A. Ful- ler and P.O. Kevan, Eds.). lUCN Publication N.S. 16. Morges, Switzerland: International Union for Conservation of Nature and Natural Resources, pp. 41-71. Brown, J., R.K. Haugen and S. Parrish (1975) Selected climatic and soil thermal characteristics of the Prudhoe Bay region. In Ecological In- vestigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alas- ka, Special Report 2, pp. 3-11. Brown, J. and N.A. Grave (1979) Physical and thermal disturbance and protection of permafrost. In Proceedings of the Third Interna- tional Conference on Permafrost, July 10-13, 1978, Edmonton, Al- berta, Canada. Vol. 2. Ottawa: National Research Council of Can- ada, pp. 51-91. Brown, J.H. and R.C. Lasiewski (1972) Metabolism of weasels: The cost of being long and thin. Ecology, 53: 939-943. Bryson, R.A. (1966) Air masses, streamlines and boreal forest. Geo- graphical Bulletin, 8: 228-269. Bryson, R.A. and P.M. Kuhn (1961) Stress-differential induced diver- gence with application to littoral precipitation. Erdkunde, 15: 287- 294. Bunnell, F.L. (1973) Theological ecology or models and the real world. Forestry Chronicle, 49 : 161-171. Bunnell, F.L. and P. Dowding (1974) ABISKO— A generalized decom- position model for comparisons between tundra sites. In Soil Organ- isms and Decomposition in Tundra: Proceedings of the Microbiol- References 493 ogy, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr., and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 227- 247. Bunnell, F.L. and D.E.N. Tait (1974) Mathematical simulation models of decomposition processes. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flan- agan, Eds.). Stockholm: International Biological Programme Tun- dra Biome Steering Committee, pp. 207-225. Bunnell, F.L., S.F. MacLean, Jr. and J. Brown (1975) Barrow, Alaska, U.S.A. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Bio- logical Productivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stock- holm: Swedish Natural Science Research Council, pp. 73-124. Bunnell, F.L. and K.A. Scoullar (1975) ABISKO II— A computer simulation model of carbon flux in tundra ecosystems. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Productivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 425-448. Bunnell, F.L., D.E.N. Tait, P.W. Flanagan and K. Van Cleve (1977a) Microbial respiration and substrate weight loss. I. A general model of the influence of abiotic variables. Soil Biology and Biochemistry, 9: 33-40. Bunnell, F.L,, D.E.N. Tait and P.W. Flanagan (1977b) Microbial respir- ation and substrate weight loss. II. A model of the influence of chemical composition. So/7 Biology and Biochemistry, 9: 41-47. Bunnell, F.L. and K.A. Scoullar (1981) Between-site comparisons of car- bon flux in tundra using simulation models. In Tundra Ecosystems: A Comparative Analysis (L.C. Bliss, O.W. Heal and J.J. Moore, Eds.). Cambridge: Cambridge University Press, pp. 685-715 (in press). Bunt, J.S. (1965) Observations on the fungi of Macquarie Island. ANARE Scientific Reports, Series B (II) Botany, 78: 22 pp. Burges, A. (1958) Micro-organisms in the Soil. London: Hutchinson, 188 pp. Byers, G.W. (1969) Evolution of wing reduction in crane flies (Diptera: Tipulidae). Evolution, 23: 346-354. Caldwell, M.M. (1970a) The effect of wind on stomatal aperture, photo- 494 References synthesis, and transpiration of Rhododendron ferrugineum L. and Pinus Cembra L. Centralblatt fur das Gesamte Forstwesen, 87: 193- 201. Caldwell, M.M. (1970b) Plant gas exchange at high wind speeds. Plant Physiology, 46: 535-537. Caldwell, M.M., L.L. Tieszen and M. Fareed (1974) The canopy struc- ture of tundra plant communities at Barrow, Alaska, and Niwot Ridge, Colorado. Arctic and Alpine Research, 6: 151-159. Caldwell, M.M., D.A. Johnson and M. Fareed (1978) Constraints on tundra productivity: Photosynthetic capacity in relation to solar radiation utilization and water stress in arctic and alpine tundras. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 323-342. Calef, G.W. (1978) Population status of caribou in the Northwest Terri- tories. In Parameters of Caribou Population Ecology in Alaska. Proceedings of a Symposium and Workshop (D.R. Klein and R.G. White, Eds.). Biological Papers of the University of Alaska, Special Report No. 3, pp. 9-16. Calef, G.W. and G.M. Lortie (1973) Observations of the Porcupine caribou herd, 1972. Section I, Appendix I, Wildlife. Interim Report No. 3. Winnipeg: Environment Protection Board, Canada, 127 pp. Callaghan, T.V. (1976) Growth and population dynamics of Carex bigelowii in an alpine environment. Oikos, 21: 402-413. Cameron, R.D. and J.R. Luick (1972) Seasonal changes in total body water, extracellular fluid, and blood volume in grazing reindeer. Canadian Journal of Zoology, 50: 107-116. Cameron, R.D. and K.R. Whitten (1979) Seasonal movements and sex- ual segregation of caribou determined by aerial survey. Journal of Wildlife Management, 43: 626-633. Cameron, R.E. (1972) Ecology of blue-green algae in Antarctic soils. In First International Symposium on Taxonomy and Ecology of Blue- Green Algae, Madras, 1970 (T.V. Desikachary, Ed.). Madras: Bangalore Press, pp. 353-384. Cameron, R.E., A.D. Knox and F.A. Morelli (1978) The role of algae in tundra soils. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 207-227. Campbell, W.B., R.W. Harris and R.E. Benoit (1973) Response of Alas- kan tundra microflora to crude oil spill. In Proceedings of the Sym- posium on the Impact of Oil Resource Development on Northern Plant Communities: Presented at the 23 rd A A AS Alaska Science Conference, Fairbanks, Alaska, 17 August 1972 (B.H. McCown and D.R. Simpson, Coords.). University of Alaska, Institute of Arc- tic Biology, Occasional Publications on Northern Life 1, pp. 53-62. References 495 Cannon, F.C., L.K. Dunican and P.V. Jennings (1970) The significance of sulphur bacteria in blanket peat. In Sulphur in Agriculture. Dublin: An Foras Taluntias, pp. 134-157. Cantlon, J.E. (1961) Plant cover in relation to macro-, meso- and micro- relief. Final report to Arctic Institute of North America on Con- tracts ONR-208 and ONR-212, 128 pp. Carey, R.W. and J. A. Berry (1978) Effects of low temperature on res- piration and uptake of rubidium ions by excised barley and corn roots. Plant Physiology, 61: 858-860. Carl, E.A. (1971) Population control in arctic ground squirrels. Ecology, 52: 395-413. Carlson, R.F., W. Norton and J. McDougall (1974) Modeling snowmelt runoff in an arctic coastal plain. University of Alaska Institute of Water Resources Report IWR-43, 72 pp. Carson, C.E. (1968) Radiocarbon dating of lacustrine strands in arctic Alaska. Arctic, 21: 12-26. Carson, C.E. and K. Hussey (1962) The oriented lakes of arctic Alaska. Journal of Geology, 70: 417-439. Chabot, B.F. and W.D. Billings (1972) Origins and ecology of the Sier- ran alpine flora and vegetation. Ecological Monographs, 42: 163- 199. Chabot, B.F., J.F. Chabot and W.D. Billings (1972) Ribulose-1,5 diphosphate carboxylase activity in arctic and alpine populations of Oxyria digyna. Photosynthetica, 6: 364-369. Challinor, J.L. and P.L. Gersper (1975) Vehicle perturbation effects upon a tundra soil-plant system. II. Effects on the chemical regime. Soil Science Society of America, Proceedings, 39: 689-695. Chang, S.C. (1940) Assimilation of phosphorus by a mixed soil popula- tion and by pure cultures of soil fungi. Soil Science, 49: 197-210. Chang, S.C. and M.L. Jackson (1957) Fractionation of soil phosphorus. Soil Science, 84: 133-144. Chapin, F.S., III (1974a) Morphological and physiological mechanism of temperature compensation in phosphate absorption along a lati- tudinal gradient. Ecology, 55: 1180-1198. Chapin, F.S., III (1974b) Phosphate absorption capacity and acclimation potential in plants along a latitudinal gradient. Science, 183: 521- 523. Chapin, F.S., III (1975) Nutrient/carbon costs associated with tundra adaptations to a cold nutrient-poor environment. Proceedings of the Circumpolar Conference on Northern Ecology, Ottawa 1975. Ot- tawa: National Research Council of Canada, Scientific Committee on Problems of the Environment, pp. 1-183 to 1-194. Chapin, F.S., III (1977) Temperature compensation in phosphate ab- sorption occurring over diverse time scales. Arctic and Alpine Re- 496 References search, 9: 139-148. Chapin, F.S., III (1978) Phosphate uptake and utilization by Barrow tundra vegetation. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer- Verlag, pp. 483-507. Chapin, F.S., III (1980a) The mineral nutrition of wild plants. Annual Review of Ecology and Systematics, 11: 233-260. Chapin, F.S., 111 (1980b) Nutrient allocation and responses to defolia- tion in tundra plants. Arctic and Alpine Research, 12: 553-563. Chapin, F.S., III, K. Van Cleve and L.L. Tieszen (1975) Seasonal nutri- ent dynamics of tundra vegetation at Barrow, Alaska. Arctic and Alpine Research, 7: 209-226. Chapin, F.S. and A.J. Bloom (1976) Phosphate absorption: Adaptation of tundra graminoids to a low temperature, low phosphorus en- vironment. Oikos, 26: 111-121. Chapin, F.S., III, R.J. Barsdate and D. Barel (1978) Phosphorus cychng in Alaskan coastal tundra: a hypothesis for the regulation of nutri- ent cycling. Oikos, 31: 189-199. Chapin, F.S., III, and K. Van Cleve (1978) Nitrogen and phosphorus dis- tribution in an Alaskan tussock tundra ecosystem: natural patterns and implications for development. In Environmental Chemistry and Cycling Processes: Proceedings of Symposium, A ugusta, Georgia, April 28-1 May, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. Department of Energy, CONF-760429, pp. 738-753. Chapin, F.S., III, K. Van Cleve and M.C. Chapin (1979) Soil tempera- ture and nutrient cycling in the tussock growth form of Eriophorum vaginatum. Journal of Ecology, 67: 169-189. Chapin, F.S., III, D.A. Johnson and J.D. McKendrick (1980) Sea- sonal movement of nutrients in plants of differing growth form in an Alaskan tundra ecosystem: Implications for herbivory. Journal of Ecology, 68: 189-209. Chapman, S.B. (1979) Some interrelationships between soil and root res- piration in lowland Calluna heathland in southern England. Journal of Ecology, 67: 1-20. Chernov, Yu.I. (1978) Adaptive features of the life cycles of tundra zone insects. Journal of General Biology, 39(3): 394-402 (in Russian). Chernov, Yu.I. and E.N. Savchenko (1965) Ecology and premarginal development phases of the arctic Tipula (Pterolachisus) carinifrons Holm. (Diptera: Tipulidae). Zoologicheskii Zhurnal, 44: 111-119. Chew, R.M. (1974) Consumers as regulators of ecosystems: an alter- native to energetics. Ohio Journal of Science, 74: 359-370. Chitty, D. (1967) The natural selection of self-regulatory behaviour in animal populations. Proceedings of Ecological Society of Australia, References 497 2: 51-78. Christensen, P.J. (1974) A microbiological study of some lake waters and sediments from the Mackenzie Valley with special reference to Cytophagas. Arctic, 27: 309-311. Christian, J.J. and D.E. Davis (1965) Endocrines, behavior and popula- tion. Science, 146: 1550-1560. Christophersen, J. (1973) I. Basic aspects of temperature action on microorganisms. In Temperature and Life {H. Precht, J. Christoph- ersen, H. Hensel and W. Larcher, Eds.). New York: Springer- Verlag, pp. 3-59. Churchill, E.D. and H.C. Hanson (1958) The concept of climax in arctic and alpine vegetation. Botanical Review, 24: 127-191. Clarholm, M., V. Lid-Torsvik and J.H. Baker (1975) Bacterial popula- tions of some Fennoscandian tundra soils. In Fennoscandian Tun- dra Ecosystems. Part I. Plants and Microorganisms (F.E. Wielgo- laski, Ed.). New York: Springer-Verlag, pp. 251-260. Clark, F.E. (1967) Bacteria in soil. In So/7 Biology (A. Burges and F. Raw, Eds.). London: Academic Press, pp. 14-49. Clark, F.E. (1977) Internal cycling of nitrogen in shortgrass prairie. Ecology, 58: 1322-1333. Clark, F.E. and E.A. Paul (1970) The microflora of grassland. /Ic^va/ices in Agronomy, 11: 375-435. Clebsch, E.E.C. and R.E. Shanks (1968) Summer climatic gradients and vegetation near Barrow, Alaska. Arctic, 21: 161-171. Cleland, R. (1971) Cell wall extension. Annual Review of Plant Physiol- ogy, 11: 197-222. Clement, L.E. (1975) The ecology of Tipula carinifrons Holm. (Diptera: Tipulidae) in the arctic coastal tundra of northern Alaska. M.S. thesis. University of Alaska, Fairbanks, 96 pp. Cochrane, V.W. (1958) Physiology of Fungi. New York: John Wiley, 524 pp. Cole, C.V., D.L. Grunes, L.K. Porter and S.R. Olsen (1963) The effect of nitrogen on short-term phosphorus absorption and translocation in corn (Zea mays). Soil Science Society of America, Proceedings, 11: 611-614. Cole, D.W., S.P. Gessel and S.F. Dice (1967) Distribution and cycling of nitrogen, phosphorus, potassium and calcium in a second-growth Douglas fir ecosystem. In Symposium on Primary Productivity and Mineral Cycling in Natural Ecosystems (H.E. Young, Ed.). Orono, Maine: University of Maine Press, pp. 197-232. Cole, D.W., J. Turner and C. Bledsoe (1977) Requirement and uptake of mineral nutrients in coniferous ecosystems. In The Belowground Ecosystem: A Synthesis of Plant-Associated Processes (J.K. Mar- shall, Ed.). Colorado State University, Range Science Department 498 References Science Series 26, pp. 171-176. Coleman, D.C., R. Andrews, J.E. Ellis and J.S. Singh (1976) Energy flow and partitioning in selected man-managed and natural ecosys- tems. Agro-Ecosystems, 3: 45-54. Coleman, D.C., C.V. Cole, R.V. Anderson, M. Blaha, M.K. Campion, M. Clarholm, E.T. Elliot, H.W. Hunt, B. Schaefer and J. Sinclair (1977) An analysis of rhizosphere-saprophage interactions in terres- trial ecosystems. In Soil Organisms as Components of Ecosystems: International Soil Zoology Colloquium, Uppsala, Sweden, June 21-25, 1976 (U. Lohm and T. Persson, Eds.). Ecological Bulletins 25. Stockholm: Swedish Natural Science Research Council, pp. 299-309. Colinvaux, P. A. (1964) Origin of ice ages: pollen evidence from arctic Alaska. Science, 145: 707-708; 147: 633. Collier, B.D., N.C. Stenseth, S. Barkley and R. Osborn (1975) A simula- tion model of energy acquisition and utilization by the brown lem- ming (Lemmus trimucronatus) at Barrow, Alaska. Oikos, 26: 276-294. CoUins, N.J. and W.C. Oechel (1974) The pattern of growth and translo- cation of photosynthate in a tundra moss, Polytrichum alpinum. Canadian Journal of Botany, 52: 355-363. Conover, J.H. (1960) Macro- and micro-climatology of the Arctic Slope of Alaska. U.S. Army, Headquarters Quartermaster Research and Engineering Command, Natick, Massachusetts, Technical Report EP-139, 65 pp. Cosgrove, D.J. (1967) Metabolism of organic phosphates in soil. In Soil Biochemistry, Vol. I (A.D. McLaren and G.H. Peterson, Eds.). New York: Marcel Dekker, pp. 216-228. Coulson, J.C. and J. Butterfield (1978) An investigation of the biotic factors determining the rates of plant decomposition on blanket bog. Journal of Ecology, 66: 631-650. Coulson, J.C. and J.B. Whittaker (1978) Ecology of moorland animals. In Production Ecology of British Moors and Montane Grasslands (O.W. Heal and D.F. Perkins, Eds.). Berlin: Springer-Verlag, pp. 52-93. Courtin, G.M. (1968) Evapotranspiration and energy budgets of two alpine microenvironments, Mt. Washington, N.H. Ph.D. disserta- tion, University of Illinois (Urbana), 177 pp. Courtin, G.M. and J.M. Mayo (1975) Arctic and alpine plant water rela- tions. In Physiological Adaptations to the Environment (F.J. Vern- berg, Ed.). New York: Intext Educational Publishers, pp. 201-224. Courtin, G.M. and C.L. Labine (1977) Microclimatological studies on Truelove Lowland. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University References 499 of Alberta Press, pp. 73-106. Coyne, P.I. and J.J. Kelley (1974) Variations in carbon dioxide across an arctic snowpack during spring. Journal of Geophysical Research, 79: 799-802. Coyne, P.I. and J.J. Kelley (1975) CO2 exchange over the Alaskan arctic tundra: Meteorological assessment by an aerodynamic method. Journal of Applied Ecology, 12: 587-611. Coyne, P.I. and J.J. Kelley (1978) Meteorological assessment of CO2 ex- change over an Alaskan arctic tundra. In Vegetation and Produc- tion Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 299-321. Crafts, A.S. and C.E. Crisp (1971) Phloem Transport in Plants. San Francisco: W.H. Freeman, 481 pp. Crossley, D.A., Jr. (1977) The roles of terrestrial saprophagous ar- thropods in forest soils: current status of concepts. In The Role of Arthropods in Forest Ecosystems (W.J. Mattson, Ed.). New York: Springer-Verlag, pp. 49-56. Custer, T.W. (1974) Population ecology and bioenergetics of the Lap- land longspur {Calcarius lapponicus) near Barrow, Alaska. Ph.D. dissertation. University of California (Berkeley), 189 pp. Custer, T.W. and F.A. Pitelka (1977) Demographic features of a Lapland longspur population near Barrow, Alaska. The Auk, 94: 505-525. Custer, T.W. and F.A. Pitelka (1978) Seasonal trends in summer diet of the Lapland longspur near Barrow, Alaska, U.S.A. Condor, 80: 295-301. Czapek, F. (1899) Zur Chemie der Zellmembranen bei den Laubund Lebermoosen. Flora, 86: 361-381. Dadykin, V.P. (1954) Peculiarities of plant behavior in cold soils. Akad- emii Nauk SSSR, Voprosy Botaniki, 2: 455-472 (in Russian), 473-489 (in French). Dahlman, R.C. and C.L. Kucera (1965) Root productivity and turnover in native prairie. Ecology, 46: 84-89. Dash, M.C. and J.B. Cragg (1972) Selection of microfungi by Enchytra- eidae (Oligochaeta) and other members of the soil fauna. Pedo- biologia, 12: 282-286. Davidson, J.L. and F.L. Milthorpe (1966) The effect of defoliation on the carbon balance in Dactylis glomerata. Annals of Botany N.S., 30: 185-198. Davidson, R.L. (1969) Effect of root/leaf temperature differentials on root/shoot ratios in some pasture grasses and clover. Annals of Botany N.S., 33: 561-569. Davis, D.E. and F.B. GoWey (1963) Principles in Mammalogy. New York: Reinhold, 335 pp. 500 References Dean, R., J.E. Ellis, R.W. Rice and R.E. Bement (1975) Nutrient re- moval by cattle from a short grass prairie. Journal of Applied Ecol- ogy, 12: 25-29. de Boois, H.M. (1974) Measurement of seasonal variations in the oxygen uptake of various litter layers of an oak forest. Plant and Soil, 40: 545-555. de la Roche, I.A., C.J. Andrews, M.K. Pomeroy and M. Kates (1972) Lipid changes in winter wheat seedlings {Triticum aestivum) at tem- peratures inducing cold hardiness. Canadian Journal of Botany, 50: 2401-2409. Deneke, F.J., B.H. McCown, P.I. Coyne, W. Rickard and J. Brown (1975) Biological aspects of terrestrial oil spills: USA CRREL oil re search in Alaska, 1970-1974. U.S. Army Cold Regions Research and Engineering Laboratory Research Report 346, 66 pp. Dennis, J.G. (1968) Growth of tundra vegetation in relation to arctic microenvironments at Barrow, Alaska. Ph.D. dissertation, Duke University, 289 pp. Dennis, J.G. (1977) Distribution patterns of belowground standing crop in arctic tundra at Barrow, Alaska. Arctic and Alpine Research, 9: 113-127. Dennis, J.G. and P.L. Johnson (1970) Shoot and rhizome root standing crops of tundra vegetation at Barrow, Alaska. Arctic and Alpine Research, 2: 253-266. Dennis, J.G., L.L. Tieszen and M.A. Vetter (1978) Seasonal dynamics of aboveground and belowground production of vascular plants at Barrow, Alaska. In Vegetation and Production Ecology of an Alas- kan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 113-140. deWit, C.T. (1965) Photosynthesis of Leaf Canopies. Institute of Biological and Chemical Research on Field Crops and Herbage, Wageningen, Agricultural Research Reports 663, 57 pp. deWit, C.T., R. Brouwer and F.W.T. Penning deVries (1970) The simu- lation of photosynthetic systems. In Prediction and Measurement of Photosynthetic Productivity: Proceedings of the IBP/PP Technical Meeting, Trebon, 14-21 September 1969. Wageningen: Center for Agricultural Publishing and Documentation, pp. 47-70. Dingman, S.L. (1973) The water balance in arctic and subarctic regions: Annotated bibliography and preliminary assessment. U.S. Army Cold Regions Research and Engineering Laboratory Special Report 187, 133 pp. Dommergues, Y. (1962) Contribution a I'etude de la dynamique micro- bienne des sols en zone semi-aride et en zone tropicale seche. An- nales Agronomiques, 13: 265-324, 391-468. Dorogostaiskaya, E.V. and L.M. Novichkova-Ivanova (1967) On the References 501 changes in the algal flora of tundra soils resulting from their cultiva- tion. Botanicheskii Zhurnal, 52: 461-468 (in Russian). Douce, G.K. (1976) Biomass of soil mites (Acari) in arctic coastal tun- dra. Oikos, 11: 324-330. Douce, G.K. and D.A. Crossley (1977) Acarina abundance and com- munity structures in an arctic coastal tundra. Pedobiologia, 17: 32-42. Douglas, L.A. and J.C.F. Tedrow (1959) Organic matter decomposition rates in arctic soils. Soil Science, 88: 305-312. Douglas, L.A. and J.C.F. Tedrow (1960) Tundra soils of arctic Alaska. Transactions of the 7th International Congress of Soil Science, Madison, Wisconsin, 4: 291-304. Douglas, L.A. and A. Bilgin (1975) Nutrient regimes of soils, land- scapes, lakes and streams, Prudhoe Bay, Alaska. In Ecological In- vestigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 60-70. Dowding, P. (1974) Nutrient losses from litter on IBP tundra sites. In So/7 Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: In- ternational Biological Programme Tundra Biome Steering Commit- tee, pp. 363-373. Dowding, P. and P. Widden (1974) Some relationships between fungi and their environment in tundra regions. In Soil Organisms and De- composition in Tundra: Proceedings of the Microbiology, Decom- position and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 123-149. Dowding, P., F.S. Chapin III, F.E. Wielgolaski and P. Kilfeather (1981) Nutrients in tundra ecosystems. In Tundra Ecosystems: A Compar- ative Analysis {L.C. Bliss, O.W. Heal and J.J. Moore, Eds.). Cam- bridge: Cambridge University Press, pp. 647-683 (in press). Doyle, W.T. (1964) Nonvascular Plants: Form and Function. Belmont, California: Wadsworth, 147 pp. Drew, J.V. and J.C.F. Tedrow (1962) Arctic soil classification and pat- terned ground. Arctic, 15: 109-116. Dugdale, R.C. and D.W. Toetz (1961) Sources of nitrogen for arctic Alaskan lakes. In Final Report on Investigations of the Nitrogen Cycle in Alaskan Lakes. Arctic Institute of North America, Subcon- tract ONR-253, University of Pittsburgh, Department of Biological Sciences, pp. 1-21. 502 References Dunbar, M.J. (1968) Ecological Development in Polar Regions: A Study in Evolution. Englewood Cliffs, N.J.: Prentice-Hall, 119 pp. Duncan, W.G., R.S. Loomis, W.A. Williams and R. Hanau (1967) A model for simulating photosynthesis in plant communities. Hilgar- dia, 38: 181-205. Dunican, L.K. and T. Rosswall (1974) Taxonomy and physiology of tun- dra bacteria in relation to site characteristics. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, De- composition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 79-92. Durrell, L.W. (1959) Algae in Colorado soils. American Midland Natur- alist, 61: 322-328. Edwards, N.T. and P. SoUins (1973) Continuous measurement of carbon dioxide evolution from partitioned forest floor components. Ecol- ogy, 54(2): 406-412. Ehleringer, J.R. and P.C. Miller (1975) Water relations of selected plant species in the alpine tundra, Colorado. Ecology, 56: 370-380. Ehon, C.S. (1966) The Pattern of Animal Communities. London: Methuen, 432 pp. Evans, L.T., I.E. Wardlaw and C.N. Williams (1964) Environmental control of growth. In Grass and Grasslands (C. Barnard, Ed.). Lon- don: Macmillan, pp. 102-125. Everett, K.R. (1978) Some effects of oil on the physical and chemical characteristics of wet tundra soils. Arctic, 31: 260-276. Everett, K.R. and R.J. Parkinson (1977) Soil and landform associations, Prudhoe area, Alaska. Arctic and Alpine Research, 9: 1-19. Everett, K.R., P.J. Webber, D.A. Walker, R.J. Parkinson and J. Brown (1978) A geoecological mapping scheme for Alaskan coastal tundra. In Proceedings of the Third International Conference on Perma- frost, July 10-13, 1978, Edmonton, Alberta, Canada. Vol. 1. Otta- wa: National Research Council of Canada, pp. 359-365. Feist, D.D. (1975) Population studies of lemmings on the arctic coastal tundra near Barrow, Alaska. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 135-143. Fjellberg, A. (1975) Organization and dynamics of Collembola popula- tions on Hardangervidda. In Fennoscandian Tundra Ecosystems. Part 2. Animals and Systems Analysis (F.E. Wielgolaski, Ed.). New York: Springer-Verlag, pp. 73-79. Flanagan, P.W. (1978) Microbial ecology and decomposition in arctic References 503 tundra and subarctic taiga ecosystems. In Microbial Ecology (M.W. Loutit and J.A.R. Miles, Eds.). New York: Springer-Verlag, pp. 161-168. Flanagan, P.W. and A.M. Scarborough (1974) Physiological groups of decomposer fungi on tundra plant remains. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, De- composition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 159-181. Flanagan, P.W. and A.K. Veum (1974) Relationships between respira- tion, weight loss, temperature and moisture in organic residues on tundra. In Soil Organisms and Decomposition in Tundra: Proceed- ings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stock- holm: International Biological Programme Tundra Biome Steering Committee, pp. 249-277. Flanagan, P.W. and F.L. Bunnell (1976) Decomposition models based on chmatic variables, substrate variables, microbial respiration, and production. In The Role of Terrestrial and Aquatic Organisms in Decomposition Processes: The 1 7th Symposium of the British Eco- logical Society, 15-18 April 1975 (J.M. Anderson and A. Macfad- yen, Eds.). London: Blackwell Scientific Publications, pp. 437-457. Flanagan, P.W. and K. Van Cleve (1977) Microbial biomass, respiration and nutrient cycling in a black spruce taiga ecosystem. In Soil Or- ganisms as Components of Ecosystems: International Soil Zoology Colloquium, Uppsala, Sweden, June 21-25, 1976 (V. Lohm and T. Persson, Eds.). Ecological Bulletins 25. Stockholm: Swedish Natu- ral Science Research Council, pp. 261-273. Flint, P.S. and P.L. Gersper (1974) Nitrogen nutrient levels in arctic tun- dra soils. In Soil Organisms and Decomposition in Tundra: Pro- ceedings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J, Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 375-387. Focht, D.D. (1978) Methods for analyses of denitrification in soils. In Nitrogen in the Environment. Vol. II. Soil-Plant-Nitrogen Rela- tionships (D.R. Nielsen and J.G. MacDonald, Eds.). New York: Academic Press, pp. 433-490. Fogg, G.E., W.D.P. Stewart, P. Fay and A.E. Walsby (1973) The Blue- Green Algae. London: Academic Press, 459 pp. Fonda, R.W. and L.C. Bliss (1966) Annual carbohydrate cycle of alpine 504 References plants on Mt. Washington, New Hampshire. Bulletin of the Torrey Botanical Club, 93: 268-277. Ford, J. A. (1959) Eskimo Prehistory in the Vicinity of Point Barrow, Alaska. Anthropological Papers of the American Museum of Nat- ural History 47, Pt. 1, 272 pp. Formozov, A.N. (1961) On the significance of the structure of the snow cover in the ecology and geography of mammals and birds. In Role of Snow Cover in Natural Processes: Presentation volume to CD. Rikhter on his 60th birthday. Moscow: Akademii Nauk SSSR, pp. 166-209. Fournelle, H.J. (1967) Soil and water bacteria in the Alaskan subarctic tundra. Arctic, 20: 104-113. French, D.D. (1974) Classification of IBP Tundra Biome sites based on climate and soil properties. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and In- vertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flan- agan, Eds.). Stockholm: International Biological Programme Tun- dra Biome Steering Committee, pp. 3-25. French, H.M. (1976) The Periglacial Environment. New York: Longman, 309 pp. Frissel, M.J. and J. A. Van Veen (1978) Critique of: Computer simula- tion modeling for nitrogen in irrigated croplands. In Nitrogen in the Environment. Vol. I. Nitrogen Behavior in Field Studies (D.R. Niel- sen and J.G. MacDonald, Eds.). New York: Academic Press, pp. 145-162. Fuller, W.A. (1967) Ecologie hivernale des lemmings et fluctuations de leurs populations. Terre et Vie, 2: 97-115. Fuller, W.A., A.M. Martell, R.F.C. Smith and S.W. Speller (1975) High arctic lemmings, Dicrostonyx groenlandicus. II. Demography. Canadian Journal of Zoology, 53: 867-879. Gaare, E. and T. Skogland (1975) Wild reindeer food habits and range use at Hardangervidda. In Fennoscandian Tundra Ecosystems. Part 2. Animals and Systems Analysis (F.E. Wielgolaski, Ed.). New York: Springer-Verlag, pp. 195-205. Garrett, S.D. (1963) Soil Fungi and Soil Fertility. Oxford: Pergamon Press, 165 pp. Gates, D.M. (1962) Energy Exchange in the Biosphere. New York: Harper and Row, 151 pp. Gates, D.M. (1965) Energy, plants and ecology. Ecology, 46: 1-13. Gerloff, G.C., D.D. Moore and J.T. Curtis (1964) Mineral content of native plants of Wisconsin. Wisconsin Experimental Research Sta- tion Report 14, 27 pp. Gersper, P.L. and J.L. Challinor (1975) Vehicle perturbation effects References 505 upon a tundra soil-plant system. I. Effect of morphological and physical environmental properties of the soils. So/7 Science Society of America, Proceedings, 39: 737-744. Gessaman, J. A. (1972) Bioenergetics of the snowy owl (Nyctea scan- diaca). Arctic and Alpine Research, 4: 223-238. Gimingham, C. and R. Smith (1971) Growth form and water relations of mosses in the maritime Antarctic. British Antarctic Survey Bulletin, 25: 1-21. Gjaerevoll, O. (1956) The Plant Communities of the Scandinavian Alpine Snow-beds. Det Kongelige Norske Videnskabers Selskabs Skrifter, 1956(1), 405 pp. Gollerbakh, M.M. and E.A. Shtina (1969) Soil Algae. Leningrad: Pub- Hshing House Nauka, Leningrad Branch, 228 pp. (in Russian). Goodman, G.T. and D.F. Perkins (1959) Mineral uptake and retention in cotton-grass (Eriophorum vaginatum L.). Nature, 184: 467-468. Goodwin, C.W. (1976) Process interaction and soil temperature near Point Barrow, Alaska. Ph.D. dissertation. University of Michigan, 258 pp. Grace, J. and J.R. Thompson (1973) The after-effect of wind on the photosynthesis and transpiration of Festuca arundinacea. Physio- logia Plant arum, 28: 541-547. Granhall, U. and H. Selander (1973) Nitrogen fixation in a subarctic mire. Oikos, 24: 8-15. Gray, T.R.G. and S.T. Williams (1971) Microbial productivity in soil. In Microbial Productivity: 21st Symposium of the Society of General Microbiology (D.E. Hughes and A.H. Rose, Eds.). Cambridge: Cambridge University Press, pp. 255-286. Grier, C.C. (1973) Organic matter and nitrogen distribution in some mountain heath communities of the Source Lake Basin, Washing- ton. Arctic and Alpine Research, 5: 261-267. Griffin, D.M. (1966) Soil water terminology in mycology and plant pathology. British Mycological Society, Transactions, 49: 367-368. Griffin, D.M. (1972) Ecology of Soil Fungi. Syracuse: Syracuse University Press, 193 pp. Grime, J. P. (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. A merican Naturalist, 111: 1 1 69- 1 1 94 . Grodziiiski, W. and B.A. Wunder (1975) Ecological energetics of small mammals. In Small Mammals: Their Productivity and Population Dynamics (F.B. Golley, K. Petrusewicz and L. Ryszkowski, Eds.). Cambridge: Cambridge University Press, pp. 173-204. Gunn, W.W. (1973) Bibliography of the Naval Arctic Research Lab- oratory. Arctic Institute of North America Technical Paper 24, 176 pp. 506 References Gupta, U.C. (1967) Carbohydrates. In Soil Biochemistry. Vol. 1. (A.D. McLaren and G.H. Peterson, Eds.). New York: Marcel Dekker, pp. 91-118. Guymon, G.L. (1976) Summer moisture-temperature for arctic tundra. ASCE: Journal of the Irrigation and Drainage Division, 102(1 R 4): 403-411. Haag, R.W. (1974) Nutrient limitations to plant production in two tun- dra communities. Canadian Journal of Botany, 52: 103-116. Haag, R.W. and L.C. Bliss (1974) Energy budget changes following sur- face disturbance to upland tundra. Journal of Applied Ecology, 11: 355-374. Hadley, E.B. and L.C. Bliss (1964) Energy relationships of alpine plants on Mt. Washington, New Hampshire. Ecological Monographs, 34: 331-357. Hall, H.A. and G. Pritchard (1975) The food of larvae of Tipula sacra Alexander in a series of abandoned beaver ponds (Diptera: Tipuli- dae). Journal of Animal Ecology, 44: 55-66. Hanson, W.C., F.W. Whicker and J.F. Lipscomb (1975) Lichen forage ingestion rates of free-roaming caribou estimated with fallout cesium-137. Proceedings of the First International Caribou and Reindeer Symposium, 9-11 August 1972, University of Alaska, Fair- banks, Alaska (J.R. Luick, P.C. Lent, D.R. Klein and R.G. White, Eds.). Biological Papers of the University of Alaska, Special Report 1, pp. 71-79. Hanssen, J.F. and J. Goksc^yr (1975) Biomass and production of soil and litter fungi at Scandinavian tundra sites. In Fennoscandian Tun- dra Ecosystems. Part 1. Plants and Microorganisms (F.E. Wiel- golaski, Ed.). Berlin: Springer-Verlag, pp. 239-243. Hare, F.K. and J. Ritchie (1972) The boreal biocHmates. Geographical Review, 62: 333-365. Harner, R.F. and K.T. Harper (1973) Mineral composition of grassland species of the Eastern Great Basin in relation to stand productivity. Canadian Journal of Botany, 51: 2037-2046. Harper, J.L. (1968) The regulation of numbers and mass in plant populations. In Population Biology and Evolution: Proceedings of the International Symposium, June 7-9, 1967, Syracuse, New York (R.C. Lewontin, Ed.). Syracuse, N.Y.: Syracuse University Press, pp. 138-158. Harris, W.F., R.S. Kinerson and N.T. Edwards (1977) Comparison of belowground biomass of natural deciduous forests and loblolly pine plantations. In The Belowground Ecosystem: A Synthesis of Plant- Associated Processes (J.K. Marshall, Ed.). Colorado State Universi- References 507 ty, Range Science Department Series 26, pp. 29-37. Hashizume, T., H. Morimoto, T. Mosubuchi, M. Abe and T. Hamada (1965) Some aspects of energy utilization by cows during gestation and lactation. In Energy Metabolism (of Farm Animals) (K.L. Blax- ter, Ed.). New York: Academic Press, pp. 111-120. Hatch, M.D., C.B. Osmond and R.O. Slatyer (Eds.) {\91\)Photosyn- thesis and Photorespiration: Proceedings of a Conference Held at Australian National University, Canberra, Australia, 23 November to 5 December 1970. New York: Wiley-Interscience, 565 pp. Hattori, I. (1973) Microbial Life in the Soil: An Introduction. New York: Marcel Dekker, 427 pp. Haugen, R.K. and J. Brown (1980) Coastal-inland distributions of sum- mer air temperature and precipitation in northern Alaska. Arctic and Alpine Research, 12:403-412. Hausenbuiller, R.L. (1972) So/7 Science Principles and Practices. Du- buque, Iowa: Wm.C. Brown, 504 pp. Hayes, A.J. and P. Rheinberg (1975) Microfungal populations of the Abisko area, northern Sweden. In Fennoscandian Tundra Ecosys- tems. Part 1. Plants and Microorganisms (F.E. Wielgolaski, Ed.). New York: Springer-Verlag, pp. 244-250. Heal, O.W. and D.D. French (1974) Decomposition of organic matter in tundra. In So/7 Organisms and Decomposition in Tundra: Proceed- ings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stock- holm: International Biological Programme Tundra Biome Steering Committee, pp. 279-309. Heal, O.W., H.E. Jones and J.B. Whittaker (1975) Moor House, U.K. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Productivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 295-320. Heal, O.W, and S.F. MacLean, Jr. (1975) Comparative productivity in ecosystems — secondary productivity. In Unifying Concepts in Ecol- ogy: Report of the Plenary Sessions of the First International Con- gress of Ecology, The Hague, The Netherlands, September 2-14, 1974 (W.H. van Dobben and P.H. Lowe-McConnell, Eds.). The Hague: Dr. W. Junk, pp. 89-108. Heal, O.W., P.M. Latter and G. Howson (1978) A study of the rates of decomposition of organic matter. In Production Ecology of British Moors and Montane Grasslands (O.'WHesd and D.F. Perkins, Eds.). Berlin: Springer-Verlag, pp. 136-159. Heal, O.W. and D.F. Perkins (Eds.) (1978) Production Ecology of Brit- 508 References ish Moors and Montane Grasslands. Berlin: Springer-Verlag, 426 pp. Heal, O.W., P.W. Flanagan, D.D. French and S.F. MacLean, Jr. (1981) Decomposition and accumulation of organic matter in tundra. In Tundra Ecosystems: A Comparative Analysis (L.C. Bliss, O.W, Heal and J.J. Moore, Eds.). Cambridge: Cambridge University Press, pp. 587-633 (in press). Healey, I.N. and A. Russell-Smith (1970) The extraction of fly larvae from woodland soils. Soil Biology and Biochemistry, 2: 119-129. Healey, I.N. and A. Russell-Smith (1971) Abundance and feeding prefer- ences of fly larvae in two woodland soils. IV Colloquium Pedobio- logiae, Dijon 1970. Annales de Zoologie-Ecologie Animale, Inst. Nat. Recherche Agr., hors serie, 177-191. Hemming, J.E. (1971) The distribution and movement patterns of cari- bou in Alaska. Alaska Department of Fish and Game, Game Tech- nical Bulletin 1, 60 pp. Hemmingsen, A. and B. Jensen (1957) The occurrence of Tipula (Vesti- plex) arctica Curtis in Greenland and its decreasing body length with increasing latitude. Meddelelser om GrSnland, 159(1), 20 pp. Henin, S., G. Monnier and L. Turc (1959) C. R. Seanc. Soc. Biol., 248: 138. (Original not seen; cited in Satchell 1974.) Herman, R.P. (1977) Root contribution to 'total soil respiration' in a tallgrass prairie. American Midland Naturalist, 98: 227-232. Hernandez, H. (1973) Natural plant recolonization of surficial distur- bances, Tuktoyaktuk Peninsula Region, Northwest Territories. Canadian Journal of Botany, 51: 2177-2196. Hicklenton, P.R. and W.C. Oechel (1976) Physiological aspects of the ecology of Dicranum fuscescens in the subarctic. I. Acclimation and acclimation potential of CO2 exchange in relation to habitat, light, and temperature. Canadian Journal of Botany, 54: 1104-1119. Hill, D.E. and J.C.F. Tedrow (1961) Weathering and soil formation in the arctic environment. American Journal of Science, 259: 84-101. Hoagland, D.R. and T.C. Broyer (1936) General nature of the process of salt accumulation by roots with description of experimental meth- ods. Plant Physiology, 11: 471-507. Hobbie, J.E. (Ed.) (1980) Limnology of Tundra Ponds, Barrow, Alaska. Stroudsburg, Pa.: Dowden, Hutchinson and Ross, 514 pp. Hocking, B. and CD. Sharplin (1965) Flower basking by arctic insects. Nature, 206: 215. Hok, J.R. (1969) A reconnaissance of tractor trails and related phenom- ena on the North Slope of Alaska. U.S. Department of Interior, Bureau of Land Management Publication, 66 pp. HoUeman, D.F., R.G. White and J. Luick (1974) New isotope methods for measuring milk intake and yield. Journal of Dairy Science, 58: References 509 1814-1821. Holling, C.S. (1959) The components of predation as revealed by a study of small-mammal predation of the European pine sawfly. The Can- adian Entomologist, 91: 293-320. Holmes, R.T. (1966) Feeding ecology of the red-backed sandpiper {Cal- idris alpina) in arctic Alaska. Ecology, 47: 32-45. Holmes, R.T. and F.A. Pitelka (1968) Food overlap among coexisting sandpipers on northern Alaskan tundra. Systematic Zoology, 17: 305-318. Holmgren, B., C. Benson and G. Weller (1975) A study of the breakup on the Arctic Slope of Alaska by ground, aircraft and satellite obser- vations. In Climate of the Arctic: Twenty-fourth Alaska Science Conference Proceedings, Fairbanks, Alaska, August 1973 (G. Wel- ler and S.A. Bowling, Eds.). Fairbanks: Geophysical Institute, Uni- versity of Alaska, pp. 358-366. Holzworth, G.C. (1974) Summaries of the lower few kilometers of rawinsonde and radiosonde observations in the United States. Pre- print, CHmatology Conference and Workshop of the American Me- teorological Society, 8-11 October 1974, Asheville, North Carolina. (Data available from National Climatic Center, Asheville, N.C., Job No. 13105, "Inversion Study.") Home, A.J. (1972) The ecology of nitrogen fixation on Signy Island, South Orkney Islands. British Antarctic Survey Bulletin, 27: 1-18. Hosokawa, T., N. Odani and H. Tagawa (1964) Causality of the distri- bution of corticolous species in forests with special reference to the physio-ecological approach. The Bryologist, 67: 396-411. Houston, W.W.K. (I97I) Carabidae (Col.) from two areas of the North Pennines. Entomologists' Monthly Magazine, 107: 1-4. Hsiao, T. (1973) Plant responses to water stress. Annual Review of Plant Physiology, 24: 519-570. Hudson, H. (1968) The ecology of fungi on plant remains above the soil. The New Phytologist, 67: 837-874. Huffaker, R.C. and L.W. Peterson (1974) Protein turnover in plants and possible means of its regulation. Annual Review of Plant Physi- ology, 25: 363-392. Hulteh, E. (1968) Flora of Alaska and Neighboring Territories: A Manu- al of the Vascular Plants. Stanford, Calif.: Stanford University Press, 1008 pp. Hutchinson, G.E. (1959) Homage to Santa Rosalia or why there are so many kinds of animals. American Naturalist, 93: 145-159. Hutchinson, T.C., J. A. Hellebust and M. Telford (1976) Oil spill effects on vegetation and soil microfauna at Norman Wells and Tuktoyak- tuk, N.W.T. Department of Indian Affairs and Northern Develop- ment, Ottawa, Arctic Land Use Research, 74-75-83, 143 pp. 510 References Hyder, D.N. (1972) Defoliation in relation to vegetative growth. In The Biology and Utilization of Grasses (V.B. Youngner and CM. McKell, Eds.). New York: Academic Press, pp. 304-317. lizuka, H., H. Okazaki and N. Seto (1969) A new sulphate-reducing bac- terium isolated from Antarctica. Journal of General and Applied Microbiology, 15: 11-18. Ingraham, J.L. (1958) Growth of psychrophilic bacteria. Journal of Bac- teriology, 76: 75-80. Ivanov, M.W. (1955) Method of determination of bacterial production in the reservoirs. Mikrobiologiia, 24: 79-89 (in Russian). Ivarson, K.C. (1965) The microbiology of some permafrost soils in the Mackenzie Valley, N.W.T. Arctic, 18: 256-260. Jarvis, P.G. and M.S. Jarvis (1963) The water relations of tree seedlings. IV. Some aspects of the tissue water relations and drought resis- tance. Physiologia Plantarum, 16: 501-516. Jenkin, J.F. (1975) Macquarie Island subantarctic. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tun- dra Biome V International Meeting on Biological Productivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 375-397. Jenkins, T.F., L.A. Johnson, CM. Collins and T.T. McFadden (1978) The physical, chemical and biological effects of crude oil spills on black spruce forests, interior Alaska. Arctic, 31: 305-323. Jenkinson, D.S. and A. Ayanaba (1977) Decomposition of carbon-14 labeled plant material under tropical conditions. Soil Science Soci- ety of America, Proceedings, 41: 912-915. Jenkinson, D.S. and J.H. Rayner (1977) The turnover of soil organic matter in some of the Rothamsted classical experiments. Soil Sci- ence, 123: 298-305. Johansson, L.-G. (1974) The distribution and fate of ""C photoassimil- ated by plants on a subarctic mire at Stordalen. In International Bio- logical Programme Progress Report 1973 (J.G.K. Flower-Ellis, Ed.). Swedish Tundra Biome Project, Technical Report 16, pp. 165-172. John, R.P. (1942) An ecological and taxonomic study of the algae of British soils. I, The distribution of the surface-growing algae. An- nals of Botany, N.S., 6: 323-349. Johnson, A.W. and J.G. Packer (1965) Polyploidy and environment in arctic Alaska. Science, 148: 237-239. Johnson, A.W., L.A. Viereck, R.E. Johnson and H. Melchior (1966) Vegetation and flora. In Environment of the Cape Thompson Re- gion, Alaska (N.J. Wilimovsky and J.N. Wolfe, Eds.). Oak Ridge, Tenn.: U.S. Atomic Energy Commission, pp. 277-354. Johnson, D.A. and M.M. Caldwell (1974) Field measurements of photo- References 511 synthesis and leaf growth rate on three alpine plant species. Arctic and Alpine Research, 6: 245-251. Johnson, D.A. and L.L. Tieszen (1976) Aboveground biomass alloca- tion, leaf growth, and photosynthesis patterns in tundra plant forms in arctic Alaska. Oecologia (Berl.), 24: 159-173. Johnson, F.L. and P.G. Risser (1974) Biomass, annual net primary pro- duction and dynamics of six mineral elements in a post oak-black- jack oak forest. Ecology, 55: 1246-1258. Johnson, L. and K. Van Cleve (1976) Revegetation in arctic and subarc- tic North America — A literature review. U.S. Army Cold Regions Research and Engineering Laboratory CRREL Report 76-15, 32 pp. Johnson, P.C. (1963) Nervous and chemical control of respiration. In Physiology (E.E. Selkurt, Ed.). 2nd edition. Boston: Little, Brown, pp. 471-487. Johnson, P.L. and F.B. Kistner (1967) Breakup of ice, Meade River, Alaska. U.S. Army Cold Regions Research and Engineering Labor- atory Special Report 118, 12 pp. Johnson, P.L. and L.L. Tieszen (1973) Vegetative research in arctic Alaska. In Alaska Arctic Tundra: Proceedings of the 25th Anniver- sary Celebration of the Naval Arctic Research Laboratory (M.E. Britton, Ed.). Arctic Institute of North America Technical Paper 25, pp. 169-198. Johnston, A. and L.M. Bezeau (1962) Chemical composition of range forage plants of the Festuca scabrella association. Canadian Journal of Plant Science, 42 : 1 05 - 1 1 5 . Jordan, W.R. and J.T. Ritchie (1971) Influence of soil water stress on evaporation, root absorption, and internal water status of cotton. Plant Physiology, 48: 783-788. Jordon, C.F. and J.R. Kline (1972) Mineral cycling: Some basic concepts and their application in a tropical rain forest. Annual Review of Ecology and Systematics, 3: 33-50. Jordon, D.C., P.S. McNichol and M.R. Marshall (1978) Biological nitrogen fixation: The terrestrial environment of a high arctic eco- system (Truelove Lowland, Devon Island, N.W.T.). Canadian Journal of Microbiology, 24: 643-649. Kaila, A. (1949) Biological absorption of phosphorus. Soil Science, 68: 279-289. Kalff, J. (1965) Primary production rates and the effects of some en- vironmental factors on algal photosynthesis in small arctic tundra ponds. Ph.D. dissertation, Indiana University, 122 pp. KaUio, P., S. Suhonen and H. Kallio (1972) The ecology of nitrogen fixa- tion in Nephroma arcticum and Solorina crocea. Reports from the Kevo Subarctic Research Station, 9: 7-14. Kallio, P. and S. Heinonen (1973) Ecology of Rhacomitrium lanugino- 512 References sum (Hedw.) Brid. Reports from the Kevo Subarctic Research Sta- tion, 10: 43-54. Kallio, P. and L. Karenlampi (1975) Photosynthesis in mosses and lichens. In Photosynthesis and Productivity in Different Environ- ments: IBP Synthesis Meeting on the Functioning of Photosynthetic Systems in Different Environments, Aberystwyth, 1973 (J. P. Coop- er, Ed.). Cambridge: Cambridge University Press, pp. 393-423. Kallio, S. (1973) The ecology of nitrogen fixation in Stereocaulon pas- chale. Reports from the Kevo Subarctic Research Station, 10: 34-42. Kane, D.L. and R.F. Carlson (1973) Hydrology of the central arctic river basins of Alaska. University of Alaska, Institute of Water Re- sources Report IWR-41, 57 pp. Kedrowski, R.A. (1976) Plant water relations in the arctic tundra near Meade River, Alaska. M.S. thesis, San Diego State University, 83 pp. Kedrowski, R.A. and F.S. Chapin, III (1978) Lipid properties of Carex aquatilis from hot springs and permafrost-dominated sites in Alaska: Implications for nutrient requirements. Physiologia Plant- arum, 44: 231-237. Kelley, J.J., Jr., D.T. Bailey and B.J. Lieske (1964) Radiative energy exchange over arctic land and sea: Data 1962. University of Wash- ington, Department of Atmospheric Science, Scientific Report, 205 pp. Kelley, J. J., Jr., D.F. Weaver and B.P. Smith (1968) The variation of carbon dioxide under the snow in the Arctic. Ecology, 49: 358-361. Kelley, J. J., Jr. and D.F. Weaver (1969) Physical processes at the surface of the arctic tundra. Arctic, 11: 425-437. Kelsall, J. P. (1968) The Migrating Barren-ground Caribou of Canada. Canada Wildlife Service Monograph 3. Ottawa: Queen's Printer, 340 pp. Kenyon, C.N., R. Rippka and R.Y. Stanier (1972) Fatty acid composi- tion and physiological properties of some filamentous blue-green algae. Archiv fur Mikrobiologie, 83: 216-236. Kessel, B. and T.J. Cade (1958) Birds of the Colville River, Northern Alaska. Biological Papers of the University of Alaska 2, 83 pp. Ketellapper, H.J. (1963) Stomatal physiology. Annual Review of Plant Physiology, 14: 249-270. Khalid, R.A., W.H. Patrick, Jr. and R.D. Delaune (1977) Phosphorus sorption characteristics of flooded soils. Soil Science Society of America, Proceedings, 41: 305-310. Kilday, G.D. (1974) Mean monthly and annual precipitation, Alaska. NOAA Technical Memorandum NWS AR-10, 13 maps. Kitchen, J.F., R.V. O'Neill, D. Webb, G.W. Gallepp, S.M. Bartell, J.F. Koonce and B.S. Ausmus (1979) Consumer regulation of nutrient cycling. BioScience, 29: 28-34. References 513 Kleiber, M. (1961) The Fire of Life: An Introduction to Animal Ener- getics. New York: John Wiley, 454 pp. Klein, D.R. (1970) Tundra ranges north of the boreal forest. Journal of Range Management, 23: 8-14. Klein, D.R. and R.G. White (Eds.) (1978) Parameters of Caribou Popu- lation Ecology in Alaska; Proceedings of a Symposium and Work- shop. Biological Papers of the University of Alaska, Special Report 3, 49 pp. Kleinendorst, A. and R. Brouwer (1970) The effect of temperature of the root medium and of the growing points of the shoot on growth, water content, and sugar content of maize leaves. Netherlands Jour- nal of Science, 18: 140-148. Klikoff, L.G. (1966) Temperature dependence of the oxidative rates of mitochondria in Danthonia intermedia, Penstemon davidsonii and Sitanion hystrix. Nature, 212: 529-530. Kobayasi, Y., H. Hiratsuku, R.P. Korf, K. Tubaki, K. Aoshima, M. Soneda and J. Sugiyama (1967) Mycological studies of the Alaskan Arctic. Annual Report of the Institute of Fermentation, Osaka, 3: 1-138. Kobayasi, Y., H. Hiratsuku, Y. Otani, K. Tubaki, S. Udagawa and M. Soneda (1969) The second report on the mycological flora of the Alaskan Arctic. Bulletin of the Natural Science Museum, Tokyo, 12: 311-316. Koller, D. and J. Kigel (1972) The growth of leaves and tillers in Oryzop- sis miliacea. In The Biology and Utilization of Grasses (V.B. Youngner and C.M.^cKell, Eds.). New York: Academic Press, pp. 115-134. Koma'rkova', V. and P.J. Webber (1978) An alpine vegetation map of Niwot Ridge, Colorado. Arctic and Alpine Research, 10: 1-29. Koppen, W. (1936) Das geographische System der Klimate. In Handbuch der Klimatologie (W. Koppen and R. Seiger, Eds.). Vol. I.C. Berlin: Borntraeger, 44 pp. Koranda, J. J., B. Clegg and M. Stuart (1978) Radio-tracer measurement of transpiration in tundra vegetation, Barrow, Alaska. In Vegeta- tion and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 359-369. Kramer, P.J. (1969) Plant and Soil Water Relationships: A Modern Syn- thesis. New York: McGraw Hill, 482 pp. Krebs, C.J. (1964) The lemming cycle at Baker Lake, Northwest Territor- ies, during 1959-62. Arctic Institute of North America Technical Paper 15, 104 pp. Krebs, C.J. and I.McT. Cowan (1962) Growth studies of reindeer fawns. Canadian Journal of Zoology, 40: 863-869. Krebs, C.J., M.S. Gaines, B.L. Keller, J.H. Myers and R.H. Tamarin 514 References (1973) Population cycles in small rodents. Science, 179: 35-41. Krebs, C.J. and J.H. Myers (1974) Population cycles in small mammals. Advances in Ecological Research, 8: 268-399. Kressin, G. (1935) Beitrage zur vergleichenden Protoplasmatik der Moss- zelle. Dissertation, Greifswald. Kriss, A.E. (1947) Microorganisms of the tundra and desert soils of the Arctic. Mikrobiologiia, 16: 437-448. Krogg, J. (1955) Notes on temperature measurements indicative of spe- cial organization in arctic and subarctic plants for utilization of radiated heat from the sun. Physiologia Plantarum, 8: 836-839. Kryuchkov, V.V. (1976) Sensitive Subarctic. Moscow: Nauka, 137 pp. (Translation published as U.S. Army Cold Regions Research and Engineering Laboratory Draft Translation 556, 130 pp., 1976.) Kuiper, P.J.C. (1964) Water uptake of higher plants as affected by root temperature. Mededelingen Landbouwhogeschool Wageningen, 63: 1-11. Kuramoto, R.T. and L.C. Bliss (1970) Ecology of subalpine meadows in the Olympic Mountains, Washington. Ecological Monographs, 40: 317-347. Lachenbruch, A.H. (1962) Mechanics of thermal contraction cracks and ice-wedge polygons in permafrost. Geological Society of America Special Paper 70, 69 pp. Lachenbruch, A.H. and B.V. Marshall (1969) Heat flow in the Arctic. Arctic, 22: 300-311. Lachenbruch, A.H. and B.V. Marshall (1977) Sub-sea temperatures and a simple tentative model for offshore permafrost at Prudhoe Bay, Alaska. U.S. Geological Survey Open-file Report 77-395, 34 pp. Langer, R.H.M. (1966) Mineral nutrition of grasses and cereals. In The Growth of Cereals and Grasses (F.L. Milthorpe and J.D. Ivins, Eds.). London: Butterworths, pp. 213-226. Larcher, W., L. Schmidt, G. Grabherr and A. Cernusca (1973) Plant biomass and production of alpine shrub heaths at Mt. Patscher- kofel, Austria. In Primary Production and Production Processes: Proceedings of the Conference, Dublin, Ireland, April 1973 (L.C. BUss and F.E. Wielgolaski, Eds.). Stockholm: International Biolog- ical Programme Tundra Biome Steering Committee, pp. 65-73. Latter, P.M., J.B. Cragg and O.W. Heal (1967) Comparative studies on the microbiology of four moorland soils in the northern Pennines. Journal of Ecology, 55: 445-464, Latter, P.M. and O.W. Heal (1971) A preliminary study of the growth of fungi and bacteria from temperate and Antarctic soils in relation to temperature. So/7 Biology and Biochemistry, 3: 365-379. Laursen, G. (1975) Higher fungi in soils of coastal arctic tundra plant communities. Ph.D. dissertation, Virginia Polytechnic Institute and References 515 State University, 370 pp. Laursen, G.A., O.K. Miller, Jr. and H.E. Bigelow (1976) A new Clito- cybe from the Alaskan Arctic. Canadian Journal of Botany, 54: 976-980. Laursen, G.A. and O.K. Miller, Jr. (1977) The distribution of fungal hyphae in arctic soil on the International Biological Programme Tundra Biome site, Barrow, Alaska. Arctic and Alpine Research, 9: 149-156. Lawrence, B.A., M.C. Lewis and P.C. Miller (1978) A simulation model of population processes of arctic tundra graminoids. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Ties- zen, Ed.). New York: Springer-Verlag, pp. 599-619. Lawson, D.E., J.Brown, K.R.Everett, A.W.Johnson, V.Komarkova, B.M. Murray, D.F. Murray and P.J. Webber (1978) Tundra dis- turbances and recovery following the 1949 exploratory drilling, Fish Creek, northern Alaska. U.S. Army Cold Regions Research and En- gineering Laboratory CRREL Report 78-28, 81 pp. LeDrew, E.F. and G. Weller (1978) A comparison of the radiation and energy balance during the growing season for an arctic and alpine tundra. Arctic and Alpine Research, 10: 665-678. Lemon, E.R. (1963) Energy and water balance of plant communities. In Environmental Control of Plant Growth: Proceedings of a Sympo- sium held at Canberra, Australia, August 1962 (L.T, Evans, Ed.). New York: Academic Press, pp. 55-78. Lent, P.C. (1966) The caribou of northwestern Alaska. In Environment of the Cape Thompson Region, Alaska (N.J. Wilimovsky and J.N. Wolfe, Eds.). Oak Ridge, Tenn.: U.S. Atomic Energy Commission, pp.481-517. Leonard, E.R. (1962) Interrelations of vegetative and reproductive growth, with special reference to indeterminate plants. Botanical Review, 28: 353-410. Letrouit-Galinou, M.-A. (1973) Sexual reproduction. In The Lichens (V, Ahmadjian and M.E. Hale, Eds.). New York: Academic Press, pp. 59-90. Levin, E.I. (1899) Les microbes dans les regions arctiques. Annales de rinstitute Pasteur, 13: 558-567. Lewellen, R.I. (1972) Studies on the Fluvial Environment, Arctic Coastal Plain Province, Northern Alaska. Published by the author, P.O. Box 1068, Littleton, Colorado 80120, 282 pp. Lewellen, R.I. (1977) Subsea permafrost research techniques. Geo- sciences and Man, XVIII: 29-34. Lewis, J.K., J.L. Dodd, H.L. Hutcheson and C.L. Hanson (1971) Abiotic and herbage dynamic studies on the cottonwood site, 1970. Grasslands Technical Report 111, Fort Collins, Colorado. 516 References Lewis Smith, R.I. and D.W.H. Walton (1975) South Georgia, subant- arctic. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Bio- logical Productivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stock- holm: Swedish Natural Science Research Council, pp. 399-423. Lieske, B.J. and L.A. Stroschein (1968) Radiative regime over arctic tun- dra. University of Washington Department of Atmospheric Sci- ences, Scientific Report on Contract ONR-477(24), 31 pp. Likens, G.E. and F.H. Bormann (1972) Nutrient cycling in ecosystems. In Ecosystem Structure and Function: Proceedings of the Annual Biology Colloquium, 31 (J.H. Wiens, Ed.). Corvallis: Oregon State University Press, pp. 25-61 . Likens, G.E., F.H. Bormann, R.S. Pierce, J.S. Eaton and N.M. Johnson (1977) Biogeochemistry of a Forested Ecosystetn. New York: Springer-Verlag, 146 pp. Linell, K.A. (1973) Long-term effects of vegetative cover on per- mafrost stability in an area of discontinuous permafrost. In Perma- frost: North American Contribution to the Second International Conference. Washington, D.C.: National Academy of Sciences, pp. 688-693. Linkins, A.E. and R.K. Antibus (1978) Ectomycorrhizal fungi of Salix rotundifolia Trautv. II. Impact of surface applied Prudhoe Bay crude oil on mycorrhizal root respiration and cold acclimation. Arc- tic, 31: 381-391. Lord, N.W., J. P. Pandolfo and M.A. Atwater (1972) Simulation of meteorological variations over arctic coastal tundra under various physical interface conditions. Arctic and Alpine Research, 4: 189- 209. Luxton, M. (1972) Studies on the oribatid mites of a Danish beech wood soil. I. Nutritional biology. Pedobiologia, 12: 434-463. McCown, B.H. (1973) The influence of soil temperature on plant growth and survival in Alaska. Proceedings of the Symposium on the Im- pact of Oil Resource Development on Northern Plant Communities: Presented at the 23rd AAAS Alaska Science Conference, Fairbanks, Alaska, 17 August 1972 (B.H. McCown and D.R. Simpson, Coor- dinators). University of Alaska, Institute of Arctic Biology, Occa- sional Publications on Northern Life 1, pp. 12-33. McCown, B.H. (1975) Physiological responses of root systems to stress conditions. In Physiological Adaptations to the Environment (F.J. Vernberg, Ed.). New York: Intext Educational Publishers, 225-237. McCown, B.H. (1978) The interaction of organic nutrients, soil nitro- gen, and soil temperature on plant growth and survival in the arctic environment. In Vegetation and Production Ecology of an Alaskan References 517 Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 435-456. McCown, D.D. and F.J. Deneke (1973) Plant germination and seedling growth as affected by the presence of crude petroleum. Proceedings of the Symposium on the Impact of Oil Resource Development on Northern Plant Communities: Presented at the 23 rd A A AS Alaska Science Conference, Fairbanks, Alaska, 17 August 1972 (B.H. Mc- Cown and D.R. Simpson, Coordinators). University of Alaska, In- stitute of Arctic Biology, Occasional Publications on Northern Life 1, pp. 44-51. McCree, K.J. (1974) Equations for the rate of dark respiration of white clover and grain sorghum as functions of dry weight, photosynthetic rate, and temperature. Crop Science, 14: 509-514. McEwan, E.H. (1970) Energy metabolism of barren-ground caribou (Rangifer tarandus). Canadian Journal of Zoology, 48: 391-392. McEwan, E.H. and P.E. Whitehead (1970) Seasonal changes in the energy and nitrogen intake in reindeer and caribou. Canadian Jour- nal of Zoology, 48: 905-913. McEwan, E.H. and P.E. Whitehead (1972) Reproduction in female reindeer and caribou. Canadian Journal of Zoology, 50: 43-46. McKendrick, J.D., C.E. Owensby and R.M. Hyde (1975) Big bluestem and Indiangrass vegetative reproduction and annual reserve carbo- hydrate and nitrogen cycles. Agro-Ecosystems, 2: 75-93. McKendrick, J.D., V.J. Ott and G.A. Mitchell (1978) Effects of nitrogen and phosphorus fertilization on the carbohydrate and nutrient levels in Dupontia fisheri and Arctagrostis latifolia. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 509-537. McKendrick, J.D., G.O. Batzli, K.R. Everett and J.C. Swanson (1980) Some effects of mammalian herbivores and fertilization on tundra soils and vegetation. Arctic and Alpine Research, 12: 565-578. McNaughton, S.J., R.S. Campbell, R.A. Freyer, J.E. Mylroie and K.D. Rodland (1974) Photosynthetic properties and root chiUing re- sponses of altitudinal ecotypes of Typha latifolia L. Ecology, 55: 168-172. Macfadyen, A. (1970) Soil metabolism in relation to ecosystem energy flow and to primary and secondary production. In Methods of Study in Soil Ecology (J. Phillipson, Ed.). Paris: UNESCO, pp. 167-172. MacLean, S.F., Jr. (1969) Ecological determinants of species diversity of arctic sandpipers near Barrow, Alaska. Ph.D. dissertation, Univer- sity of California (Berkeley), 188 pp. MacLean, S.F., Jr. (1973) Life cycle and growth energetics of the arctic cranefly Pedicia hannai antenatta. Oikos, 24: 436-443. 518 References MacLean, S.F., Jr. (1974a) Primary production, decomposition, and the activity of soil invertebrates in tundra ecosystems: A hypothesis. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: In- ternational Biological Programme Tundra Biome Steering Commit- tee, pp. 197-206. MacLean, S.F., Jr. (1974b) Lemming bones as a source of calcium for arctic sandpipers {Calidris spp.). Ibis, 116: 522-557. MacLean, S.F., Jr. (1975a) Ecological adaptations of tundra inverte- brates. In Physiological Adaptation to the Environment (F.J. Vern- berg, Ed.). New York: Intext Educational Publishers, pp. 269-300. MacLean, S.F., Jr. (1975b) Ecology of tundra invertebrates at Prudhoe Bay, Alaska. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 114-123. MacLean, S.F., Jr. and F.A. Pitelka (1971) Seasonal patterns of abund- ance of tundra arthropods near Barrow. Arctic, 24: 19-40. MacLean, S.F., Jr., B.M. Fitzgerald and F.A. Pitelka (1974) Population cycles in arctic lemmings: Winter reproduction and predation by weasels. Arctic and Alpine Research, 6: 1-12. MacLean, S.F., Jr., G.K. Douce, E.A. Morgan and M.A. Skeel (1977) Community organization in the soil invertebrates of Alaskan arctic tundra. In Soil Organisms as Components of Ecosystems: Interna- tional Soil Zoology Colloquium, Uppsala, Sweden, June 21-25, 1976 (U. Lohm and T. Persson, Eds.). Ecological Bulletins 25. Stock- holm: Swedish Natural Science Research Council, pp. 90-101. MacLean, S.F., Jr., V. Behan and A. Fjellberg (1978) Soil Acari and Collembola from Chaun Bay, northern Chukotka. Arctic and Al- pine Research, 10: 559-568. MacLean, S.F., Jr. and I.D. Hodkinson (1980) The distribution of psyl- lids (Homoptera: Psylloidea) in arctic and subarctic Alaska. Arctic and Alpine Research, 12: 369-376. MacPherson, A.H. (1969) Reproduction of Canadian arctic fox popula- tions. Canadian Wildlife Service Report 8, 52 pp. Magdefrau, K. (1937) Der Wasserhaushalt der Moose. Annales Bryo- logici, 10: 141-150. Maher, W.J. (1967) Predation by weasels on winter population of lem- mings, Banks Island, Northwest Territories. Canadian Field Natur- alist, 81: 248-250. Maher, W.J. (1970) The pomarine jaeger as a brown lemming predator in northern Alaska. Wilson Bulletin, 82: 130-157. Maher, W.J. (1974) Ecology of pomarine, parasitic and long-tailed jae- References 519 gers in northern Alaska. Pacific Coast Avifauna, 37: 148 pp. Mani, M.S. (1962) Introduction to High Altitude Entomology. London: Methuen, 302 pp. Mani, M.S. (1968) Ecology and Biogeography of High Altitude Insects. The Hague: Dr. W. Junk, 527 pp. Marks, P.L. and F.H. Bormann (1972) Revegetation following forest cutting: mechanisms for return to steady-state nutrient cycling. Science, 176: 914-915. Marr, A.G., E.H. Nelson and D.J. Clark (1963) The maintenance re- quirement of Escherichia coli. Annals of the New York Academy of Sciences, 102: 536-548. Marshall, C. and G.R. Sagar (1968) The distribution of assimilates in Lolium multiflorum Lam. following differential defoliation. Annals of Botany, N.S., 32: 715-719. Maruyama, K. (1967) Blue-green algae in the Alaskan Arctic. In Phycological Report of the Japanese Microbiological Expedition to the Alaskan Arctic (Y. Kabayashi, Compiler). Bulletin of the Na- tional Science Museum, Tokyo, 10: 221-239. Mather, J.R. and C.W. Thornthwaite (1958) Microclimate Investiga- tions at Point Barrow, Alaska, 1957-1958. Drexel Institute of Tech- nology, Publications in Climatology, XI(2), 239 pp. Mattheis, P.J., L.L. Tieszen and M.C. Lewis (1976) Responses of Dupontia fisheri to lemming grazing in an Alaskan arctic tundra. Annals of Botany, N.S., 40: 179-197. May, D.C.E. (1973) Model for predicting composition and production of alpine tundra vegetation from Niwot Ridge, Colorado. M.A. the- sis. University of Colorado (Boulder), 99 pp. Maykut, G.A. and P.E. Church (1973) Radiation climate of Barrow, Alaska, 1962-66. Journal of Applied Meteorology, 12: 620-628. Mayo, J.M., D.G. Despain and E.M. van Zinderen Bakker, Jr. (1973) CO2 assimilation by Dryas integrifolia on Devon Island, Northwest Territories. Canadian Journal of Botany, 5: 581-588. Mayo, J.M., A. P. Hartgerink, D.G. Despain, R.G. Thompson, E.M. van Zinderen Bakker, Jr. and S.D. Nelson (1977) Gas exchange studies of Carex and Dryas, Truelove Lowland. In Truelove Low- land, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 265-280. Millar, J.S. (1977) Adaptive features of mammalian reproduction. Evo- lution, 31: 370-386. Miller, J.F. (1963) Probable maximum precipitation and rainfall fre- quency data for Alaska for areas to 400 square miles, durations to 24 hours, and return periods from 1 to 100 years. U.S. Weather Bu- reau Technical Paper No. 47, 69 pp. Miller, J.H., J.E. Giddens and A. A. Foster (1957) A survey of the fungi 520 References of forest and cultivated soils of Georgia. Mycologia, 49: 779-808. Miller, M.C., R.T. Prentki and R.J. Barsdate (1980) Physics. In Lim- nology of Tundra Ponds, Barrow, Alaska (J.E. Hobbie, Ed.). Stroudsburg, Pa,: Dowden, Hutchinson and Ross, pp. 51-75. Miller, O.K., Jr., G.A. Laursen and B.M. Murray (1973) Arctic and alpine agarics from Alaska and Canada. Canadian Journal of Botany, 51: 43-49. Miller, O.K., Jr. and D.E. Farr (1975) Index of the common fungi of North America (synonymy and common names). Bibliotheca My- cologia, 44: p. 206. Miller, O.K., Jr. and G.A. Laursen (1978) Ecto- and endomycorrhizae of arctic plants at Barrow, Alaska. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 229-237, Miller, P.C. (1978) Problems of synthesis in mineral cycling studies: The tundra as an example. In Environmental Chemistry and Cycling Processes: Proceedings of Symposium, Augusta, Georgia, April 28-May I, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. Department of Energy, CONF-760429, pp. 59-71. Miller, P.C, (1979) Quantitative plant ecology. In Analysis of Ecosys- tems (D. Horn, G.R, Stairs and R.D. Mitchell, Eds.), Columbus: Ohio State University Press, pp, 179-232. Miller, P.C. and L.L, Tieszen (1972) A preliminary model of processes affecting primary production in the arctic tundra. Arctic and Alpine Research, 4: 1-18. Miller, P.C, B.D. Collier and F.L. Bunnell (1975) Development of ecosystem modeling in the Tundra Biome. In Systems Analysis in Ecology, 3: 95-115. Miller, P.C, W.A. Stoner and L.L. Tieszen (1976) A model of stand photosynthesis for the wet meadow tundra, Barrow, Alaska. Ecol- ogy, 57: 411-430. Miller, P.C, W,C, Oechel, W,A. Stoner and B, Sveinbjbrnsson (1978a) Simulation of CO2 uptake and water relations of four arctic bryo- phytes at Point Barrow, Alaska. Photosynthetica, 12: 7-20. Miller, P.C, W.A. Stoner and J.R. Ehleringer (1978b) Some aspects of water relations of arctic and alpine plants. In Vegetation and Pro- duction Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 343-357. Miller, P.C, W.A. Stoner, L.L. Tieszen, M. Allessio, B. McCown, F.S, Chapin and G. Shaver (1978c) A model of carbohydrate, nitrogen, phosphorus allocation and growth in tundra production. In Vegeta- tion and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag., pp. 577-598. Miller, P.C, W.D. Billings and W.C Oechel (1979) A modehng ap- References 521 proach to low temperatures. In Comparative Mechanics of Cold Ad- aptation in the Arctic (L.S. Underwood, L.L. Tieszen, A.B. Calla- han and G.E. Folk, Eds.). New York: Academic Press, pp. 181-214. Miller, P.C. and W.A. Stoner (1979) Canopy structure and environment- al interactions. In Topics in Plant Population Biology (O. Solbrig, S. Jain, G.B. Johnson and P.H. Raven, Eds.). New York: Colum- bia University Press, pp. 428-458. Milthorpe, F.L. and J. Moorby (1974) An Introduction to Crop Phys- iology. Cambridge: Cambridge University Press, 202 pp. Minderman, G. (1968) Addition, decomposition and accumulation of or- ganic matter in forests. Journal of Ecology, 56: 355-362. Mishustin, E.W. and V.A. Mirzoeva (1972) Soil microflora in the north- ern USSR. In Microflora of the Soils of the Northern and Central USSR. Jerusalem: Israel Programme for Scientific Translations. Mitchell, W.W. (1973) Adaptations of species and varieties of grasses for potential use in Alaska. In Proceedings of the Symposium on the Impact of Oil Resource Development on Northern Plant Communi- ties: Presented at the 23 rd A A AS Alaska Science Conference, Fair- banks, Alaska, 17 August 1972 (B.H. McCown and D.R. Simpson, Coordinators). University of Alaska, Institute of Arctic Biology, Occasional Publications on Northern Life 1, pp. 2-6. Monk, CD. (1966) An ecological significance of evergreenness. Ecology, 47: 504-505. Monteith, J.L. (1965) Light distribution and photosynthesis in field crops. Annals of Botany, N.S., 29: 17-37. Monteith, J.L. (1974) Principles of Environmental Physics. London: Edward Arnold, 241 pp. Mooney, H.A. (1972) The carbon balance of plants. Annual Review of Ecology and Systematics, 3: 315-346. Mooney, H.A. (1977) Convergent Evolution in Chile and California, Mediterranean Climate Ecosystems. Stroudsburg, Pa.: Dowden, Hutchinson and Ross, 224 pp. Mooney, H.A. and W.D. Billings (1960) The annual carbohydrate cycle of alpine plants as related to growth. American Journal of Botany, 47: 594-598. Mooney, H.A. and W.D. Billings (1961) Comparative physiological ecology of arctic and alpine populations of Oxyria digyna. Ecologi- cal Monographs, 31: 1-29. Mooney, H.A. and A.W. Johnson (1965) Comparative physiological ecology of an arctic and an alpine population of Thalictrum alpinum. Journal of Ecology, 46: 721-727. Mooney, H.A., S.L. Gulmon, D.J. Parsons and A.T. Harrison (1974) 522 References Morphological changes within the chaparral vegetation type as re- lated to elevational gradients. MadroJio, 11: 281-316. Moore, J.J., P. Dowding and B. Healy (1975) Glenamoy, Ireland. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Pro- ductivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 321-343. Moore, P.D. and D.J. Bellamy (1974) Peatlands. New York: Springer- Verlag, 221 pp. Moritz, R.E. (1977) On a possible sea-breeze circulation near Barrow, Alaska. Arctic and Alpine Research, 9: 427-431. Morton, A.J. (1977) Mineral nutrient pathways in a Molinietum in autumn and winter. Journal of Ecology, 65: 993-999. Moser, T.J., T.H. Nash, III and J.W. Thomson (1979) Lichens of Anak- tuvuk Pass, Alaska, with emphasis on the impact of caribou graz- ing. Bryologist, 82: 393-408. Mosser, J.L., G.M. Herdrich and T.D. Brock (1976) Temperature op- tima for bacteria and yeasts from cold-mountain habitats. Canadian Journal of Microbiology, 11: 324-325. Muc, M. (1977) Ecology and primary production of sedge-moss meadow communities, Truelove Lowland. In Truelove Lowland, Devon Is- land, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmon- ton: The University of Alberta Press, pp. 157-184. Mullen, D.A. (1965) Physiologic correlations with population density and other environmental factors in the brown lemming, Lemmus tri- mucronatus. Ph.D. dissertation. University of California (Berkeley), 171 pp. Mullen, D.A. (1968) Reproduction in brown lemmings {Lemmus trimu- cronatus) and its relevance to their cycle of abundance. University of California Publications in Zoology, 85: 1-24. Mullen, D.A. and F.A. Pitelka (1972) Efficiency of winter scavengers in the Arctic. Arctic, 25: 225-231. Muller, K. (1909) Untersuchung uber die Wasseraufnahme durch Moose und verschiedene andere Pflanzen und Pflanzenteile. Jahrbuch fur Wissenschaftliche Botanik, 46: 587-598. Munervar, F. and A.G. Wollum II (1977) Effects of the addition of phos- phorus and inorganic nitrogen on carbon and nitrogen mineraliza- tion in Andepts from Colorado. Soil Science Society of America, Journal, 41: 540-545. Munn, N.R. (1973) An examination of the inorganic nitrogen status of a soil of the Alaskan coastal tundra plain. M.Sc. thesis, Utah State University, 52 pp. Murray, B.M. and D.F. Murray (1978) CheckHsts of vascular plants. t References 523 bryophytes, and lichens for the Alaskan U.S. IBP Tundra Biome study areas; Barrow, Prudhoe Bay, Eagle Summit. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Ties- zen, Ed.). New York: Springer-Verlag, pp. 647-677. Murray, D.F. (1978) Ecology, floristics, and phytogeography of north- ern Alaska. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 19-36. Myers, J. P. and F.A. Pitelka (1979) Variations in summer temperature patterns near Barrow, Alaska: analysis and ecological interpreta- tion. Arctic and Alpine Research, 11: 131-144. Nakano, Y. and J. Brown (1972) Mathematical modeling and validation of the thermal regimes in tundra soils, Barrow, Alaska. Arctic and Alpine Research, 4: 19-38. National Academy of Sciences (1969) United States-Canadian Tables of Eeed Composition. 2nd revision. Washington, D.C.: U.S. National Academy of Sciences, Publication 1684, 92 pp. National Academy of Sciences (1972) Nutrient Requirements of Domes- tic Animals. 2nd revised edition. Washington, D.C.: U.S. National Academy of Sciences, No. 10, 117 pp. National Academy of Sciences (1974) U.S. Participation in the Interna- tional Biological Program. Report No. 6 of the U.S. National Com- mittee for the International Biological Program. Washington, D.C.: U.S. National Academy of Sciences, 166 pp. Nelson, L. (1977) Growth and survival characteristics of three arctic soil bacteria on Truelove Lowland. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 547-565. Ng, E. and P.C. Miller (1975) A model of the effect of tundra vegetation on soil temperatures. In Climate of the Arctic: Twenty-fourth Alaska Science Conference Proceedings, Fairbanks, Alaska, August 1973 (G. Weller and S.A. Bowling, Eds.). Fairbanks: Geophysical Institute, University of Alaska, pp. 222-226. Ng, E. and P.C. Miller (1977) Validation of a model of the effects of tun- dra vegetation on soil temperatures. Arctic and Alpine Research, 9: 89-104. Ng, E. and P.C. Miller (1980) Soil moisture relations in the southern Cal- ifornia chaparral. Ecology, 61: 98-107. Nielsen, K.F. and E.C. Humphries (1966) Effects of root temperature on plant growth. So/75 and Fertilizers, 29: 1-7. Nielson, CO. (1962) Carbohydrases in soil and litter invertebrates. Oikos, 13: 200-215. Norton, D.W. (1972) Incubation schedules of four species of Calidridine sandpipers at Barrow, Alaska. Condor, 74: 164-176. 524 References Norton, D.W. (1974) Ecological energetics of Calidridine sandpiper breeding in northern Alaska. Ph.D. dissertation, University of Alaska, 163 pp. Norton, D.W., I.W. Ailes and J. A. Curatolo (1975) Ecological relation- ships of the inland tundra avifauna near Prudhoe Bay, Alaska. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 124-133. Novichkova-Ivanova, L.N. (1972) Soil and aerial algae of polar deserts and arctic tundra. In Proceedings, IV International Meeting on the Biological Productivity of Tundra, Leningrad, USSR (F.E. Wielgo- laski and T. Rosswall, Eds.). Stockholm: International Biological Programme Tundra Biome Steering Committee, pp. 261-265. Oberbauer, S. (1978) The influence of plant water relations on species distribution in alpine and tussock tundra. M.S. thesis, San Diego State Univeristy, 172 pp. Oberbauer, S. and P.C. Miller (1979) Plant water relations in montane and tussock tundra vegetation types in Alaska. Arctic and Alpine Research, 11: 69-81. O'Connor, F.B. (1967) The Enchytraeidae. In Soil Biology (A. Burges and F. Raw, Eds.). London: Academic Press, pp. 213-257. Odum, H.T. (1970) Rain forest structure and mineral-cycling home- ostasis. In A Tropical Rain Forest: A Study of Irradiation and Ecol- ogy at El Verde, Puerto Rico (H.T. Odum, Ed.). Oak Ridge, Tenn.: U.S. Atomic Energy Commission, pp. H-3 to H-52. Oechel, W.C. (1976) Seasonal patterns of temperature response of CO2 flux and acclimation in arctic mosses growing in situ. Photosyn- thetica, 10: 447-456. Oechel, W.C. and N.J. ColUns (1973) Seasonal patterns of CO2 exchange in bryophytes at Barrow, Alaska. In Primary Production and Pro- duction Processes: Proceedings of the Conference, Dublin, Ireland, April 1973 (L.C. Bhss and F.E. Wielgolaski, Eds.). Stockholm: In- ternational Biological Programme Tundra Biome Steering Commit- tee, pp. 197-203. Oechel, W.C, P. Hicklenton, B. Sveinbjornsson, P.C. Miller and W.A. Stoner (1975) Temperature acclimation of photosynthesis in D. fus- cescens growing in situ in the arctic and subarctic. In Proceedings of the Circumpolar Conference on Northern Ecology, Ottawa, 1975. Ottawa: National Research Council of Canada, pp. 1-131 to 1-144. Oechel, W.C. and N.J. Collins (1976) Comparative CO2 exchange pat- terns in mosses from two tundra habitats at Barrow, Alaska. Cana- dian Journal of Botany, 54: 1355-1369. Oechel, W.C. and B. Sveinbjornsson (1978) Primary production pro- cesses in arctic bryophytes at Barrow, Alaska. In Vegetation and References 525 Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 269-298. Ohmura, A. (1972) Heat and water balance on arctic tundra. In Interna- tional Geography 1972: Papers Submitted to the 22nd International Geographical Congress, Canada (W.P. Adams and F.M. Helleiner, Eds.). Toronto: University of Toronto Press, pp. 175-176. Ohmura, A. and F. Muller (1976) Heat balance measurement on arctic tundra. Axel Heiberg Island, Canadian Arctic Archipelago. Pro- ceedings, 23rd International Geographical Congress, Moscow 1976, pp. 80-84. Olsen, S.R., W.D. Kemper and R.D. Jackson (1962) Phosphate diffu- sion to plant roots. Soil Science Society of America, Proceedings, 26: 222-227. Olson, J.S. (1963) Energy storage and the balance of producers and de- composers in ecological systems. Ecology, 44: 322-331. Olson, J.S. (1970) Carbon cycles and temperate woodlands. In Analysis of Temperate Forest Ecosystems (D.E. Reichle, Ed.). New York: Springer-Verlag, pp. 226-241. Omdhansky, V.L. (1911) Etude bacteriologique du Mammouth de Sanga Jorach et du sol adjacent. Archives des Sciences Biologiques publi^es par l' Institute Imperial de M^decine Exp^rimentale a St. Petersbourg, 16: 355-367. Oosting, H.J. (1956) 77?^ Study of Plant Communities: An Introduction to Plant Ecology. 2nd edition. San Francisco: W.H. Freeman, 440 pp. Orians, G.H. and O.T. Solbrig (1977) A cost-income model of leaves and roots with special reference to arid and semiarid areas. American Naturalist, HI: 677-690. Osborn, R.G. (1975) Models of lemming demography and avian preda- tion near Barrow, Alaska. M.S. thesis, San Diego State University, 106 pp. Ostbye, E. (Ed.) (1975) Hardangervidda, Norway. In Structure and Func- tion of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Productivity of Tun- dra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 225-264. Outcalt, S.I. (1974) Gradient mapping of patterned ground characteris- tics from a photomosaic of the IBP Tundra Biome site near Barrow, Alaska. Mathematical Geology, 6: 235-244. Outcalt, S.I., C. Goodwin, G. Weller and J. Brown (1975) A digital com- puter simulation of the annual snow and soil thermal regimes at Bar- row, Alaska. Water Resources Research, 11: 709-715. Ovington, J.D. (1968) Some factors affecting nutrient distribution within 526 References ecosystems. In Functioning of Terrestrial Ecosystems at Primary Production Level: Proceedings of the Copenhagen Symposium (F.E. Eckardt, Ed.). Natural Resources Research 5. Paris: UNESCO, pp. 95-105. Parinkina, O.M. (1974) Bacterial production in tundra soils. In Soil Organisms and Decomposition in Tundra: Proceedings of the Mi- crobiology, Decomposition and Invertebrate Working Groups Meet- ing, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stockholm: Interna- tional Biological Programme Tundra Biome Steering Committee, pp. 65-77. Parker, G.R. (1972) Biology of the Kaminuriak population of barren- ground caribou. Part I. Total numbers, mortality, recruitment, and seasonal distribution. Canadian Wildlife Service Report Series 20, 95 pp. Parkhurst, D.F. and O.L. Loucks (1972) Optimal leaf size in relation to environment. Journal of Ecology, 60: 505-537. Parkinson, D. (1971) Studies on fungi in Canadian aspen forest soils. In Productivity of Forest Ecosystems: Proceedings of the Brussels Symposium (P. Duvigneaud, Ed.). Paris: UNESCO, pp. 425-430. Parkinson, D., S. Visser and J.B. Whittaker (1977) Effects of Collem- bolan grazing on fungal colonization of leaf litter. In Soil Organ- isms as Components of Ecosystems: International Soil Zoology Col- loquium, Uppsala, Sweden, June 21-25, 1976(1]. Lohn and T. Pers- son, Eds.). Ecological Bulletins 25. Stockholm: Swedish Natural Science Research Council, pp. 75-79. Patterson, P.M. (1943) Some ecological observations on bryophytes. The Bryologist, 46: 1-13. Patterson, R.P., D.L. Grunes and D.J. Lathwell (1972) Influence of root-zone temperature and P supply on total and inorganic P, free sugars, aconitase and soluble amino N in corn. Crop Science, 12: 227-230. Paul, E.A., V.O. Biederbeck, W.E. Lowe and J.R. Willard (1973) Soil microorganisms. I: Population dynamics of bacteria and actinomy- cetes. Canadian Committee for the International Biological Pro- gram Technical Report No. 37, Matador Project. Payne, T.C. et al. (1951) Geology of the Arctic Slope of Alaska. Oil and Gas Investigation Map OM 126. U.S. Geological Survey, Washing- ton, D.C. Payton, I.J. and D.J. Brasch (1978) Growth and nonstructural car- bohydrate reserves in Chionochloa rigida and C macra, and their short-term response to fire. New Zealand Journal of Botany, 16: 435-460. Pearson, O.P. (1966) The prey of carnivores during one cycle of mouse References 527 abundance. Journal of Animal Ecology, 35: 217-233. Penning de Vries, F.W.T. (1972a) A model for simulating transpiration of leaves with special attention to stomatal functioning. Journal of Applied Ecology, 9: 57-77. Penning de Vries, F.W.T. (1972b) Respiration and growth. In Crop Pro- cesses in Controlled Environments (A.R. Rees, K.E. Cockshull, D.W. Hand and R.G. Hurd, Eds.). London: Academic Press, pp. 327-347. Penning de Vries, F.W.T. (1973) Substrate utilization and respiration in relation to growth and maintenance in higher plants. Ph.D. disserta- tion. Agricultural University, Wageningen, 357 pp. Penning de Vries, F.W.T. (1974) Substrate utiHzation and respiration in relation to growth and maintenance in higher plants. Netherlands Journal of Agricultural Sciences, 11: 40-44. Penning de Vries, F.W.T. (1975) The cost of maintenance processes in plant cells. Annals of Botany, N.S., 39: 77-92. Penning de Vries, F.W.T., A.H.M. Brunsting and H.H. van Laar (1974) Products, requirements and efficiency of biosynthesis: a quantita- tive approach. Journal of Theoretical Biology, 45: 339-377. Person, S.J. (1975) Digestibility of indigenous plants utilized by Rangifer tarandus. Ph.D. dissertation. University of Alaska, 97 pp. Person, S.J., R.G. White and J.R. Luick (1975) In vitro digestibility of forages utilized by Rangifer tarandus. In Proceedings of the First International Caribou and Reindeer Symposium, 9-11 August 1972, University of Alaska (J.R. Luick, P.C. Lent, D.R. Klein and R.G. White, Eds.). Biological Papers of the University of Alaska, Special Report 1, pp. 251-256. Person, S.J., R.E. Pegau, R.G. White and J.R. Luick (1980) In vitro and nylon bag digestibilities of reindeer and caribou forages. Journal of Wildlife Management, 44: 613-622. Petersen, H. (1971)Thenutritionalbiology of Collembolaand its ecologi- cal significance. A review of recent literature with a few original observations (in Danish, English summary). Entomologiske Med- delelser, 39: 97-118. Petersen, H. (1975) Estimation of dry weight, fresh weight, and calorific content of various Collembolan species. Pedobiologia, 15: 222-243. Peterson, K.M. (1978) Vegetational successions and other ecosystemic changes in two arctic tundras. Ph.D. dissertation, Duke University, 305 pp. Peterson, K.M. and D.W.. Billings (1975) Carbon dioxide flux from tun- dra soils and vegetation as related to temperature at Barrow, Alaska. American Midland Naturalist, 94: 88-98. Peterson, R.M, Jr. and G.O. Batzh (1975) Activity patterns in natural populations of the brown lemming (Lemmus trimucronatus). Jour- 528 References nal of Mammalogy, 56: 718-720. Peterson, R.M, Jr., G.O. Batzli and E.M. Banks (1976) Activity and energetics of the brown lemming in its natural habitat. Arctic and Alpine Research, 8: 131-138. Peterson, W.L. and J.M. Mayo (1975) Moisture stress and its effect on photosynthesis in Dicranum polysetum. Canadian Journal of Botany, 53: 2897-2900. Petrides, G.A. and W.G. Swank (1966) Estimating the productivity and energy relations of an African elephant population. In Proceedings of the Ninth International Grasslands Congress, Sao Paulo, Brazil, pp. 832-842. Pianka, E.R. (1970) On r- and K- selection. American Naturalist, 140: 594-597. Pilat, A. and J. A. Nannfeldt (1954) Notulae ad cognitionem Hymeno- mycetum Lapponiae Tornensis (Sueciae). Friesia, 1: 6-38. Pitelka, F.A. (1957a) Some aspects of population structure in the short- term cycle of the brown lemming in northern Alaska. Cold Spring Harbor Symposium on Quantitative Biology, 22: 237-251. Pitelka, F.A. (1957b) Some characteristics of microtine cycles in the Arc- tic. In Arctic Biology: Eighteenth Annual Biology Colloquium (H.P. Hansen, Ed.). Corvalhs: Oregon State University Press, pp. 73-88. Pitelka, F.A. (1964) The nutrient-recovery hypothesis for arctic micro- tine cycles. I. Introduction. In Grazing in Terrestrial and Marine Environments: A Symposium of the British Ecological Society, Bangor, 11-14 April 1962 (D.J. Crisp, Ed.). Oxford: Blackwell Scientific Publications, pp. 55-56. Pitelka, F.A. (1973) Cycling pattern in lemming populations near Bar- row, Alaska. In Alaskan Arctic Tundra: Proceedings of the 25th Anniversary Celebration of the Naval Arctic Research Laboratory (M.E. Britton, Ed.). Arctic Institute of North America Technical Paper 25, pp. 199-215. Pitelka, F.A. (1974) An avifaunal review for the Barrow region and North Slope of arctic Alaska. Arctic and Alpine Research, 6: 161-184. Pitelka, F.A., P.Q. Tomich and G.W. Treichel (1955) Ecological rela- tions of jaegers and owls as lemming predators near Barrow, Alaska. Ecological Monographs, 25: 85-117. Pitelka, F.A., R.T. Holmes and S.F. MacLean, Jr. (1974) Ecology and evolution of social organization in arctic sandpipers. American Zoologist, 14: 185-204. Pitt, J.I. and J.H.B. Christian (1968) Water relations of xerophilic fungi isolated from prunes. Applied Microbiology, 16: 1853-1858. Poelt, J. (1973) Appendix A: Classification. In The Lichens (V. Ahmad- References 529 jian and M.E. Hale, Eds.). New York: Academic Press, pp. 599-632. Polunin, N. (1935) The vegetation of Akpatok Island, Part II. Journal of Ecology, 23: 161-209. Porter, L. (1975) Nitrogen transformation in ecosystems. In Soil Bio- chemistry. Vol. 4 (E.A. Paul and A.D. McLaren, Eds.). New York: Marcel Dekker, pp. 1-30. Prentki, R.T. (1976) Phosphorus cycling in tundra ponds. Ph.D. disser- tation. University of Alaska, 275 pp. Prentki, R.T., M.C. Miller, R.J. Barsdate, V. Alexander, J. Kelley and P. Coyne (1980) Chemistry. In Limnology of Tundra Ponds, Bar- row, Alaska (J.E. Hobbie, Ed.). Stroudsburg, Pa.: Dowden, Hutch- inson and Ross, pp. 76-178. Prescott, G.W. (1953) Preliminary notes on the ecology of freshwater algae of the Arctic Slope, Alaska, with descriptions of some species. American Midland Naturalist, 50: 463-470. Prescott, G.W. (1961) Ecology of freshwater algae in the Arctic. In Re- cent Advances in Botany: From Lectures and Symposia Presented to the IX International Botanical Congress, Montreal, 1959. Vol. 1. Toronto: University of Toronto Press, pp. 201-207. Procter, D.L.C. (1977) Nematode densities and production on Truelove Lowland. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 347-361. Pruitt, W.O. (1960) Animals in the snow. Scientific American, 202(1): 61-68. Pugh, G.J.F. (1974) Terrestrial fungi. In Biology of Plant Litter Decom- position (C.H. Dickinson and G.J.F. Pugh, Eds.). London: Aca- demic Press, pp. 303-336. Putnins, P. (1966) The sequences of baric weather patterns over Alaska. In Studies on the Meteorology of Alaska. First Interim Report of the Environmental Data Service, Environmental Science Services Administration, Washington, D.C., 81 pp. Raisbeck, J.M. and M.F. Mohtadi (1974) The environmental impacts of oil spills on land in the arctic regions. Water, Air and Soil Pollution, 3: 195-208. Rail, G. (1965) Soil fungi from the alpine zone of the Medicine Bow Mountains, Wyoming. Mycologia, 57: 872-881. Rastorfer, J.R. (1972) Comparative physiology of four West Antarctic mosses. In Antarctic Terrestrial Biology (G.A. Llano, Ed.). Wash- ington, D.C.: American Geophysical Union, pp. 143-161. Rastorfer, J.R. (1978) Composition and bryomass of the moss layers of two wet-tundra meadow communities near Barrow, Alaska. In Veg- etation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 169-183. 530 References Raunkiaer, C. (1934) The Life Forms of Plants and Statistical Plant Geography. Oxford: Clarendon Press, 632 pp. Rausch, R. (1950) Observations of a cyclic decline of lemmings (Lem- mus) on the arctic coast of Alaska during spring 1949. Arctic, 3: 166-177. Raw, F. (1967) Arthropoda (except Acari and Collembola). In Soil Biol- ogy (A. Burges and F. Raw, Eds.). London: Academic Press, pp. 323-362. Rayner, J.M. (Ed.) (1960a) Temperature and wind frequency tables for North America and Greenland. Vol. 1. January-June. McGill Uni- versity, Arctic Meteorology Research Group, Publication in Mete- orology 24, 275 pp. Rayner, J.M. (Ed.) (1960b) Temperature and wind frequency tables for North America and Greenland. Vol. 2. July-December. McGill Uni- versity, Arctic Meteorology Research Group, Publication in Mete- orology 25, 275 p. Redmann, R.E. and Z.M. Abouguendia (1978) Partitioning of respira- tion from soil, litter and plants in a mixed-grassland ecosystem. Oecologia (Berl.), 36: 69-79. Reed, J.C. and A.G. Ronhovde( 1971) Arctic laboratory: Ahistory of the Naval Arctic Research Laboratory at Point Barrow, Alaska. Arctic Institute of North America Technical Report, 748 pp. Rickard, W.E., Jr. and J. Brown (1974) Effects of vehicles on arctic tundra. Environmental Conservation, 1: 55-62. Ricklefs, R.E. (1973) Ecology. Newton, Mass.: Chiron Press, 861 pp. Roby, D.D. (1978) Behavioral patterns of barren-ground caribou of the Central Arctic herd adjacent to the Trans-Alaska oil pipeline. M.S. Thesis, University of Alaska, 200 pp. Rochow, T.F. (1967) The ecology of Thlaspi alpestre in the central Rocky Mountains along altitudinal gradients. Ph.D. dissertation, Duke University, 265 pp. Rodin, L.E., N.I. Bazilevich and N.N. Rozov (1975) Productivity of the world's main ecosystems. In Productivity of World Ecosystems: Proceedings of Symposium, August 31 -September 1, 1972, Seattle, Washington. Washington, D.C.: National Academy of Sciences, pp. 13-26. Rogers, J.C. (1978) Meteorological factors affecting interannual vari- ability of summertime ice extent in the Beaufort Sea. Monthly Weather Review, 106: 890-897. Ross, R. (1956) The cryptogamic flora of the Arctic. III. Algae: Plank- tonic. Botanical Review, 20: 400-416. Rosswall, T. (1974) Decomposition of plant litter in Stordalen — A sum- mary. In Progress Report 1973 (J.G.K. Flower-Ellis, Ed.). Interna- tional Biological Programme, Swedish Tundra Biome Project Tech- ■If, References 531 nical Report 16, pp. 207-212. Rosswall, T. and B.H. Svensson (1974) Chemolithotrophic and photo- synthetic bacteria at Stordalen. In Progress Report 1973 (J.G.K. Flower-Ellis, Ed.). International Biological Programme, Swedish Tundra Biome Project Technical Report 16, pp. 73-79. Rosswall, T., J.G.K. Flower-Ellis, L.G. Johansson, S. Jonsson, B.E. Ryde'n and M. Sonesson (1975) Stordalen (Abisko), Sweden. In Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Pro- ductivity of Tundra, Abisko, Sweden, April 1974 (T. Rosswall and O.W. Heal, Eds.). Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, pp. 265-294. Rosswall, T. and O.W. Heal (Eds.) (1975) Structure and Function of Tundra Ecosystems: Papers Presented at the IBP Tundra Biome V International Meeting on Biological Productivity of Tundra, Abisko, Sweden, April 1974. Ecological Bulletins 20. Stockholm: Swedish Natural Science Research Council, 450 pp. Rouse, W.R., P.F. Mills and R.B. Stewart (1977) Evaporation in high latitudes. Water Resources Research, 13:909-914. Russell, R.S. (1940) Physiological and ecological studies on an arctic vegetation. III. Observations on carbon assimilation, carbohydrate storage and stomatal movement in relation to the growth of plants on Jan Mayen Island. Journal of Ecology, 28: 289-309. Russell, R.S., D.W. Cutler, S.E. Jacobs, A. King and A.G. Pollard (1940) Physiological and ecological studies on an arctic vegetation. II. The development of vegetation in relation to nitrogen supply and soil micro-organisms on Jan Mayen Island. Journal of Ecology, 28: 269-288. Ryan, J.K. (1977) Synthesis of energy flows and population dynamics of Truelove Lowland invertebrates. In Truelove Lowland, Devon Is- land, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmon- ton: The University of Alberta Press, pp. 325-346. Ryddn, B.E. (1977) Hydrology of Truelove Lowland. In Truelove Low- land, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 107-136. Saebc^, S. (1968) The autecology of Rubus chamaemorus L. I. Phosphor- us economy of Rubus chamaemorus in an ombrotrophic mire. Sci- entific Reports from the Agricultural College of Norway, 47(1): 1-67. Saebc(, S. (1969) On the mechanism behind the effect of freezing and thawing on dissolved phosphorus in Sphagnum fuscum peat. Scien- tific Reports from the Agricultural College of Norway, 48(14): 1-10. Safriel, U.N. (1975) On the significance of clutch size in nidifugous birds. Ecology, 56: 703-708. 532 References Satchell, J.E. (1971) Feasibility study of an energy budget for Meathop Wood. In Productivity of Forest Ecosystems: Proceedings of the Brussels Symposium (P.E. Duvigneaud, Ed.). Paris: UNESCO, pp. 619-630. Satchell, J.E. (1974) Introduction: Litter-interface of animate/inanimate matter. In Biology of Plant Litter Decomposition (C.H. Dickinson and G.J.F. Pugh, Eds.). New York: Academic Press, pp. xiii-xliv. Savile, D.B.O. (1972) Arctic adaptations in plants. Research Branch, Canada Department of Agriculture Monograph 6, 81 pp. Schamel, D. and D. Tracy (1977) Polyandry, replacement clutches, and site tenacity in the red phalarope {Phalaropus fulicarius) at Barrow, Alaska. Bird-Banding, 48: 314-324. Schlesinger, W.H. (1977) Carbon balance in terrestrial detritus. Annual Review of Ecology and Systematics, 8: 51-81. Scholander, P.P., V. Walters, R. Hock and L. Irving (1950) Body insula- tion of some arctic and tropical mammals and birds. Biological Bul- letin, 99: 225-236. Schramm, U. (1972) Temperature-food interactions in herbivorous in- sects. Oecologia (Berl.), 9: 399-402. Schultz, A.M. (1964) The nutrient recovery hypothesis for arctic microtine cycles. II. Ecosystem variables in relation to arctic micro- tine cycles. In Grazing in Terrestrial and Marine Environments: A Symposium of the British Ecological Society, Bangor, 1 1 -14 April 1962 (D.J. Crisp, Ed.). Oxford: Blackwell Scientific Publications, pp. 57-68. Schultz, A.M. (1969) A study of an ecosystem: The arctic tundra. In The Ecosystem Concept in Natural Resource Management (G.M. Van Dyne, Ed.). New York: Academic Press, pp. 77-93. Schwerdtfeger, W. (1975) Mountain barrier effect on the flow of stable air north of the Brooks Range. In Climate of the Arctic: Twenty- fourth Alaska Science Conference Proceedings, Fairbanks, Alaska, 15-17 August 1973. Fairbanks: Geophysical Institute, University of Alaska, pp. 204-208. Scott, D. (1970) Relative growth rates under controlled temperatures of some New Zealand indigenous and introduced grasses. New Zealand Journal of Botany, 8: 76-81. Scott, D. and W.D. Billings (1964) Effects of environmental factors on standing crop and productivity of an alpine tundra. Ecological Monographs, 34: lAli-llQ. Scuhhorpe, CD. (1967) The Biology of Aquatic Vascular Plants. New York: St. Martin's Press, 610 pp. Searby, H.W. and M. Hunter (1971) Climate of the North Slope, Alaska. NOAA Technical Memorandum AR-4, 54 pp. Seastedt, T.R. and S.F. MacLean, Jr. (1979) Territory size and composi- References 533 tion in relation to resource abundance in Lapland longspurs breed- ing in arctic Alaska. Auk, 96: 131-142. Sellmann, P.V., K.L. Carey, C. Keeler and A.D. Hartwell (1972) Terrain and coastal conditions on the Arctic Alaska Coastal Plain: Arctic environmental data package supplement 1. U.S. Army Cold Re- gions Research and Engineering Laboratory Special Report 165, 72 pp. Sellmann, P.V. and J. Brown (1973) Stratigraphy and diagenesis of perennially frozen sediment in the Barrow, Alaska, region. In Per- mafrost: North American Contribution to the Second International Conference. Washington, D.C.: National Academy of Sciences, pp. 171-181. Sellmann, P.V., J. Brown, R.L Lewellen, H.L. McKim and C.J. Merry (1975) The classification and geomorphic implications of thaw lakes on the Arctic Coastal Plain, Alaska. U.S. Army Cold Regions Re- search and Engineering Laboratory Research Report 344, 20 pp. Sestak, Z., J. Catsky and P.O. Jarvis (Eds.) (1971) Plant Photosynthet- ic Production. Manual of Methods. The Hague: Dr. W. Junk, 818 pp. Sextone, A., K. Everett, T. Jenkins and R.M. Atlas (1978) Fate of crude and refined oils in North Slope soils. Arctic, 31: 339-347. Shaver, G.R. (1976) Ecology of roots and rhizomes in graminoid plants of the Alaskan coastal tundra. Ph.D. dissertation, Duke University, 213 pp. Shaver, G.R. and W.D. BilHngs (1975) Root production and root turn- over in a wet tundra ecosystem, Barrow, Alaska. Ecology, 56(2): 401-409. Shaver, G.R. and W.D. Billings (1976) Carbohydrate accumulations in tundra graminoid plants as a function of season and tissue age. Flora, 165: 247-267. Shaver, G.R. and W.D. BiUings (1977) Effects of day length and temp- erature on root elongation in tundra graminoids. Oecologia (Berl.), 28: 57-65. Shtina, E.A. (1972) Some peculiarities of the distribution of nitrogen- fixing blue-green algae in soils. In First International Symposium on Taxonomy and Ecology of Blue-Green Algae, Madras, 1970 (T.V. Desikachary, Ed.). Madras: Bangalore Press, pp. 294-295. Skoog, R.D. (1968) Ecology of the caribou {Rangifer tarandus granti) in Alaska. Ph.D. dissertation. University of California (Berkeley), 699 pp. Slaughter, C.W., M. Mellor, P.V. Sellmann, J. Brown and L. Brown (1975) Accumulating snow to augment the fresh water supply at Barrow, Alaska. U.S. Army Cold Regions Research and Engineer- ing Laboratory Special Report 217, 21 pp. 534 References Small, E. (1972) Photosynthetic rates in relation to nitrogen cycling as an adaptation to nutrient deficiency in peat bog plants. Canadian Jour- nal of Botany, 50: 2227-2233. Smirnov, N.N. (1958) Some data about the food consumption of plant production of bogs and fens by animals. Verhandlung Internation- ale Vereinigung fur Theoretische und Angewandte Limnologie, 13: 363-368. Smirnov, N.N. (1961) Food cycles in sphagnous bogs. Hydrobiologia, 17: 175-182. Smith, F.E. (1954) Quantitative aspects of population growth. In Dynamics of Growth Processes (E.J. Boell, Ed.). Princeton, N.J.: Princeton University Press, pp. 277-294. Sohlenius, B. (1980) Abundance, biomass, and contribution to energy flow by soil nematodes in terrestrial ecosystems. Oikos, 34: 186-194. Soil Survey Staff (1975) Soil Taxonomy: A Basic System of Soil Classifi- cation for Making and Interpreting Soil Surveys. Soil Conservation Service, U.S. Department of Agriculture Handbook 436, 754 pp. Sonesson, M. (Ed.) (1980) Ecology of a Subarctic Mire. Ecological Bulletins 30. Stockholm: Swedish Natural Science Research Coun- cil, 313 pp. Sdrenson, T. (1941) Temperature relations of phenology of the north- east Greenland flowering plants. Meddelelser om Grdnland, 125(9): 1-305. Scirenson, T. (1948) A method of establishing groups of equal amplitude in plant sociology based on similarity of species content and its ap- plication to analyses of the vegetation of the Danish commons. Bio- logiske Skrifter, 5: 1-34. Spetzman, L.A. (1959) Vegetation of the Arctic Slope of Alaska. U.S. Geological Survey Professional Paper 302B, pp. 19-58. Springett, J. A., J.E. Brittain and B.P. Springett (1970) Vertical move- ment of Enchytraeidae (Oligochaeta) in moorland soils. Oikos, 21: 16-21. Stalfelt, M.G. (1937) Der Gasaustausch der Moose. Planta, 27: 30-60. Standen, V. (1978) The influence of soil fauna on decomposition by micro-organisms in blanket bog litter. Journal of Animal Ecology, 47: 25-38. Stanley, D.W. and R.J. Daley (1976) Environmental control of primary productivity on an Alaskan tundra pond. Ecology, 51: 1025-1033. Starkey, R.L. (1950) Relations of microorganisms to transformations of sulfur in soils. Soil Science, 70: 55-65. Starkey, R.L. (1966) Oxidation and reduction of sulfur compounds in soils. Soil Science, 101: 297-306. Steere, W.C. (1974) The status and geographical distribution of Voitia hyperborea in North America (Musci: Splachnaceae). Bulletin of the References 535 Torrey Botanical Club, 101: 55-63. Steere, W.C. (1976) Ecology, phytogeography, and floristics of arctic Alaskan bryophytes. Journal of the Hattori Botanical Laboratory, 41: 47-72. Steere, W.C. (1978a) Floristics, phytogeography, and ecology of arctic Alaskan bryophytes. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer- Verlag, pp. 141-167. Steere, W.C. (1978b) 77?^ Mosses of Arctic Alaska. Bryophytorum Bibliotheca 14. Vaduz: J. Cramer, 508 pp. Steere, W.C. and H. Inoue (1978) The Hepaticae of arctic Alaska. Jour- nal of the Hattori Botanical Laboratory, 44: 251-345. Stewart, R.G. and W.R. Rouse (1976) Simple models for calculating evaporation from dry and wet tundra surfaces. Arctic and Alpine Research, 8: 263-274. Stewart, W.D.P. (1968) Nitrogen input into aquatic ecosystems. In Algae, Man and the Environment: Proceedings of an International Symposium held at Syracuse University, June 18-30, 1967 (D.F. Jackson, Ed.). Syracuse, N.Y.: Syracuse University Press, pp. 53-72. Stewart, W.D.P. (1969) Biological and ecological aspects of nitrogen fix- ation by free-living micro-organisms. Proceedings of the Royal So- ciety B, 172: 367-388. Stewart, W.D.P. (1973) Nitrogen fixation by photosynthetic micro- organisms. Annual Review of Microbiology, 27: 283-316. Stoner, W.A. and P.C. Miller (1975) Water relations of plant species in the wet coastal tundra at Barrow, Alaska. Arctic and Alpine Re- search, 7: 109-124. Stoner, W.A., P.C. Miller and P.M. Miller (1978a) A test of a model of irradiance within vegetation canopies at northern latitudes. Arctic and Alpine Research, 10: 761-767. Stoner, W.A., P.C. Miller and W.C. Oechel (1978b) Simulation of the effect of the tundra vascular plant canopy on the productivity of four moss species. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer- Verlag, pp. 371-387. Stoner, W.A., P.C. Miller, S.P. Richards and S.A. Barkley (1978c) Inter- nal nutrient recycling as related to plant life form: A simulation ap- proach. In Environmental Chemistry and Cycling Processes: Pro- ceedings of Symposium, Augusta, Georgia, April 28-May 1, 1976 (D.C. Adriano and I.L. Brisbin, Jr., Eds.). U.S. Department of Energy, CONF-760429, pp. 165-181. Stoner, W.A., P.C. Miller and L.L. Tieszen (1978d) A model of plant growth and phosphorus allocation for Dupontia fisheri in coastal. 536 References wet tundra. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 559-576. Streten, N.A. (1974) A satellite view of weather systems over the north- ern American Arctic. Weather, 29: 369-380. Stutz, R.C. (1973) Nitrogen fixation in a high arctic ecosystem. Ph.D. dissertation, University of Alberta, 62 pp. Stutz, R.C. (1977) Biological nitrogen fixation in high arctic soils, Truelove Lowland. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 301-314. Sutton, CD. (1969) Effect of low soil temperature on phosphate nutri- tion of plants: a review. Journal of Science, Food, and Agriculture, 20: 1-3. Sveinbjornsson, B. (1979) Controls of CO2 exchange in arctic Polytri- chum mosses. Ph.D. dissertation, McGill University, Montreal, 190 pp. Svoboda, J. (1977) Ecology and primary production of raised beach communities, Truelove Lowland. In Truelove Lowland, Devon Is- land, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmon- ton: The University of Alberta Press, pp. 185-216. Swanson, C.A. and D.R. Geiger (1967) Time course of low temperature inhibition of sucrose translocation in sugar beets. Plant Physiology, 42: 751-756. Tamm, CO. (1964) Growth of Hylocomium splendens in relation to tree canopy. The Bryologist, 67: 423-426. Taylor, W.R. (1956) The cryptogamic flora of the Arctic. II. Algae: Non-planktonic. Botanical Review, 20: 363-399. Tedrow, J.C.F. (1977) Soils of the Polar Landscapes. New Brunswick, N.J.: Rutgers University Press, 638 pp. Tedrow, J.C.F. and J.E. Cantlon (1958) Concepts of soil formation and classification of arctic regions. Arctic, 11: 166-179. Thompson, D.Q. (1951) Summer food preferences of the brown and col- lared lemmings. In Science in Alaska 1951: Proceedings Second Alaska Science Conference, Fairbanks, p. 347. Thompson, D.Q. (1955a) The role of food and cover in population fluc- tuations of brown lemming at Point Barrow, Alaska. In Transac- tions of the 20th North American Wildlife Conference, pp. 166-176. Thompson, D.Q. (1955b) The 1953 lemming emigration at Point Bar- row, Alaska. Arctic, 8: 37-45. Thompson, D.Q. (1955c) The ecology and population dynamics of the brown lemming (Lemmus trimucronatus) at Point Barrow, Alaska. Ph.D. dissertation, University of Missouri, 138 pp. Thomson, J.W. (1979) Lichens of the Alaskan Arctic Slope. Toronto: References 537 University of Toronto Press, 314 pp. Thomson, L.W. and S. Zalik (1973) Lipids in rye seedlings in relation to vernalization. Plant Physiology, 52: 268-273. Tieszen, L.L. (1972a) Photosynthesis in relation to primary production. In Proceedings IV International Meeting on the Biological Produc- tivity of Tundra, Leningrad, USSR (F.E. Wielgolaski and T. Ross- wall, Eds.). Stockholm: International Biological Programme Tun- dra Biome Steering Committee, pp. 52-62. Tieszen, L.L. (1972b) The seasonal course of aboveground production and chlorophyll distribution in a wet arctic tundra at Barrow, Alaska. Arctic and Alpine Research, 4: 307-324. Tieszen, L.L. (1973) Photosynthesis and respiration in arctic tundra grasses: field light intensity and temperature responses. Arctic and Alpine Research, 5: 239-251. Tieszen, L.L. (1974) Photosynthetic competence of the subnivean vege- tation of an arctic tundra. Arctic and Alpine Research, 6: 253-256. Tieszen, L.L. (1975) CO: exchange in the Alaskan arctic tundra: Sea- sonal changes in the rate of photosynthesis of four species. Photo- synthetica, 9: 376-390. Tieszen, L.L. (Ed.) (1978a) Vegetation and Production Ecology of an Alaskan Arctic Tundra. New York: Springer-Verlag, 686 pp. Tieszen, L.L. (1978b) Photosynthesis in the principal Barrow, Alaska, species: A summary of field and laboratory responses. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Ties- zen, Ed.). New York: Springer-Verlag, pp. 241-268. Tieszen, L.L. (1978c) Summary. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 621-645. Tieszen, L.L. and D.C. Sigurdson (1973) Effect of temperature on car- boxylase activity and stability on some Calvin cycle grasses from the Arctic. Arctic and Alpine Research, 5: 59-66. Tieszen, L.L. and D.A. Johnson (1975) Seasonal pattern of photosyn- thesis in individual grass leaves and other plant parts in arctic Alaska with a portable '"CO: system. Botanical Gazette, 136: 99-105. Tieszen, L.L. and N.K. Wieland (1975) Physiological ecology of arctic and alpine photosynthesis and respiration. In Physiological Adapta- tion to the Environment (F.J. Vernberg, Ed.). New York: Intext Educational PubHshers, pp. 157-200. Tieszen, L.L. and S. Archer (1979) Physiological responses of plants in tundra grazing systems. In Special Management Needs of Alpine Ecosystems (D.A. Johnson, Ed.). Symposium of the Society for Range Management, February 14, 1977, Casper, Wyoming. Range Science Series No. 5, Denver, Colorado, pp. 22-42. 538 References Tieszen, L.L., M.C. Lewis, P.C. Miller, J. Mayo, F.S. Chapin, III and W. Oechel (1981) An analysis of processes of primary production in tundra forms. In Tundra Ecosystems: A Comparative Analysis (L.C. Bliss, O.W. Heal and J.J. Moore, Eds.). Cambridge: Cam- bridge University Press, pp. 285-356 (in press). Tikhomirov, B.A. (1963) Contribution to the Biology of Arctic Plants. Doklady Akademii Nauk SSSR, 154 pp. Trabant, D. and C. Benson (1972) Field experiments on the development of depth hoar. In Studies in Mineralogy and Precambrian Geology. Geological Society of America Memoir 135, pp. 309-322. Treharne, K.J. (1972) Biochemical limitations to photosynthetic rates. In Crop Processes in Controlled Environments (A.R. Rees, K.E. Cockshull, D.W. Hand and R.G. Hurd, Eds.). London: Academic Press, pp. 280-303. Trudell, J. and R.G. White (1980) The effect of forage structure and availability on food intake, biting rate, bite size and daily eating time of reindeer. Journal of Applied Ecology (in press). Tukey, H.B., Jr. (1970) The leaching of substances from plants. Annual Review of Plant Physiology, 21: 305-324. Turner, F.B. (1970) The ecological efficiency of consumer populations. Ecology, 51: 741-742. Ulrich, A. and P.L. Gersper (1978) Plant nutrient limitations of tundra plant growth. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 457-481. Underwood, L.S. (1971) The bioenergetics of the arctic fox (Alopexlago- pus L.). Ph.D. dissertation, Pennsylvania State University, 92 pp. Underwood, L.S. (1975) Notes on the arctic fox (Alopex lagopus) in the Prudhoe Bay area of Alaska. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 144-149. University of Alaska, Arctic Environmental Information and Data Cen- ter (1975) Alaska Regional Profiles: Arctic Region (L.L. Selkregg, Ed.). Office of the Governor, 218 pp. U.S. Department of Commerce. Local climatological data. Annual sum- mary with comparative data (Barrow and Barter Island, Alaska). National Oceanic and Atmospheric Administration, Asheville, North Carolina. U.S. Geological Survey (1971-1976) Water resources data for Alaska. U.S. Geological Survey, Water Resources Division, Anchorage, Alaska. Vaizey, J.R. (1887) On the absorption of water and its relation to the con- stitution of the cell wall in mosses. Annals of Botany, 1: 147-152. References 539 Van Cleve, K. (1974) Organic matter quality in relation to decomposi- tion. In Soil Organisms and Decomposition in Tundra: Proceedings of the Microbiology, Decomposition and Invertebrate Working Groups Meeting, Fairbanks, Alaska, August 1973 (A.J. Holding, O.W. Heal, S.F. MacLean, Jr. and P.W. Flanagan, Eds.). Stock- holm: International Biological Programme Tundra Biome Steering Committee, pp. 311-324. Van Cleve, K. (1977) Recovery of disturbed tundra and taiga surfaces in Alaska. In Recovery and Restoration of Damaged Ecosystems (J. Cairns, Jr., K.L. Dickson and E.E. Herricks, Eds.). Proceedings of the International Symposium on the Recovery of Damaged Ecosys- tems held at Virginia Polytechnic Institute and State University, Blacksburg, Virginia, March 23-25, 1975. Charlottesville: Univer- sity Press of Virginia, pp. 422-455. Van Cleve, K., L.A. Viereck and R.L. Schlenter (1971) Accumulation of nitrogen in alder {Alnus) ecosystems near Fairbanks, Alaska. Arctic and Alpine Research, 3: 101-114. Van Cleve, K. and L.A. Viereck (1972) Distribution of selected chemical elements in even-aged alder {Alnus) ecosystems near Fairbanks, Alaska. Arctic and Alpine Research, 4: 239-255. van der Drift, J. (1959) The role of the soil fauna in the decomposition of forest litter. In Proceedings of 15 th International Congress of Zo- ology (H.R. Hewer and N.D. Riley, Eds.). London: Burlington House, pp. 357-360. Van Dyne, G., F.M. Smith, R.L. Czaplewski and R.G. Woodmansee (1978) Analyses and syntheses of grassland ecosystem dynamics. In Glimpses of Ecology (J.S. Singh and B. Gopal, Eds.). Jaipur, India: International Scientific Publications, pp. 1-79. Verstraete, W. (1978) Critique of "Methods for analysis of denitrifica- tion in soils." In Nitrogen in the Environment. Vol. 2. Soil-Plant- Nitrogen Relationships (D.R. Nielsen and J.G. MacDonald, Eds.). New York: Academic Press, pp. 491-501. Viereck, L.A. and E.L. Little, Jr. (1975) Atlas of United States trees. Vol. 2. Alaska trees and common shrubs. U.S. Department of Agri- culture, Forest Service, Miscellaneous Publication No. 1293, 19 pp. and 81 maps. Vlassak, K., E.A. Paul and R.E. Harris (1973) Assessment of biological nitrogen fixation in grassland and associated sites. Plant and Soil, 38: 637-649. Wales, R.A., L.P. Milligan and E.H. McEwan (1975) Urea recycling in caribou, cattle and sheep. In Proceedings of the First International Reindeer and Caribou Symposium, 9-11 August 1972, University of 540 References Alaska, Fairbanks, Alaska (J.R. Luick, P.C. Lent, D.R, Klein and R.G. White, Eds.). Biological Papers of the University of Alaska, Special Report 1, pp. 297-307. Walker, D.A. (1977) Analysis of effectiveness of a television scanning densitometer for indicating geobotanical features in an ice wedge polygon at Barrow, Alaska. M.A. thesis. University of Colorado, 132 pp. Walker, D.A., P.J. Webber, K.R. Everett and J. Brown (1978) Effects of crude and diesel oil spills on plant communities at Prudhoe Bay, Alaska, and the derivation of oil spill sensitivity maps. Arctic, 31: 242-259. Walker, D.A. and P.J. Webber (1979) Relationships of soil acidity and air temperature to the wind and the effect on vegetation at Prudhoe Bay, Alaska. Arctic, 32: 224-236. Walker, H.J. (1973) Morphology of the North Slope. In Alaskan Arctic Tundra: Proceedings of the 25th Anniversary Celebration of the Naval Arctic Research Laboratory (M.E. Britton, Ed.). Arctic In- stitute of North America Technical Paper 25, pp. 49-92. Walker, T.W. (1957) The sulphur cycle in grassland soils. Journal of British Grassland Society, 12: 10-18. Wallace, L.L. and A.T. Harrison (1978) Carbohydrate mobilization and movement in alpine plants. American Journal of Botany, 65: 1035- 1040. Wallwork, J. A. (1967) Acari. In Soil Biology (A. Burges and F. Raw, Eds.). London: Academic Press, pp. 363-395. Walsh, J.E. (1977) Measurement of the temperature, wind and moisture distribution across the northern coast of Alaska. Arctic and Alpine Research, 9: 175-182. Warren Wilson, J. (1957) Arctic plant growth. Advancement of Science, 13: 383-388. Warren Wilson, J. (1966a) An analysis of plant growth and its control in arctic environments. Annals of Botany, N.S., 30: 383-402. Warren Wilson, J. (1966b) Effect of temperature on net assimilation rate. Annals of Botany, N.S., 30: 753-761. Watson, A. (1957) The behavior, breeding and food-ecology of the snowy owl Nyctea scandiaca. Ibis, 99: 419-462. Watson, C.E. (1959) Climates of the states, Alaska. Climatography of the United States No. 60-49. U.S. Department of Commerce, 24 pp. Watson, D.G., W.C. Hanson, J.J. Davis and C.E. Cushing (1966) Lim- nology of tundra ponds and Ogotoruk Creek. In Environment of the Cape Thompson Region, Alaska (N.J. Wilimovsky and J.N. Wolfe, Eds.). Oak Ridge, Tenn.: U.S. Atomic Energy Commission, pp. 415-435. Weaver, D.F. (1969) Radiation regime over arctic tundra, 1965. Univer- I References 541 sity of Washington, Department of Atmospheric Sciences, Scientific Report, 260 pp. Weaver, D.F. (1970) Radiation regime over arctic tundra and lakes, 1966. University of Washington, Department of Atmospheric Sciences, Scientific Report, 112 pp. Weaver, J.E. (1954) North American Prairie. Lincoln, Neb.: Johnsen, 348 pp. Weaver, J.E. (1958) Summary and interpretation of underground devel- opment in natural grassland communities. Ecological Monographs, 28: 55-78. Weber, N.A. (1950a) The role of lemmings at Point Barrow, Alaska. Science, 111: 552-553. Weber, N.A. (1950b) A survey of the insects and related arthropods of arctic Alaska. Part I. Transactions of the American Entomological Society, 76: 147-206. Webb, D.P. (1977) Regulation of deciduous forest litter decomposition by soil arthropod feces. In The Role of Arthropods in Forest Eco- systems (W.J. Mattson, Ed.). New York: Springer-Verlag, pp. 57-69. Webber, P.J. (1971) Gradient analysis of the vegetation around the Lewis Valley, north-central Baffin Island, Northwest Territories, Canada. Ph.D. dissertation. Queen's University at Kingston, 366 pp. Webber, P.J. (1978) Spatial and temporal variation of the vegetation and its productivity, Barrow, Alaska. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 37-112. Webber, P.J. and D.A. Walker (1975) Vegetation and landscape analysis at Prudhoe Bay, Alaska: A vegetation map of the Tundra Biome study area. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 80-91. Webber, P.J., J.C. Emerick, D.C.E. May and V. Komarkova (1976) The impact of increased snowfall on alpine vegetation. In Ecological Im- pacts of Snowpack Augmentation in the San Juan Mountains, Col- orado (H.W. Steinhoff and J.D. Ives, Eds.). Final Report to the Bu- reau of Reclamation, Colorado State University FNR-7052-1. Fort Collins: College of Forestry and Natural Resources, Colorado State University, pp. 201-264. Webber, P.J. and D.E. May (1977) The distribution and magnitude of belowground plant structures in the alpine tundra of Niwot Ridge, Colorado. Arctic and Alpine Research, 9: 157-174. Webber, P.J. and J.D. Ives (1978) Damage and recovery of tundra vege- tation. Environmental Conservation, 5: 171-182. Wein, R.W. and L.C. Bliss (1973) Experimental crude oil spills on arctic 542 References plant communities. Journal of Applied Ecology, 10: 671-682. Wein, R.W. and L.C. Bliss (1974) Primary production in arctic cot- tongrass tussock tundra communities. Arctic and Alpine Research, 6: 261-274. Weiser, C.J. (1970) Cold resistance and injury in woody plants. Science, 169: 1269-1278. Weller, G., S. Cubley, S. Parker, D. Trabant and C. Benson (1972) The tundra microclimate during snow-melt at Barrow, Alaska. Arctic, 25: 291-300. Weller, G. and B. Hoimgren (1974a) The microclimates of the arctic tun- dra. Journal of Applied Meteorology, 13: 854-862. Weller, G. and B. Holmgren (1974b) Summer global radiation and al- bedo— Data for three stations in the Arctic Basin: Ice Island T-3, Barrow, Prudhoe Bay, 1971-1973. University of Alaska, Geophysi- cal Institute Technical Report 2, 31 pp. Wendler, G., N. Ishikawa and N. Streten (1974) The climate of the McCall Glacier, Brooks Range, Alaska, in relation to its geographical set- ting. Arctic and Alpine Research, 6: 307-318. West, G.C. and M.S. Meng (1966) Nutrition of willow ptarmigan in northern Alaska. Auk, 83: 603-615. West, G.C. and D.W. Norton (1975) Metabolic adaptations of tundra birds. In Physiological Adaptation to the Environment (F.J. Vern- berg, Ed.). New York: Intext Educational Publishers, pp. 301-329. White, L.M. (1973) Carbohydrate reserves in grasses: a review. Journal of Range Management, 26: 13-18. White, R.G. (1975) Some aspects of nutritional adaptations of arctic herbivores. In Physiological Adaptation to the Environment (F.J. Vernberg, Ed.). New York: Intext Educational Publishers, pp. 239-268. White, R.G. (1979) Nutrient acquisition and utilization in arctic herbi- vores. In Comparative Mechanisms of Cold Adaptation (L.S. Underwood, L.L. Tieszen, A.B. Callahan and G.E. Folk, Eds.). New York: Academic Press, pp. 13-50. White, R.G., B.R. Thomson, T. Skogland, S.J. Person, D.E. Russell, D.F. Holleman and J.R. Luick (1975) Ecology of caribou at Prud- hoe Bay, Alaska. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the University of Alaska, Special Report 2, pp. 150-201. White, R.G. and M.K. Yousef (1978) Energy expenditures in reindeer walking on roads and on tundra. Canadian Journal of Zoology, 56: 215-223. White, R.G. and J. Trudell (1980) Habitat preference and forage con- sumption by reindeer and caribou near Atkasook, Alaska. Arctic and Alpine Research, 12: 511-529. I I I References 543 White, R.G., F.L. Bunnell, E. Gaare, T. Skogland and B. Hubert (1981) Ungulates on arctic ranges. In Tundra Ecosystems: A Comparative Analysis (L.C. Bliss, Jr., O.W. Heal and J.J. Moore, Eds.). Cam- bridge: Cambridge University Press, pp. 397-483 (in press). Whitehead, P.E. and E.H. McEwan (1973) Seasonal variation in the plasma testosterone concentration of reindeer and caribou. Cana- dian Journal of Zoology, 51: 651-658. Whitfield, D.W.A. (1977) Energy budgets and ecological efficiencies on Truelove Lowland. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 607-620. Whittaker, R.H. (1967) Gradient analysis of vegetation. Biological Re- views of the Cambridge Philosophical Society, 42: 207-264. Whittaker, R.H. (1975) Communities and Ecosystems. 2nd edition. New York: Macmillan, 385 pp. Whittaker, R.H., G.E. Likens, F.H. Bormann, J.S. Eaton and T.G. Sic- cama (1979) The Hubbard Brook ecosystem study: forest nutrient cycling and element behavior. Ecology, 60: 203-220. Widden, P. (1977) Microbiology and decomposition on Truelove Low- land. In Truelove Lowland, Devon Island, Canada: A High Arctic Ecosystem (L.C. Bliss, Ed.). Edmonton: The University of Alberta Press, pp. 505-530. Widden, P., T. Newell and D. Parkinson (1972) Decomposition and microbial populations of Truelove Lowland, Devon Island. Devon Island LB. P. Project, High Arctic Ecosystem, Project Report 1970 and 1971 (L.C. Bliss, Ed.). Edmonton: Department of Botany, Uni- versity of Alberta, pp. 341-358. Wielgolaski, F.E. (Ed.) {\915a) Eennoscandian Tundra Ecosystems. Part 1. Plants and Microorganisms. Berlin: Springer-Verlag, 366 pp. Wielgolaski, F.E. (Ed.) (1975b) Fennoscandian Tundra Ecosystems. Part 2. Animals and Systems Analysis. Berhn: Springer-Verlag, 337 pp. Wielgolaski, F.E. (1975c) Productivity of tundra ecosystems. In Produc- tivity of World Ecosystems: Proceedings of a Symposium, August 3 1 -September I, 1972, Seattle, Washington. Washington, D.C.: National Academy of Sciences, pp. 1-12. Wielgolaski, F.E., S. Kjelvik and P. Kallio (1975) Mineral content of tundra and forest tundra plants in Fennoscandia. In Fennoscandian Tundra Ecosystems. Part 1. Plants and Microorganisms {¥ .E. Wiel- golaski, Ed.). Berlin: Springer-Verlag, pp. 316-332. Wiggins, I.L. (1951) The distribution of vascular plants on polygonal ground near Point Barrow, Alaska. Contributions from the Dudley Herbarium, 4: 41-56. Wilkinson, S.R. and R.W. Lowrey (1973) Cycling of mineral nutrients in 544 References pasture ecosystems. In Chemistry and Biochemistry of Herbage, Vol. II (G.W. Butler and R.W. Bailey, Eds.). New York: Academic Press, pp. 247-315. Williams, J.R. (1970) Ground water in the permafrost regions of Alaska. U.S. Geological Survey Professional Paper 696, 83 pp. Williams, M.E., E.D. Rudolph and E.A. Schofield (1975) Selected data on lichens, mosses and vascular plants on the Prudhoe Bay tundra. In Ecological Investigations of the Tundra Biome in the Prudhoe Bay Region, Alaska (J. Brown, Ed.). Biological Papers of the Uni- versity of Alaska, Special Report 2, pp. 213-215. Williams, M.E., E.D. Rudolph, E.A. Schofield and D.C. Prasher (1978) The role of lichens in the structure, productivity and mineral cycling of the wet coastal Alaskan tundra. In Vegetation and Production Ecology of an Alaskan Arctic Tundra (L.L. Tieszen, Ed.). New York: Springer-Verlag, pp. 185-206. Williams, R.F. (1955) Redistribution of mineral elements during develop- ment. Annual Review of Plant Physiology, 6: 25-42. Witkamp, M. (1966) Rates of carbon dioxide evolution from the forest floor. Ecology, 47: 492-493. Woo, M.-K. (1976) Evaporation and water level in the active layer. Arc- tic and Alpine Research, 8: 213-217. Woodmansee, R.G., J.L. Dodd, R.A. Bowman, F.E. Clark and C.E. Dickinson (1978) Nitrogen budget of a shortgrass prairie ecosystem. Oecologia (Berl.), 34: 363-376. Young, B.A. and J.L. Corbett (1972) Maintenance energy requirement of grazing sheep in relation to herbage availability. Australian Jour- nal of Agricultural Research, 23: 57-76. Younkin, W.E. (1976) Revegetation studies in the northern Mackenzie Valley region. Arctic Gas Biological Report Series, 38. Various pagination. Zinkler, D. (1969) Vergleichende Untersuchungen zum Wirkungs- spekdrum der Carbohydrasen von Collembolen (Aplerygola). Ver- handlingen der Deutsche Zoologischen Gesellschaft, Zoologischer Anzeiger 32, Supplementband, pp. 640-644. APPENDIX 1 U.S. IBP Tundra Biome Projects, Personnel, Site Locations 1970-1974 Location and years Ar Alpine ctic Eagle Niwot Project title Personnel and affiliation Barrow Prudhoe Summit Ridge Other Producers Primary production. Dr. Larry L. Tieszen, 70-73 70 73 photosynthesis and nutri- Augustana College (Sioux ient dynamics in tundra Falls) vegetation Dr. Mary L. Allessio Leek, Rider College 72 Dr. John G. Dennis, Na- 70-71 71 tional Park Service Dr. Brent McCown, Univ. 71-72 73L of Wisconsin John Ahrendt 70 David Albright 72 Michael W. Battrum 71 71 Rick Bohnsack 71 Dr. Terry V. Callaghan, 72 British Antarctic Survey Dr. Nigel Collins, The 72 University, Birmingham, England Elizabeth Collins 72 Carol Dennis 71 Margaret Dillon 74 Gary Fischer 71 71 David Greiner 70 Dale Harrison 70 Donald L. Hazlett 71 71 Dr. Douglas A. Johnson 70-71 71 72-73 (M.S., Ph.D.), USDA, SEA, Utah Claire Lewellen 71 Dr. Martin C. Lewis, York 73 Univ., Ontario Richard Mandsager 72-73 Phihp Mattheis 73 Corrine Mikhelson 71 Gregory E. Mowers 71 71 Richard Nelson 71-72 71 Bruce Oksol 72 Ken Olson 70 70 Robert Pritchard 72L Donna C. Sigurdson 71 Jerel Tieszen 71 Sharon Tieszen 70-71 Robert Vaughn 70 Mary Vetter 74 Nancy Wieland 73 Carbon dioxide dynamics Dr. Patrick I. Coyne, 71-72 on arctic coastal tundra USDA, SEA, Okla. Dr. John J. Kelley, Univ. 70-72 of Alaska Barry Corell 72 Mary Ann Coyne 71-72 Principal investigators listed first, remaining personnel alphabetically with senior personnel's current af- filiation. ( ) degrees awarded as part of Biome research (* cooperative with other programs). L = Laboratory or office; dates primarily refer to period of field activity and not subsequent lab and of- fice analyses. 545 546 Appendix 1 Project title Location and years Alpine Arctic Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Producers (cont'd) Root and rhizome growth and respiration in arctic tundra soils Water relations of selected arctic tundra plants Tundra lichen productivity and the role of lichens in nutrient cycling and tundra structure Physiological ecology of arctic bryophytes Patterns of carbon dioxide exchange in principal arctic Dr. W. Dwight Billings, 71-74 Duke Univ. Dr. Kim Peterson (Ph.D.) 73-74 Dr. Gaius R. Shaver 72-74 (Ph.D.), Marine Biology Laboratory, Woods Hole Alex W. Trent (M.A.) 71 Dr. Philip C. Miller, San 72-73 Diego State Univ. James Ehleringer (M.S.) Patsy Miller 73 Russ Moore Wayne Stoner (M.S.) 72 Nancy Wieland Dr. Emanuel D. Rudolph, Ohio State Univ. Douglas C. Prasher Dr. Edmund A. Schofield, Sierra Club (formerly) Michael S. Williams (M.S.) Dr. James R. Rastorfer, Chicago State Univ. James Haeberlin Allen Skorepa Dr. David K. Smith (Ph.D.*), Univ. of Ten- nessee Dr. Harold J. Webster, Fordham Univ. Dr. Walter C. Oechel, McGill Univ. 72 72-74 bryophyte species Patrick W.C. Leung 73 Janet Mailey 74 Therese Ruszczynski 74 Dr. Bjartmar Sveinbjorns- son (Ph.D.*), University of Alaska (Anchorage) 73-74 Halldora Sveinbjornsson 74 Measurement of transpira- Dr. John J. Koranda, Law- 73 tional water flux by tritium rence Radiation Labora- method tory Bruce Clegg 73 John Martin 73 Marshall Stuart 73 Annual nitrogen and re- Dr. Jay D. McKendrick, 73-74 serve carbohydrate cycle in Univ. of Alaska two arctic grass species as Gary Michaelson 74 affected by nitrogen and George A. Mitchell 73-74 phosphorus fertilization Valerie Ott 73 Peter C. Scorup 74 Phosphate uptake and in- Dr. F. Stuart Chapin III 72-73 ternal cycling in tundra (Ph.D.*), Univ. of Alaska plants Arnold Bloom 73 Critical plant analysis Dr. Albert Ulrich, Univ. of 72 values for growth of California (Berkeley) tundra plants Clifford Carlson Kwok Fong Carlos Llano M.S. Mustafa Jeffrey Tennyson 73 71 71-72 72 72 72 72-73 72-73 73 72 72 72-73 72-73 72-73 72 73 73 72 72 72 72-74L 72-74L 72-74L 72-74L Appendix 1 547 Location and years Alpine Arctic Project title Eagle Personnel and affiliation Barrow Prudhoe Summit Niwoi Ridge Other Canopy structure, gas ex- change and water relations in alpine-arctic species Response of arctic, boreal, and alpine biotypes in re- ciprocal transplants Growth rates, phenology and production of certain alpine and arctic plants in Colorado transplant gar- dens Gradient analysis and pri- mary production in tran- sects from arctic and al- pine tundra to taiga Ordination, productivity and mapping of tundra vegetation Producers (cont'd) Dr. Martyn M. Caldwell, 70 Utah State Univ. James Ehleringer (M.S.) Marcee Fareed (M.S.) Diane Hanson Roger Hanson Dr. Douglas A. Johnson Thomas Shoemaker James R. Vickland Dr. William W. Mitchell, Univ. of Alaska Dr. Frank J. Wooding, Univ. of Alaska Dr. Jay D. McKendrIck, Univ. of Alaska William Folkstead Charles Knight Paul Michaelson Keith Poppert David Scharon Dr. Erik K. Bonde, Univ. of Colorado Dr. Maxine Foreman, Community College of Denver James K. Mitchell Dr. James H. Anderson, Univ. of Alaska Frank Bogardus Bruce K. Bright Bruce P. Burba Patrick D. Cahill Dr. Aureal T. Cross, Christopher Cross David Densmore Roseann Densmore Kent P. Gormley Rebecca A. Ludlum J. Page Spencer Stephen E. Tilmann Dr. Patrick J. Webber, Univ. of Colorado Dr. John Andrews, Univ. of Colorado John Batty John Davidson Dr. Diane Ebert (May) (M.S., Ph.D.), Husson College Dr. John C. Emerick (Ph.D.), Colorado School of Mines Dr. Jo Ann W. Flock (Ph.D.*), Univ. of Color- ado Dr. Vera Komarkova (Ph.D.*), Univ. of Color- ado Gregory E. Mowers 72 Rachel Sherer 73 Marie Slack Anne Stilson 73 Kim Sutherland 72-73 72-73 72 71 71 71 72-73 71 73 71-74 72-73 72-73 71-73 71-74 72-73 71 71 71-74L 7I-73L 71-74L 72L 72L 71-73 71-72 71-73 70 70 71 71 71 71 72 72 71 71 71 70 70 70 71-74 71-74 71 71-74 74 74 72-74 72-74 74 73 72 71-74 71-74 71-74 73 72 548 Appendix 1 Project title Location and years Arctic Alpine Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Producers (cont'd) Taxonomy, biogeography and documentation of tundra flora Vegetation survey of the Prudhoe Bay region A cytological and floristic survey of the vascular plants in the vicinity of Barrow Ecological effects of oil spills and seepages in cold- dominated environments Natural landmarks of the Alaskan arctic lowlands Sue Vetter (Clark) 72 Donald A. Walker (M.A.) 73-74 73-74 J. Wied Robyn Willey Dr. David F. Murray, 71-72 71 Univ. of Alaska Dr. Palle Gravensen, Univ. 71 of Alberta Dr. Albert W. Johnson, 72 San Diego State Univ. Alan Batten Janet Leon Barbara Murray 71 Dr. William Steere, New 71-72 71-72 York Botanical Garden Dr. Bonita J. Neiland, 71 Univ. of Alaska Jerome Hok 71 71 Dr. John G. Packer, Univ. 72 of Alberta Marc Galeski 72 Gordon D. McPherson 72 Dr. Jerry Brown, CRREL 70-73 Dr. Brent McCown, Univ. 71-73 of Wisconsin Dr. Patrick I. Coyne 70 Dr. Frederick Deneke, 71-72 USPS, Minn. Richard Haugen 70 Dr. Patrick J. Hunt, 71 USDA/SEA, S.C. Fleetwood Koutz 71 Dr. R.P. Murrmann, 70 70 USDA/SEA, Calif. Warren E. Rickard, Jr. 71-72 Donald Victor 70 Dr. John J. Koranda 74 Charles D. Evans, Univ. of 74 Alaska Dr. George C. West, Univ. of Alaska 73 72 73 71 71-74 72 72-74 72 71-73 70 71-73 73L 71-73 70 71 71-72 70-73 74 74 74 Population studies of the brown lemming Population determination and nutrient flux through lemmings in the tundra ecosystem Consumers Dr. Frank A. Pitelka, 70-74 Univ. of Calif. (Berkeley) Dr. Guy N. Cameron, 70 Univ. of Houston William E. Glanz 71 Lawrence S. Goldstein 72 Dr. George O. Batzli 72-74 Univ. of Illinois (Urbana) David Best 73-74 David Cappetta 74 Ronald Cherry 73 Fred R. Cole (M.S.) 72 Joan Fitzgerald 73 Mark Ginder 73 Glenn Gunterspergen 74 Nancy Hikes 74 Roy M. Peterson (M.S.) 72-73 Gary Ullinsky 73 70-7 IL Appendix 1 549 Project title Location and years Alpine Arctic Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Determination of lemming home range by radio- tracking The interaction of brown lemmings, vegetation and habitat characteristics during the winter Lemming herbivory in arc- tic tundra Small mammal populations and energetics in arctic and subarctic ecosystems The role of weasels as lem- ming predators on arctic coastal tundra Ecology and bioenergetics of small mammalian carni- vores on the Arctic Coastal Plain Food intake, energy ex- penditure and ecology of caribou in arctic tundra 72-73 72 73 73 72 72-73 73 72-73 72-73 70-72 72 Consumers (cont'd) Dr. Edwin M. Banks, Univ. of Illinois (Urbana) Ronald J. Brooks Michael Kirton Tom Kron Michael Melampy Jay Schnell Virginia Schnell Dr. Stephen F. MacLean, Jr., Univ. of Alaska Herbert R. Melchior, Univ. of Alaska Herbert R. Melchior, Univ. of Alaska Dr. Boyd D. Collier, San Diego State Univ. Edward F. Cheslak (M.S.) Jon Keely Walter Koenig Stephen Temple Dr. Dale D. Feist, Univ. of Alaska Dr. George C. West Dr. John W. Coady (Ph.D.*), Alaska Depart- ment of Fish and Game Wayne Couture Lawrence Frank Ray Kendel Tom Lahey Randy Pitney Dr. Paul H. Whitney (Ph.D.*), Lombard North Group, Ltd., Calgary Dr. Stephen F. MacLean, Jr. 72-73 Dr. B. Michael Fitzgerald, 72-73 DSIR, New Zealand Andrew Grossman 73 Dr. Larry S. Underwood, Univ. of Alaska Dr. Robert E. Henshaw, N.Y. State Dept. of En- vironmental Conservation Dr. Robert G. White Univ. of Alaska Dr. Jack R. Luick, Univ. of Alaska Dianne Caley Dr. Raymond D. Camer- on, Alaska Dept. of Fish and Game Robert A. Dieterich, DVM, Univ. of Alaska Paul Frelier A.M. Gau Dr. Dan F. Holleman, Univ. of Alaska Dr. Steven J. Person (Ph.D.*), Univ. of Alaska Donald E. Russell (M.S.) Dr. Terje Skogland, Nor- wegian State Wildlife Re- search Institute 71 71 70 1-72 71 73 70 71 71-72 71-72 71-72 70 70 71-72 71-72 71-72 70 70 71-72 71-72 71-72 70 70 70 70 70 71-72 72-73 72-73 72 72 72 72 72-73 72 73 71 71 71 71L 71 72-73 71 550 Appendix 1 Location and years Alpine Arctic Project title Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Social organization and habitat utilization in the Lapland longspur at Bar- row, Alaska Avian populations and bio- energetics in arctic and subarctic ecosystems Ecology and current status of cliff-nesting raptors in arctic tundra Population ecology and habitat selection of whist- ling swans on the Alaskan Arctic Coastal Plain Population and habitat utilization of white-tailed ptarmigan in the Colorado alpine Population ecology and energetics of tundra soil arthropods and their role in decomposition processes Consumers (cont'd) Dr. Brian R. Thomson, 72 Univ. of Edinburgh Dr. M.K. Yousef, Univ. of 73 Nevada Dr. Frank A. Pitelka, 71-74 Dr. Thomas W. Custer 71-73 (Ph.D.), U.S. Fish and Wildlife Service Michael W. Monroe 73 James V. Remsen 72 Alan P. Romspert 71 Dr. George C. West 70 70 Dr. David W. Norton 70-72 71-72 (Ph.D), Univ. of Alaska Dr. Uriel N. Safriel, 71-73 Hebrew Univ., Jerusalem Zvi Abramski 72 Irwin W. Ailes 71 71 David Anderson 72 Shoshana Ashkenazie 73 (M.S.) James Curatolo 72 Dr. Barbara B. DeWolfe, Univ. of Calif. (Santa Barbara) Scott Kronberg 72 Stephen McDonald 70 71 Jeffrey O. Myll 71-72 Arnold Newman 73 Douglas Schamel 73 Dr. Tom J. Cade, Cornell Univ. Dr. John R. Haugh, U.S. Geological Survey Dr. Clayton M. White, Brigham Young Univ. Kim Sikoryak Paul R. Spitzer Dr. Walter R. Spofford Dr. L.G. Swartz Stanley A. Temple James D. Weaver Dr. Clait E. Braun, Color- ado Division of Wildlife Dr. Terry A. May (Ph.D.), Univ. of Maine Juliet Gee Jeff Norton Gary Siagel Marie A. Vendeville Dr. Stephen F. MacLean, Jr. Lawrence E. Clement (M.S.) James Bumgartner Mark Deyrup Craig Hollingsworth Margaret Miller (Skeel) 70-73 72-73 70 72 70 71 71 Dr. William J.L. Sladen, 71 71 Jr., Johns Hopkins Univ. John Moore 71 71 Robert Munro 71 71 Frank Pines 71 71 Peter Whitehouse 71 73 70-73 71 71 70-71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71-73 71-73 71 72 72 71 70-72 71 72 Appendix 1 551 Project title Location and years Alpine Arctic Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Consumers (cont'd) Abundance, age structure and respiration rates of ground-dwelling spiders in arctic coastal tundra Energy budgets and ele- mental turnover of tundra invertebrates near Barrow, Alaska Abundance and trophic function of tundra nema- todes Composition and abun- dance of the lepidopteran fauna of arctic and alpine Alaskan tundra Abundance of soil arthro- pods and their effects on soil microorganisms in al- pine tundra Populations and produc- tion of major arthropods Thomas McGrath 73 Edward A. Morgan 71-72 71 Ray Price Donald W. Smith 70 J. Ward Testa 73 Louis Verner 71 71 April D. Volk Dr. Boyd D. Collier, San 72 72 Diego State Univ. 72 71 Dr. D.A. Crossley, Jr., 72 Univ. of Georgia Dr. G. Keith Douce (M.S., 72&74 Ph.D.), Univ. of Georgia Dr. Grover C. Smart, Jr., 71 Univ. of Florida Thomas H. Atkinson 72 Dr. Kenelm W. Philip, 71 Univ. of Alaska Dr. John S. Edwards, 71 Univ. of Washington Paul Banko David A. Dailey Dr. Ronald Schmoller 71 Univ. of Tennessee Richard Ambrose Martha Commins Dayle Donaldson Barry Lumpkin Mary McGrade Irene Rosenberg Virginia Tolbert Wayne Tolbert 71 71-72 72 71 71 71 71 71-72 71 72 71 71 71 71 71 72 Metabolic activities of bac- teria and yeast in wet mea- dow tundra soil Basidiomycetes biomass and function in the arctic and alpine tundras The role of algae and pro- tozoan associates in tundra soils Decomposer, Soils, and Nutrient Flux Dr. Robert E. Benoit, Vir- 70-73 ginia Polytechnic Institute Robert Breedlove Walton B. Campbell (M.S.) Gregory Dickenson Richard W. Harris (M.S.) Tryree Kessler Robert E. Moffit Leland S. Warren Ton Rau Dr. Orson K. Miller, Vir- ginia Polytechnic Institute Dr. Gary A. Laursen (Ph.D.) Office of Naval Research Dr. Roy E. Cameron, Argonne National Labor- atory Anne Dalton (Knox) Frank A. Morelli 70 70-72 73 71-72 74 72 71 73 71-74 71-74 71-73 73 73 70 71 71 72L 552 Appendix 1 Locat on and years Arctic Alp ne Eagle Niwot Project title Personnel and affiliation Barrow Prudhoe Summit Ridge Other Decomposer, Soils, and Nutrient Flux (cont'd) Decomposition processes in Dr. Patrick W. Flanagan 70-73 70 70-73 70 arctic and alpine tundra Univ. of Alaska Alan Crawford 71-72 71 71-72 71L Diane Duvall 71L John Ho 70 Aria Scarborough 71-73 70 70-73 Soil-plant interactions and Dr. Paul L. Gersper, Univ. 70-72 71 71 associated multivariate of Calif. (Berkeley) analyses Dr. Rodney J. Arkley, Univ. of Calif. (Berkeley) Douglas Anderson William Bauman Josephine L.Challinor Dr. Harvey F. Donner, Univ. of Calif. (Berkeley) Philip S. Flint Karen Fuller Dr. Rudi Glaser, Univ. of Calif. (Berkeley) Brian Hicks Dr. L. Jacobsen, Univ. of Calif. (Berkeley) Dr. Hans Jenny, Univ. of Calif. (Berkeley) Merril Misam Norton R. Munn (M.S.) Dr. Arnold M. Schultz, Univ. of Calif. (Berkeley) Alex P. Simons (M.S.) Dean Williams John N. Zorich 70-72 73 70 72 71-73 70 72 72 70 72 71 70 70 71 71 71 71 71 72L 73-74 Nitrogen fixation in arctic and alpine terrestrial and aquatic ecosystems Rates of nitrification and denitrification in tundra soils Nitrogen cycling in tundra and taiga soils as influ- enced by temperature and moisture (and laboratory plant nutrient analysis) Phosphorus characteriza- tion in tundra soil Dr. Vera Alexander, Univ. 70-73 Dr. Donald M. Schell, 70-73 Univ. of Alaska Margaret Billington 70-74 Sinikka Kallio 73 Linda Peyton 70-73 G.B. Threlkeld 70 Dr. Stephen A. Norrell, 72-73 Univ. of Alaska Carol M. Anderson 72-73 Mary Johnston (M.S.) 73 George Lindholm (M.S.) 71 Dr. Keith Van Cleve, 72 Univ. of Alaska Nancy Anthony Carol Brass Carol Campbell Rudolph Candler Robert Hardy Joan Holland Joyce Meador Mark Metcalf Lorraine Noonan Tom Pearce Harold Piene (M.S.) Benjamin Sands Robert Schlentner Bonita Snarski Patricia Troth Dr. Robert J. Barsdate 73 Univ. of Alaska 71 71 71 71-73 72-73 71-72 70-71 71 72 71 71 70 70 73 70 71-73L 72-73L 73 L 71-72L 70-74L 70L 70L 71 70-7 IL 71-72L 71L 70-72L 71 72 70L 71-72L 70L 72L I Appendix 1 553 Project title Location and years Alpine Arctic Eagle Niwot Personnel and affiliation Barrow Prudhoe Summit Ridge Other Decomposer, Soils and Nutrient Flux (cont'd) Tundra soil land-water nutrient interactions Peat structure and soil mapping of arctic soils Dr. Dirk Barel, Arheim, Netherlands Dr. Lowell A. Douglas, Rutgers Univ. Dr. Aytekin Bilgin (Ph.D.) (Deceased) Dr. K. R. Everett, Ohio State Univ. Robert Parkinson (M.S.) 73 71 71-72 71-72 71-72 72-74 72-74 74 Abiotic Interface The micro- and regional Dr. Gunter Weller, Univ. 70-72 71 climate of the Alaskan arc- of Alaska tic tundra Dr. Carl S. Benson, Univ. of Alaska 70-74 72-74 Stewart Cubley 71 Gary Hess Dr. Bjorn Holmgren, 73-74 73-74 Meteorological Institute. Sweden Gilberr Mimken 70 Stanley Parker 71-72 72 Scott Parrish (Deceased) 73 72-73 Richard Schwartz 71 71 Robert Timmer 72 Dennis Trabant (M.S.) 70-71 Dr. Gerd Wendler, Univ. 70 of Alaska Prediction and validation Dr. Jerry Brown 70-71 of temperature and thaw in Dr. Yoshisuke Nakano, 71 arctic tundra soils CRREL Robert Arnold 71 Gregor Fellers Richard McGaw 70 William Powell 70 Vaughn Rockney 70-71 David Schaeffer 70 Leander Stroschein Donald Vietor 70 Soil and plant canopy tem- perature regimes in alpine tundra Soil moisture balance in the alpine tundra Ground temperature re- gimes and energy budgets in the Colorado alpine Dr. Erwin R. Berglund, Oregon State Univ. Dr. Dwane J. Sykes, Eyring Research Institute, Provo, Utah Thomas R. Ford William A. Quirk Dr. John D. Ives, Univ. of Colorado Dr. Roger G. Barry, Univ. of Colorado Carol Batty John Clark Jed Fuhrman Bonnie Gray Edway Hinckley Barbara Jarvis Patricia Jenson Kathleen Laughlin Dr. Ellsworth LeDrew (M.S.), Univ. of Waterloo 73 70-71 72 70 70 70-7 1 L 70L 71 71 71 71 71-72 73 72-73 71 72 73 72 73 71 72 73 72-73 554 Appendix 1 Project title Location and years Alpine Arctic Eagle Niwot Personnel and affiliation Barrow Prudhoe Summit Ridge Other Abiotic Interface (cont'd) Evan F. Meltzer David Rowe 72 71 Terrestrial Modeling and Synthesis The analysis of the struc- ture and function of the wet tundra ecosystem at Point Barrow, Alaska A tundra model Simulation modeling of the Barrow tundra ecosystem Modeling tundra primary production processes Modeling vertebrate con- sumers in the arctic tundra Decomposition-nutrient flux and caribou model de- velopment Simulation of meteorologi- cal variation over arctic coastal tundra under per- turbed physical interface conditions Geomorphic and near- surface thermal regime model of tundra terrain Snow-melt and hydrologic modeling of coastal tundra Dr. Harry N. Coulombe, USFWS, Colorado 70 Dr. Jerry Brown, CRREL 70 Dr. Philip C. Miller 70 Dr. K.W. Bridges, Univ. 71 of Hawaii Dr. Mitchell E. Timin, San 71-72 Diego State Univ. Paul Nobbs 72 Jean Rannells Jon S. Zich 72 Dr. Phihp C. Miller 70-73 Dr. Edward G. Brittain, 72-73 Australian National Univ. John Hom 73 Bruce Lawrence (M.S.) 73 Edward Ng (M.S.*) 73 Wayne Stoner (M.S.) 71-73 Dr. Boyd D. Collier 73 Sylvia A. Barkley (M.S.) 73 73-74 Ronald Osburn (M.S.) 72-73 73-74 Nils Stenseth 72 Dr. Fred L. Bunnell (Ph.D.*), Univ. of British Columbia Dr. Pille Bunnell, Univ. British Columbia Penny Lewis Donald E. Russell (M.S.) David Tait Dr. Joseph Pandolfo, Cen- ter for the Environment and Man Dr. Norman W. Lord, Center for the Environ- ment and Man Dr. Marshall Atwater, Center for the Environ- ment and Man 70&73 73 73 73 73 Administrative, technical and logistic support Dr. Samuel 1. Outcalt, 72-73 Univ. of Michigan Dr. Cecil W. Goodwin 72-73 (M.S., Ph.D.), Pennsyl- vania State Univ. Roger Lachenbruch 73 Dr. S. Lawrence Dingman, 72 Univ. of New Hampshire Dr. Robert 1. Lewellen, Littleton, Colorado Central Program Dr. Jerry Brown 70-74 70 70 70 70 71 71-73 72 70-74 73 73 73 73 73 72-74 72 72-74 73-74 73-74 73-74 72-74 70-72 70-72 72-74 72-74 72-74 I 71-74 Dr. George C. West James Baldridge 71 73 70-80 1-72 70-74 70 Appendix 1 555 Location and years Ar Alpi ne ctic Eagle Niwot Project title Personnel and affiliation Barrow Prudhoe Summit Ridge Other Central Program (cont'd) Dr. Patrick I. Coyne 71 70 Christopher C. Cross 71 Dann Farquhar 72 Bobby Fox 72-73 David Grantz 72-73 Johanne Harper 73-74 Dr. John E. Hobbie 73 Jule Loftus 72 Dr. Stephen F. MacLean, Jr. 73 Jean Herb Moore 71 Scott Parrish 72-74 73-74 John Polhemus 72 James Prill 72 Donval Simpson 71 71 Dr. Larry L. Tieszen 71-73 C. Ray Vest 71 Dr. Patrick J. Webber 71-74 David Witt • 72-74 Data processing and stor- Barry Campbell 73 73 age, computer research de- Brucilla Campbell 73 73 sign and modeling t Dr. Frederick C. Dean, Univ. of Alaska James Dryden 71 71-73 71-72 Stephen Geller 72 71-72 Dr. Samuel J. Harbo, Jr., 71 71-72 Univ. of Alaska Dr. Jere Murray, Homer, 71-72 71-72 Alaska Evelyn M. Porter 71 71-73 Robert A. Porter 71 71-73 Frances Randall 71-74 David VanAmburg 73 Dr. Keith Van Cleve 73 71-74 Publications (CRREL) Stephen L. Bowen, CRREL Mary Aho, University of Alaska, Anchorage Cheryl Clark Norma Coutermanche Harold Larsen Laurie McNicholas, Univ. of Alaska Donna Murphy Matthew Pacillo Cheryl Richardson Hazel Sanborn Sandra Smith Audrey Vaughan 70-80 74 74 73 70-80 74 70-80 70-74 70-73 70-74 70 74 Primary productivity and phytoplankton dynamics in arctic ponds and lakes Bacteria and benthic algae activity and productivity in arctic ponds Aquatic Program (see Hobbie 1980) Dr. Vera Alexander, Univ. 70-74 of Alaska Cathleen M. Chmielowski 71 Robert Clasby 71 Christopher Coulon 71-73 Elizabeth Coulon 73 Dr. Staffan Holmgren, 71 Uppsala University Dr. John E. Hobbie, Marine Biological Lab- oratory, Woods Hole Mary E. Bennett Michael Crezee 71-73 71 71 71 556 Appendix 1 Project title Location and years Alpine Arctic Eagle Personnel and affiliation Barrow Prudhoe Summit Niwot Ridge Other Aquatic (cont'd) Dr. Ralph Daley, Inland 73 Waters, W. Vancouver, B.C. Dr. Tom Fenchel, Univ. of 73 Aarhus, Denmark Primary production of vas- cular aquatic plants in arc- tic ponds Daily rhythms, productiv- ity, and seasonal cycles of zooplankton in arctic coastal ponds and lakes Polly Penhale 71-72 Park Rublee (M.S.) 73 Dr. Donald W. Stanley 71-73 (Ph.D.), University of East Carolina Tor S. Traaen 72 Dr. C. Peter McRoy, 71-72 Univ. of Alaska Thomas Leue 72 Dr. Raymond G. Stress, 71-73 State Univ. of New York Dr. Sally W. Chisholm (Ph.D.), Univ. of Calif. (Berkeley) Benthic carbon dynamics and the ecological role of Lepidurus arcticus in coastal tundra ponds 72 Dr. Michael C. Miller, 71-73 Univ. of Cincinnati T.W. Federle (M.S.) Kathleen Kallendorf 73 Robert J. Kallendorf (M.S.) Constance Menefee 73 James P. Reed (M.S.) 71-72 Linda R. Reed 72 D.J. Stromme (M.S.) Nutrient metabolism and Dr. Robert J. Barsdate, 70-73 water chemistry in ponds Univ. of Alaska and lakes of the Arctic Alex Fu 70 Coastal Plain Norman A. King Mary Nebert 71 Cathy Prentki 71 Dr. Richard Prentki 70-73 (Ph.D.), Univ. of Wis- consm Thomas Tribble 70 71 Thomas Casady 73 Dr. Stanley Dodson 73 Univ. of Wisconsin James C. Edwards 72 Stephen Goldstone 73 Gary V. Gulezian 72 Dr. Donald Kangas, 73 Northeast Missouri State Univ. William A. Shapse 71 71 Peter L. Starkweather 71 71 Seasonal cycles and ener- Dr. Raymond D. Dillon, 71 71 getics of benthic communi- Univ. of South Dakota ties in coastal tundra Dennis G. Buechler (M.S.) 71 71 ponds and lakes John T. Hobbs 71-72 Dynamics of freshwater Dr. Donald A. Bierle, 71-72 71 microbenthic communities St. Paul Bible College in arctic ponds and lakes Sharon Bierle Ted Boal 72 Donald G. Hardy 71-72 Karen L. Hart 71 72L 71 71 70-73L Appendix 1 557 Project title Location and years Alpine Arctic Eagle Niwot Personnel and affiliation Barrow Prudhoe Summit Ridge Other Aquatic (cont'd) Energetics of the fish pop- ulation in Ikroavik Lake Modeling the aquatic pond system at Barrow, Alaska Dr. James N. Cameron, 71-72 Univ. of Texas (Port Aransas) Jon Kostoris 72 Polly A. Penhale (M.S.) 71-72 Dr. Jawahar Tiwari, 73 Univ. of California (L.A.) Paul Nobbs 73 71 73-74 73-74 APPENDIX 2 Location of Principal Biome Plots ?sE = ;;-SHiS5SSS533SSSgZi5siiiillslf:: = i I -- = ^"coooooaaooooo SaCCCSCCCCCCCC ^iTooo^ooooaooo OUUUU-iUUUUUUUU ■Coo 9 9 » c c c c 0 0 0 0 (JUUU w I umbe 1306, 1317, 0405 1315, 0402, 1314, 1326 c> 1306. 1307 1312. 1316 1325. 0403 1303. 1304 0206, 0401 1311, 1313. 0412 0431 1321. 1322. ■1 '^ 11 ] 0316 1302 1305 1324 1301 1309 1310 0411 1319 1320 si >. 559 * Subject Index Acari abundance, 421 biomass, 415-416 density, 415 distribution in soil, 419-420 feeding habits, 436 Albedo, 46-47, 61, 65 Algae biomass, 463 distribution, 256, 266-268, 271-272 fungal parasites, 261 nitrogen fixation, 234-235 numbers, 263 soil, 263 species diversity, 263 Allocation biomass, 145 carbohydrates, 142 carbon compounds, 148-158 Dupontia, 152, 154 environmental influence, 155 grazing, 177-181 inflorescences, 140 leaf and stem growth, 140 nutrients, 142, 169-175 patterns, 184 photosynthate, 147 reproduction, 175 rhizome production, 140 root production, 140 see also Translocation Animals Arctic Slope species, 6-7 Aquatic systems movement by nutrients, 481 Arctic Coastal Plain see Coastal Plain Arctic foxes density, 365 predators, 360-361, 365 prey selection, 367 Arctic Front, 31, 34-35 Arctic Slope physiographic provinces, 1-2 plant and animal species, 5-7 Atmospheric circulation, 31 Bacteria, 333 abundance on polygons, 421 anaerobic, 258 biomass distribution, 269-270 decomposition, 473 denitrification, 259 distribution, 255-256, 268, 272 growth, 299-302 iron, 260 microfloral, 289 nitrogen-fixing, 258 numbers, 257 production, 280-281 respiration, 299-302 species diversity, 257, 288 sulfur, 259 utilization of substrate, 259-297 see also Microbial activity Barrow research area geographic description, 10-16 research sites, 12-13 Biomass, 232 allocation, 145 belowground, 76-78 carbon pool size, 458-459 distribution, 195-199 flux, 162-165 microbial, 245 rate of change, 158-159 root, 147 turnover rates, 193 561 562 Index vertical distribution, 74-78 Birds breeding, 445-447, 450-452 carnivorous standing crop, 464 density, 450-452 distribution of species, 444-445 energetics, 452-456 feeding habits, 447-450, 453-456 habitat use, 448-450, 457 incubation, 447-449, 456 predation, 452 predators, 445-456 premigratory movement, 449 see also Jaegers, Short-eared owls, Snowy owls Brooks Range geographic description, 1 Calcium allocation, 169-171 Canopy air temperature, 84 foliage area index, 75-76, 90 humidity, 84 influence on environment, 78 photosynthesis, 120-123, 131 soil temperatures, 86-87 structure, 74-78, 100-101, 131-134 wind, 85 Carbohydrates see TNC Carbon allocation, 140 annual input, 472 budget, 462, 468-469 content in ecosystems, 460-461 content in moss, 463 fixation, 158, 459, 466-467 flux, 467 organic matter, 220-221 pool size, 458 Carbon dioxide assimilation, 102 evolution, 311, 314-316, 325, 330 see also Photosynthesis Carboxylation activity, 106, 108, 173-174 Caribou availability of forage, 393 body size, 399 breeding, 388 comparison with lemmings, 399 density, 388 digestibility of forage, 390-391 food intake, 392 grazing, 386, 389, 402-405 habitat use, 386 insect harassment, 386, 389, 393 life history characteristics, 400 migration, 385-386, 402 mortality of calves, 388 numbers, 387 nutrition, 388, 398-399 population dynamics, 409 population energetics model, 395-398 Prudhoe Bay, 385-399 reproductive rates, 399-401 rumen function model, 394-395 thermoregulation, 397 Carnivores avian, 444-445 invertebrate, 411 Cation exchange capacity, 227-229 Chemical analysis substrate, 293 Chlorophyll, 83 see also Photosynthesis Chmate, 30-37 Coastal Plain, 4-5 Index 563 Cloud cover, 33-34, 46 Coastal Plain Province climate, 4, 5 geographic description, 3-4 geology, 3 physiography, 11 Collembola biomass, 416 density, 416 distribution in soil, 419-420 feeding habits, 435-436 influence on nutrients, 443 Colonization see Succession Decomposer organisms carbon and energy flow, 459 response to environmental conditions, 309 standing crop, 463 Decomposition, 333, 467 bacteria, 260, 295-297 fungi, 297-298 graminoids, 309 methods of study, 292 microbial activity, 291-333 model, 309-325 nutrient cycling, 482 organic matter, 411 rates, 220, 472 soil invertebrates, 442-443 substrate chemistry, 318-321 weight loss, 293 Depth hoar, 38-41 Detritivores, 411 Detritus-based trophic system, 411-457 bioenergetics, 436-441 Diptera abundance, 416 biomass, 416 feeding habits, 436 life cycle, 424-428, 431-432 see also Tipulidae Disturbance, 218 cover removal, 214-215 experimental, 211 lemming grazing, 210 oil spills, 215 recovery, 210 runoff, 210 thaw lake cycle, 210 vegetation removal, 214 vehicle tracks, 212-213 Dunlin see Birds Dupontia allocation, 140, 154 biomass flux, 162-165 growth form, 141-142 growth patterns, 163-164 photosynthesis, 106 population dynamics, 176-177 rhizome growth, 145 root growth, 147 shoot growth, 142-145 tillers, 141-142, 175 TNC, 148-151 see also Graminoids Ecosystem model, 90 Enchytraeidae abundance, 421 biomass, 415 density, 415 distribution in soil, 419-420 energetics, 433-434 feeding habits, 434-435 growth rate, 429-430 life cycle, 428-429 Energy respiration, 159 translocation, 161 Energy balance in microclimate, 49-51 Energy budget in biophysical processes, 66-67 Environmental factors 564 Index control on nitrogen fixation, 236-239 influence on vegetation, 121-122, 186-187, 202-206, 217 Enzyme activity, 303 Ermine density, 366 Evapotranspiration, 51, 53-55, 61-62, 80, 91-93 see also Transpiration Fertilization effect on lemming reproduc- tion, 382 root-to-shoot ratio, 155 Flooding lemming habitat, 371 Fog, 34 Foliage area index, 209 Foothills Province geographic description, 3 Forage nutrient content, 356-357 see also Grazing Forbs growth form, 194 Freeze-thaw, 252, 267 Freeze-up, 64 Fungi, 333 biomass distribution, 268-269 concentration on polygons, 421 decomposition, 473 density, 271 diet of nematodes, 414 distribution, 255-256, 268, 272-276 growth, 299-302 influence of moisture, 286-287 influence of phosphorus, 282-284 influence of substrate, 281-282 influence of temperature, 284-285 numbers, 260 oil spills, 215 production, 278-280 respiration, 299-302 species diversity, 260-262 taxonomic structure, 264-266 utilization of substrate, 295-297 see also Microbial activity Geology Barrow area, 14-16 Geomorphology Coastal Plain, 3-4, 14-16 Graminoids carbohydrate content, 161 growth form, 105, 194 growth patterns, 141-148, 181-184 nutrients, 171-172 production, 144, 148 Grazing allocation, 177-181 caribou, 386 limit on CO2 uptake, 103 photosynthesis, 119, 134 plant production, 178 Growth forms, 217 characteristics, 186-188, 200 definition, 194 distribution, 195-199 photosynthesis, 105 species diversity, 199-200 Growth patterns Dupontia, 163-164 graminoids, 141-148 Herbivores, 409 birds, 336 brown lemmings, 337-384 mammals, 337 Prudhoe Bay, 384-385 Hydrology, 51-65 \ Index 565 Ice wedge formation, 15 Infiltration, 55 Insolation see Radiation Invertebrates, 10 abundance, 412-424 biomass, 412-424, 437 decomposition, 482 diets, 437-439 distribution, 418, 456 energetics, 432-436 feeding habits, 421 habitats, 418 life cycles, 429, 456 nutrient content, 465 respiration, 439-440 role in nutrient cycles, 441-443 soil, 441 spatial variation, 413 species diversity, 412 standing crop, 464 see also Acari, Coilembola, Enchytraeidae, Nematoda, Rotifers, Tardigrades, Tipulidae Irradiance photosynthesis, 127 see also Radiation Jaegers density, 363-364 food consumption, 369 nutrient transport, 408 predators, 360-364 prey selection, 366-367 Landforms, 16-21 Lapland longspurs see Birds Leaf absorptance, 82-88 area index, 108 growth, 153, 200-201, 217 photosynthesis, 106-108 resistance, 106, 129 temperature, 85-86 Lemmings abiotic influences on popula- tion, 371-374 activity patterns, 345-346 body size, 399 breeding, 340-341 brown at Barrow, 337-384 burrows, 43, 346, 350-351, 356, 404 collared at Barrow, 337-338, 347 comparison with caribou, 399-409 cycles, 338-340, 384, 409 demography, 340-343 density, 338-340, 363, 374, 401 density of predators, 363-366 diet, 347-348 digestibihty of forage, 353, 358 effect on soil, 252 energy requirements, 348-358, 381 feces and urine in soil, 263, 377 forage, 375-377, 381 grazing, 177, 179, 212, 346, 354-356,402-405,408-409 482 growth rate, 349, 350, 399 habitat use, 343-347, 371, 405 intrinsic influences on popula- tion, 382 life cycle, 400 metabolism, 349, 351 mortality, 370-371, 373 nutrient content, 465 nutrient-recovery hypothesis, 375-380 nutrient-transport hypothesis, 406-408 566 Index nutrition, 356-360, 375, 380, 409 predation, 360-371, 374-375 predators, 339 Prudhoe Bay, 384 reproduction, 350-351, 360, 377, 380, 382-383, 399-401 standing crop, 464 survival rate, 341-343 thermoregulation, 350 Lichens biomass, 463 nitrogen fixation, 238 Lipid see TNC Mesophyll resistance, 106 Microbial activity, 291 moisture, 305, 307 organic matter turnover, 423 oxygen concentrations, 306 temperature, 399 Microbial biomass, 291, 293, 473 model, 327-328 Microbial growth temperature, 299-302 Microbial maintenance, 328-330 Microbial metabolism, 334 oxygen, 307-309 Microbial respiration, 330, 333 influence of temperature, 299-302 model, 310-324 substrate chemistry, 318-321 weight loss, 321-324 Microbivores, 441 Microclimate, 4, 30, 44-51, 74 Microflora, 255-290 biomass, 289 environmental influences, 287-288, 298-309 species diversity, 288 utilization of substrate, 294-298 see also Algae, Bacteria, Fungi, Yeasts Microorganisms consumption, 457 production, 441 respiration, 439 Microtine rodents, 337 see also Lemmings, Weasels Microtopography, 16-21 Models avian predation, 370-371 box and arrow, viii canopy photosynthesis, 71-74, 79 caribou population energetics, 395-398 caribou rumen function, 394-395 decomposition, 309-326 ecological, viii lemming energetics, 348 lemming nest temperatures, 350-351 lemming nutrients, 357-360 microbial biomass, 327-328 nitrogen fixation, 238-239 plant-water relations, 97-100 snowmelt, 58-61 soil thaw, 87-90 stand photosynthesis, 120-137 surface equilibrium tempera- ture, 37 Mosses, 237 photosynthesis, 112-114, 115-119 role in nutrient cycling, 477-478 Nematoda abundance, 421 biomass, 414-415 density, 414 distribution in soil, 419-420 trophic function, 414 Index 567 Nitrogen, 231-233 absorption, 165-168 allocation, 169-171 amount in ecosystem, 464 budget, 470-471 caribou urea cycling, 399 content in ecosystems, 460-461 denitrification, 243-245, 473 distribution in biomass, 464-465 immobilization, 246-249 input from precipitation, 239-240 loss in runoff, 240-242 mineralization in soil, 245-249 nitrification of soil, 249-250 transport in soil, 250-252 see also Nutrients Nitrogen fixation, 253, 473 algae, 234-235, 238 bacteria, 258-259 biomass of organisms, 235-236 environmental control, 236-239 Nutrient cycling, 411, 472, 475, 477-478 long-term changes, 479 microorganisms, 331-333 Nutrient-recovery hypothesis, 375-380 Nutrients, 220-221 absorption, 165-168, 477 allocation, 140, 169-171 annual input, 492 contribution by lemmings, 377-378 forage, 356-357 graminoid growth, 150 input from atmosphere, 469 invertebrate biomass, 442 invertebrate production, 441 limit on CO2 uptake, 103 loss, 473 photosynthesis, 109-110, 173, 184 release, 477 reservoir in vegetation, 464 shoot growth, 171-173 soil, 253 transfer within ecosystem, 473-478 turnover rates, 476 uptake, 477 see also Nitrogen, Phosphor- us, etc. Nutrient-transport hypothesis, 406-408 Ordination see Vegetation Organic matter, 331 accumulation, 473 carbon, 459 concentration of invertebrates, 441 Oxygen influence on fungi, 287 microbial metabolism, 307-309 soil, 227 Permafrost, 1, 214, 219-220, 420, 464, 469 Perturbations see Disturbance Phalaropes see Birds Phosphorus, 231, 233-234 absorption, 165-168 allocation, 169-171 amount in ecosystem, 460-461, 464 availability, 90 budget, 470-471 distribution in biomass, 464-465 input, 240 loss in runoff, 242-243 mineralization in soil, 245-249 568 Index transport in soil, 251-252 see also Nutrients Photosynthesis, 82 adapted to Arctic, 102 canopy structure, 131-134 Cj and C4 plants, 102 carbon fixation, 466-467 competency, 127, 138 diurnal course, 116-117 efficiency, 465 evergreens, 203 foliage area index, 132-133 grazing, 119 irradiance, 110-112, 127-128 leaf area, 138 leaf development, 106-107 models, 120-123 mosses, 112-115, 119 nutrients, 184 plant pathogens, 119 photosynthetically active radi- ation (PAR), 78-79, 465 potential, 108, 140 primary production, 71-74 rate, 103, 126 seasonal course, 108, 116-119, 135-138 spatial variability, 109 temperature, 112-115, 123-126, 138 water, 115-116, 128-131 Plants growing season, 4 number of species, 6-7 physiological adaptation, 67 Polygons characteristics, 16-20 decomposer activities, 308 decomposition in troughs, 405 development, 20-21 fungi, 264-266, 268-269, 274, 283 invertebrates, 421 lemming habitat, 344-345, 405 nutrient-transport hypothesis, 406-408 plant succession, 207 snow cover, 41 soils, 22, 219-234, 405 species diversity, 209 types, 16, 20-21 vegetation, 209, 405 Polysaccharides see TNC Population dynamics Dupontia, 176-177 Potassium absorption, 165-168 Precipitation, 31-32, 35 annual average, 52 interannual variability, 35-37, 65 nutrient input, 239-240 Predation model of avian, 370 on lemmings, 337, 360-371 Predators, 410 avian, 360 mammalian, 360 numbers, 362-366 nutrient transport, 408 prey selection, 366 Primary production, 100, 209, 481 aboveground, 69-70 affected by short summer, 467 belowground, 69-70 efficiency, 465 invertebrate, 422-423 limitations on, 140, 464 measurement, 67 rates, 144, 472 seasonal progression, 71-74 species diversity, 199 Production bacteria, 280-281 fungi, 278-280 microbial, 288, 328-330 trophic systems, 335 4 I Index 569 see also Primary production Radiation, 30, 34, 38, 46, 56, 61, 64, 78-84, 481 Radiation balance, 44-49 Respiration, 68, 70 metabolic processes, 159 microbial, 305, 318, 439 model of microbial, 310-324 plant, 467 rate, 145 soil, 466 whole-system, 325-326 Rhizomes carbon content, 459 Roots carbon content, 459 growth, 147-148, 153 nutrient uptake, 158 resistance, 129 root-to-shoot ratio, 155 Rotifers density, 416 Runoff, 53-55, 57, 63, 472 Sandpipers see Birds Short-eared owls predators, 364 Shrubs growth form, 194 Snow cover, 37-44, 203 drifts, 38, 41-42 structures, 39-41 temperature, 43 types, 39 water equivalent, 52, 57 Snowmelt model, 58-61 processes, 56-61 runoff, 472 Snowpack lemming habitat, 372-373 Snowy owls density, 363-364 food consumption, 368 nutrient transport, 408 predators, 360-364 prey selection, 367 Soil acidity, 229 aeration, 227 bacteria, 258, 277, 295-296 biological processes, 220 bulk density, 222-223, 253, 288 carbon, 220-221, 459 cation exchange capacity, 227-229 characteristics, 24 conditions of formation, 219 denitrification, 243-245 description, 21-25 distribution, 21-25 effect of lemmings, 252 freezing, 64, 252 fungi, 269, 276-277, 281-284 immobilization, 245-249 invertebrate distribution, 412, 418-421 major cations, 229-231 map, 23 mineralization, 245-249 moisture, 63, 129, 225-227 nitrification, 249-250 nitrogen, 231-254 nitrogen fixation, 234-239 nutrients, 253-254 orders and suborders, 22 organic matter, 21, 90, 220-222, 232, 245, 253, 411 oxygen, 168 phosphorus, 231-254 porosity, 223-224 respiration, 331 temperature, 37, 62-63, 74, 570 Index 86-87 texture, 224 thaw, 25, 57, 87-90 water potential, 130 weathering, 252 Soil invertebrates see Invertebrates Solar radiation see Radiation Species diversity growth forms, 199-200 polygons, 208 Standing crop, 69-71 aboveground vascular, 459 belowground vascular, 463 graminoids, 140 meadows, 458 reduction by lemmings, 379 Succession, 206-217 Sugars see TNC Sulfur bacteria, 259 ordination axes, 190 Tardigrades, 416 Temperature air, 4, 32-34, 482 canopy, 84 ground, 51 influence on fungi, 284-285, 290 interannual variability, 35-36, 65 inversion, 38 leaf, 85-86 lemming habitat, 372 microbial activity, 298-299 microbial energetics, 299-302 microbial substrate utilization, 302-305 optimum, 142 photosynthesis, 123-126 regime, 31 rhizome growth, 146 root growth, 148 snow, 43 soil, 37, 42, 62-63, 420 summer, 34 Thaw equation, 61-62 season, 36 Thaw lake cycle, 15-16, 479 disturbance by, 210 succession, 206, 212-214 Thermoregulation birds, 453 lemmings, 350 Tillers Dupontia, 175 Tipulidae abundance, 416 consumption by birds, 453-456 energetics, 433-434 growth rate, 429-430 life cycle, 424-428 TNC (total nonstructural carbo- hydrate) concentration in Dupontia, 148-151 frost tolerance, 151 graminoids, 184, 353 growth respiration, 159-161 nutrients, 155 storage, 150 translocation, 161-162 Translocation, 173 carbon, 466 energy cost, 161 low temperature, 156 photosynthate, 152, 465 root growth, 154 see also Allocation Transpiration, 79, 94-97 Transport soil nutrients, 250-252 Trophic systems detritus-based, 335-336 Index 571 herbivore-based, 335-336 Tundra definition, 1 distribution, vii long-term changes, 478-481 Turnover rates, 193 Ungulates, 337 Vegetation, 8 characteristics, 25-29 growth forms, 195-199, 217 ordination axes, 188, 190 patterns, 186-206 recovery, 210, 214 sampling site, 188 succession, 479 turnover rates, 193 types, 26-29, 191-193, 217 Vertebrate fauna, 8-9 see also Arctic foxes, Birds, Caribou, Lemmings, Weasels Water balance, 51-56, 61 Water relations density, 365 mosses, 93-97 simulation in plants, 97-100 vascular plants, 91-93 Weasels food consumption, 369-370 predators, 360-361, 365 prey selection, 367 Weathering, 469 Yeasts biomass distribution, 270-271 numbers, 261 V 1 35