US/IBP SYNTHESIS SERIES 1 13 LIMNOLOGY OF TUNDRA PONDS Barraw, Alaska John E. Hobbie den, Hutchinc Ross.Inc ^ on LIMNOLOGY OF TUNDRA PONDS 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 Foundation. Views expressed in this volume do not necessarily represent those of the National Academy of Sciences or the National Science Foundation. Volume 1 MAN IN THE ANDES: A Multidisciplinary Study of High-AUitude Quechua / Paul T. Baker and Michael A. Little 2 CHILE-CALIFORNIA MEDITERRANEAN SCRUB ATLAS: A Comparative Analysis / 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 / Gordon 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 / Harold A. Mooney 6 CREOSOTE BUSH: Biology and Chemistry of Larrea in New World Deserts / T. J. Mabry, J. H. Hunziker. and D. R. DiFeo. Jr. 7 BIG BIOLOGY: The US/ IBP/ W. Frank Blair 8 ESKIMOS OF NORTHWESTERN ALASKA: A Biological Perspective/ Paul 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 / Robert L. Edmonds 11 WATER IN DESERT ECOSYSTEMS / Daniel D. Evans and John L. Thames 12 AN ARCTIC ECOSYSTEM: The Coastal Tundra at Barrow, Alaska / Jerry Brown, Philip C. Miller, Larry L. Tieszen, and Fred L. Bunnell 13 LIMNOLOGY OF TUNDRA PONDS: Barrow Alaska / John E. Hobbie 14 ANALYSIS OF CONIFEROUS FOREST ECOSYSTEMS IN THE WESTERN UNITED STATES / Robert L. Edmonds 15 ISLAND ECOSYSTEMS: Biological Organization in Selected Hawaiian Communities / Dieter Mueller- Dombois, Kent W. Bridges, and Hampton L. Carson US/IBP SYNTHESIS SERIES 1 13 LIMNOLOGY OF TUNDRA PONDS Barrow, Alaska Edited by John E. Hobbie Marine Biological Laboratory Dowden, Hutchincon (^ Ross, Inc. Stroudsburg Pennsylvania Dedicated to three scientists who shaped and supported ecological research at Barrow, Alaska, during the !950s and 1960s: Max Britton, Max Brewer, and Frank Pitelka. Copyright © 1980 by The Institute of Ecology Library of Congress Catalog Card Number: 80-26373 ISBN: 0-87933-386-3 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 photo- copying, recording, taping, or information storage and retrieval systems — with- out written permission of the publisher. 82 81 80 1 2 3 4 5 Manufactured in the United States of America Library of Congress Cataloging in Publication Data Hobbie, John E. Limnology of tundra ponds, Barrow, Alaska. (U.S./ IBP synthesis series ; v. 13) Bibliography: p. Includes indexes. 1. Limnology — Alaska — Barrow. 2. Pond ecology — Alaska — Barrow. 3. Tundra ecology — Alaska — Barrow. I. Hobbie, John E. II. Series. QH105.A4H62 574.5'2632'097987 80-26373 ISBN 0-87933-386-3 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 fifty-eight nations taking part in the IBP during the period of July 1967 to June 1974, the United States organized a number of large, multidisciplinary studies pertinent to the central IBP theme of "the biological basis of productivity and human welfare." These multidisciplinary studies (Integrated Research Programs) directed 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 accpmplish 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: U.S. /IBP Gabriel Lasker Robert B. Piatt Frederick E. Smith W. Frank Blair, Chairman Preface This book is a report of investigations of several small ponds on the arctic tundra near Barrow, Alaska. The main study, which ran from 1971 through 1973, was funded from three sources: The National Science Foundation, the State of Alaska through the University of Alaska, and individual companies and members of the petroleum industry. 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 Program). The U.S. Tundra Biome Program was under the overall direction of Jerry Brown of the U.S. Army Cold Regions Research and Engineering Laboratory and consisted of aquatic and terrestrial sections. A companion volume to this reports the findings of the terrestrial projects (Brown et al. in press). The principal investigators of the aquatic projects were: Vera Alexander, University of Alaska Robert J. Barsdate, University of Alaska Donald A. Bierle, Sioux Falls College James N. Cameron, University of Alaska Raymond D. Dillon, State University of South Dakota John E. Hobbie, North Carolina State University C. Peter McRoy, University of Alaska Michael C. Miller, University of Alaska Raymond G. Stress, The University (SUNY) at Albany. Other scientists who took part in the project were Staffan Holmgren (Uppsala University), Tom Fenchel (Aarhus University), Stanley Dodson (University of Wisconsin), John Kelley (University of Alaska), Patrick Coyne (U.S. Army, CRREL), Ralph Daley (North Carolina State University), Richard Prentki (University of Alaska), Tor Traaen (Norwegian Institute for Water Research), Donald Stanley (North Carolina State University), and Jawahar Tiwari (North Carolina State University). Additional information on the macrobenthos came from a study in 1975, 1976, and 1977 carried out by Samuel Mozley and Malcolm Butler (University of Michigan) which was funded by the U.S. Department of Energy. Vll viii Preface There are a number of possible approaches to the study of the ecology of tundra ponds. We concentrated first on measuring the fluxes of carbon, nitrogen, and phosphorus through the ecosystem. Next, we used a variety of manipulations, ranging from changed light conditions for plankton in a small bottle to an increase in phosphate in a whole pond, to investigate the controls of various processes. While the field work was going on, we also constructed a mathematical model of the ecosystem. This left little time for detailed studies of the ecology of individual species although several dominants, such as Daphnia middendorffiana and a Chironomus sp., were examined. Most effort was put into nutrient cycling studies and into investigations of the lower trophic levels. Some areas, such as the control of zooplankton species composition or the physiology of individual species of algae, were not well studied. Most of this book was written during the summer of 1974, and was later edited for consistency of style and overall integration. The portion of Chapter 7 on the insect larvae was completely rewritten in 1978, and, by the author's request, the section of Chapter 6 by R. Stross was not edited. The book is organized in a conventional fashion with the physical and chemical information first followed by the descriptions of the primary producers, secondary consumers, etc. Each chapter ends with an extensive summary; a good idea of the important parts of the limnology of the pond can be gained from these. Chapter 1 consists of a summary of the conclusions of the overall study but only those conclusions that are most interesting to an ecologist. In this way, we attempt to answer the question, "What new things did you discover?" Excellent logistics and laboratory support were provided by the Office of Naval Research through its Naval Arctic Research Laboratory at Barrow. Two former directors of this laboratory. Max C. Brewer and John Schindler, deserve particular credit for facilitating this support. Gene E. Likens of Cornell University reviewed the first draft of this book and Colleen M. Cavanaugh and Kate Eldred of the Marine Biological Laboratory provided valuable editing services. Harold Larsen, USA CRREL, prepared the illustrations and the CRREL editorial staff under the direction of Stephen Bowen provided valuable assistance throughout the preparation of this book. Kathleen Salzberg, editor of Arctic and Alpine Research, provided assistance in editing the references. Typesetting was done by The Job Shop, Woods Hole, Massachusetts, and by Donna Murphy of CRREL. Special thanks go to George Llano, formerly of the National Science Foundation, for his administrative guidance and his sympathetic understanding of the difficulties of running a large scientific study in the Arctic, and to Jerry Brown for his assistance in the field aspects of the program and in publication of this volume. John E. Hobbie Contents Foreword v Preface vii List of Contributors xii 1^ 1 : Major Findings J. E. Hobbie Introduction, 1 Description of the Ponds, 2 Flux of Carbon, 5 Flux of Phosphorus, 10 Flux of Nitrogen, 1 3 Effects of the Arctic Environ- ment, 13 Modeling, 15 ^ 2: Introduction and Site Description 19 J. E. Hobbie The Tundra Biome Project, 19 Limnology of the Arctic, 2 1 Geography and Geomorphology, 25 Biology, 36 Energy and Nutrient Cycling, 42 Summary, 46 3: Physics 51 M. C. Miller. R. T. Prentki, and R. J. Barsdate Geography, 51 Temperature Studies, 54 Hydrology, 62 Light, 70 Currents, 71 Summary, 72 . IX X Contents ^^ 4: Chemistry 76 R. T. Prentki, M. C. Miller. R. J. Barsdate, V. Alexander, J. Kelley, and P. Coyne Sediments, 76 Major Ions, 79 Trace Metals, 86 Carbon Dioxide Systems, 87 Oxygen, 97 Nitrogen, 100 Phosphorus, 115 Control of Phosphorus, 136 Organic Carbon, 151 Summary, 169 5: Primary Producers 179 V. Alexander, D. W. Stanley, R. J. Daley, and C. P. McRoy Phytoplankton, 179 Epipelic Algae, 193 Factors Controlling Algae, 201 Rooted Aquatic Plants, 224 Factors Controlling Rooted Aquatic Plants, 234 Summary, 243 6: Zooplankton 251 R. G. Stross, M. C. Miller, and R. J. Daley Communities, Life Cycles, and Production, 251 Control of Zooplankton Production (I), 274 Control of Zooplankton Production (II), 288 Summary, 293 7: Macrobenthos 297 M. Butler, M. C. Miller, and S. Mozley Introduction to Arctic Benthos, 297 Chironom- idae Studies, 303 Tadpole Shrimp, 323 Summary, 335 8: Decomposers, Bacteria, and Microbenthos 340 J. E. Hobbie, T. Traaen, P. Rublee, J. P. Reed, M. C. Miller, and T. Fenchel Bacteria, 340 Fungi, 356 Decomposition of Macrophytes, 357 Sediment Respiration, 363 Microbenthos, 372 Summary, 384 Contents xi 9: Oil Spill Effects 388 R. J. Barsdate, M. C. Miller. V. Alexander, J. R. Vestal, and J. E. Hobbie Introduction, 388 Physical and Chemical Measurements, 391 Biological Measurements, 395 Summary, 405 10: Modeling 407 J. L. Tiwari. R. J. Daley, J. E. Hobbie. M. C. Miller, D. W. Stanley, and J. P. Reed Modeling in the Aquatic Program of the Tundra Biome, 407 Whole Systems Models, 409 Benthic Carbon Flow Model, 430 Planktonic Carbon Flow Model, 434 Results and Discussion, 447 Summary, 456 References 457 Taxonomic Index 493 Subject Index 499 List of Contributors Vera Alexander Institute of Marine Science, University of Alaska, Fairbanks, Alaska 99701 Robert J. Barsdate Institute of Marine Science, University of Alaska, Fairbanks, Alaska 99701 Malcolm Butler Department of Ecological and Evolutionary Biology, Division of Biological Sciences, University of Michigan, Ann Arbor, Michigan 48109 Patrick Coyne USDA/SEA/AR; Southern Plains Research Station, 2000 18th Street, Woodward, Oklahoma 73801 Ralph J. Daley Canada Center for Inland Waters, Pacific Region, 4160 Marine Drive, West Vancouver, British Columbia, Canada V7V 1N6 Thomas Fenchel Institute of Ecology and Genetics, Ny Munkegade, Aarhus Universi- ty, Aarhus C, Denmark John E. Hobbie Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 John K el ley Naval Arctic Research Laboratory, Barrow, Alaska 99623 C. Peter McRoy Institute of Marine Science, University of Alaska, Fairbanks, Alaska 99701 xiu xiv List of Contributors Michael C. Miller Biological Science Department, University of Cincinnati, Cincinnati, Ohio 45221 Sam Mozley Department of Zoology, North Carolina State University, Raleigh, North Carolina 27650 Richard T. Prentki Department of Biological Sciences, University of Nevada, Las Vegas, Nevada 89154 James P. Reed Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 Parke A . Rublee CBCES, Smithsonian Institution, P. O. Box 28, Edgewater, Maryland 21037 Donald W. Stanley Institute of Coastal Studies, University of East Carolina, Greenville, North Carolina 27834 Raymond G. Stross Department of Biology, State University of New York, Albany, New York 12203 Jawahar Tiwari Department of Surgery, School of Medicine, University of California, Los Angeles, California 90024 Tor Traaen Institute of Water Research, P. O. Box 260, BUndern, Oslo 3, Norway J. Robie Vestal Department of Biological Science, University of Cincinnati, Cincin- nati, Ohio 45221 1 Major Findings John E. Hobbie INTRODUCTION Studies in the extreme environments of mountains, tropics, and the Arctic have long been an important part of ecological research. Apart from the stimulation and enjoyment of visiting new places, ecologists have compared these extreme habitats with one another and with temperate habitats in order to test hypotheses about general principles. This approach of comparative natural history requires a large body of data collected from many habitats; both descriptions and a good understanding of processes are required. The data from extreme environments are especially valuable as they extend the range of important variables and may even allow analyses of the effect of certain factors that always vary together in temperate regions. The IBP study of arctic ponds reported in this book is primarily a description of the habitat, the biota, and the processes by which organisms interact with other organisms and with their physical and chemical environments. In the report, the comparative aspects of the study have been deliberately de-emphasized, as constant reference to temperate and tropical lakes would have quickly doubled the size of the book. The value of the study in this comparative sense will become apparent later, when this study is referred to to find out what controlled photosynthesis, how rapidly a sedge leaf decomposed, or what the community structure was in an arctic pond. In addition to the comparative importance of the arctic ponds, there are certain advantages to investigating aquatic processes in the Arctic. For example, low diversity of the higher plants and animals allows cohorts and age classes to be identified and followed through time; this simplifies productivity measurements. In some groups, such as most of the zooplankton, there may be only a single generation each year which also greatly simplifies growth measurements. The low diversity also permits a more complete study to be carried out with fewer scientists but does not, of course, make the study of an individual process any easier. There are other tactical advantages to arctic research. First, the ponds freeze completely in mid-September, so they need to be studied for only 3 months a year (which fits into academic schedules quite well). During the field research months the scientists were working in crowded laboratories in an isolated location where there were few distractions from 1 2 J. E. Hobbie beaches, families, or television. The resulting intense interactions and scientific excitement could only be maintained for a month or so, but helped immensely to stimulate creativity and to integrate the various projects. Description of the Ponds Small ponds formed on old lake beds are abundant on the flat coastal plain of northern Alaska (Figure 1-1). A number of these ponds, several kilometers from the Naval Arctic Research Laboratory at Barrow, were studied for several years to improve our understanding of the controls of aquatic populations and processes that operate in this extreme environment. The ponds are small, only about 30 x 40 m, and shallow, up to 0.5 m deep (Figure 1-2, 1-3). Each pond is surrounded by wet tundra, mostly low grasses and sedges, and is cut off from adjoining ponds by a network of 71-20'. 7\'-\5'- 156° 50' 156-40' 1 56° 30' FIGURE 1-1. Location of IBP Tundra Biome Project, showing the Naval Arctic Research Laboratory, the village of Barrow, and the re- search sites (cross-hatched area). Major Findings 3 small ridges pushed up by the growth of underlying ice wedges. The total area enclosed by the ridges is about double the pond area. Despite the minuscule drainage basins and the desert-like levels of precipitation (12 cm annual, 50% falls as rain in June through September), the surrounding tundra is often saturated and the ponds do not dry up. Water flows from one basin to another for only a few days during the spring runoff. There is no belowground water movement from basin to basin or into the sediments because of the underlying permafrost. From late September until mid-June, the ponds and their underlying sediments are solidly frozen. Melting of the ice in the ponds occurs over a few days in the spring and water temperatures can reach as high as 16°C any time thereafter. Thawing of the sediments continues throughout the summer until 30 cm are thawed. The ponds are so shallow that the water temperature can change as much as 10°C per day in response to sunlight, air temperature, and wind. June, July, and August are cool (mean air temperature is 2.8°C), cloudy (83% cloud cover), and windy (an average of 6.1 m sec ~ ^) so the mean water temperature is low, around 6°C. FIGURE 1-2. The intensively studied ponds near Barrow, Alaska. A small field lab is at the upper left of the picture and experimental sub- ponds in Pond B are at right. An aerial cable car is suspended above the subponds. 4 J. E. Hobbie FIGURE 1-3. To avoid disturbing the water and sediments of Pond B, an investigator takes samples from a cable car. The pond waters contain small amounts of salts and have a pH around 7.3. The light, flocculant sediments are made up of 80% organic matter. The dominant primary producers of the ponds (Table 1-1) area sedge {Carex aquatilis) and a grass {Arctophila fulva) which live in the shallow margins of the pond and cover 30% of the surface (Figure 1-4). Benthic microalgae, mostly diatoms and blue-greens, are also important producers but their numbers are kept low by the continual mixing of the upper few centimeters by animals, which keeps most of the algae away from light. Algae in the water above are all small flagellated nannoplankton, especially greens and chrysophytes. Their total productivity is low and they are heavily grazed by zooplankton such as Daphnia and fairyshrimp. In turn, these herbivores are preyed upon by predaceous zooplankton {Cyclops, Heterocope) but there are no important vertebrate predators (although shorebirds do feed on zooplankton). The leaves, rhizomes, and Major Findings TABLE 1 - 1 Annual Production of Tundra Pond Communities Type of community Production gC m yr Method of measurement Phytoplankton 1.1 Benthic algae 8.4 Macrophytes 16.4 Zooplankton 0.20 Macrobenthos 1.65 Planktonic bacteria 0.01 Benthic bacteria 8.6 Benthic bacteria 4-16 Benthic bacteria 20 Protozoa 0.3 Microbenthos 0.2 carbon-14 carbon-14 biomass changes biomass changes biomass changes biomass changes biomass changes CO2 evolution in cores CO2 exchange with atmosphere biomass changes, lab growth rates biomass changes, lab growth rates roots of the grasses and sedges enter the detritus food chain as there are no grazers on the Hve plants. Most of the detritus is mineraHzed by bacteria and fungi but some is consumed by chironomid larvae, the dominant animals of the sediment. These larvae eat a few percent of the bacteria and algae per day as do the microfauna of nematodes, harpacticoid copepods, and protozoans. Flux of Carbon The measurement of the flux of carbon is a useful way to begin an ecosystem study, as all the important elements can be identified. The techniques we used for the carbon flux and standing stock are standard ecological measurements such as '''C for the primary productivity of the algae, biomass changes for rooted plants, CO 2 partial pressures by gas analysis to obtain water-air exchange, laboratory respiration studies of larger organisms, and acridine orange direct counts for the bacteria. The only component of the biota not measured was the fungi. A single measurement indicates that in sediments the mass of fungal hyphae is about equal to the mass of bacteria. The rooted plants in the pond provide most of the input of organic carbon (Figure 1-5, Table 1-1). They release dissolved organic carbon into the water, release a large quantity of CO 2 via root respiration, and add dead leaves, stems, and roots to the detrital pool. Once it reaches the sediments, a leaf of Carex takes 4 years to decompose. One reason for this rather long life-after-death is the lack of shredders in the pond ecosystem. Another reason is the 9 months of cold storage each year (however, freezing and thawing does mechanically damage the leaves). When J. E. Hobbie 100 FIGURE 1-4. A cross section of a typical pond. decomposition is calculated as percent per month of open water, then it appears that the rate is very similar to temperate rates. Algal photosynthesis in the sediment surface is also an important input of organic carbon. Although photosynthesis occurs only in the top 2 mm, the algal cells were found throughout the upper 5 cm and deeper. The buildup of these benthic algae is prevented by the downward mixing of sediment and algal cells by the animals; in the absence of this mixing, an algal mat would develop, which would have a very high productivity. Algal photosynthesis in the water column is extremely low, as low as any in the world. This is not caused by low numbers of cells, as millions per liter are always present. These are all very small cells, however, and their mass is also small. Thus, in temperate lakes the algal mass is 100 times greater than the bacterial mass; in the arctic ponds the algal mass was about equal to the bacterial mass (Note: The bacterial mass is the same in both systems). These planktonic algae also show a paradox found in other extremely oligotrophic systems, such as the Sargasso Sea. This is the rapid turnover of cells (the amount of carbon produced per day equals the carbon of the cells) in oligotrophic waters. Grazing by zooplankton, especially Daphnia, is likely responsible for much of the rapid turnover in the ponds. The result of the distribution of primary production, low in the plankton but high in the macrophytes and benthic algae, is a shift of Major Findings Air -0. j- 1 . 1 U. 1 I .i CO , 400 O.h 1 — f 0.7 0.8 I DOC 12bh — r Zooplankton 20 .0.1 liat- tL-r ixi 1 .4 -14.5- -12.1- Det r Itus 48 — v./- 50 Water Algae 700 -T 0.2 L Bacteria 1500 n DOC 1500 -250- Protozoa | 36 C CO 01 «-» o ex C 3 5C tn o I- " o o c — e tT3 - _D — ' ^^ 3 0.— 770 512 * CO, 460 400 I— 300- New Detritus 24,000 Old Detritus 3,816,000 Micro- metazoa 35 Sediment _^3^__J -284- 100 -• IS) o u o O CD O - cc o CO 380 2 50 CO^ 400 n Animals 1350 16- ■170- FIGURE 1-5. Carbon flux through a typical tundra pond. Measurements were made on 12 July 1971. On this date, the average depth of the water was 10 cm and the depth of the sediment was taken as 5 cm. Units of the standing crop (in boxes) are mg C m~^ and transfer rates (arrows) are mg C m'^ day'\ organic carbon to the sediments where it enters the detritus food web (Figure 1-5). Here, the abundant detritus is a large reservoir of food for animals while the decomposition of the organic matter provides a steady supply of nutrients for algae. The contrast between the sediment and water 8 J. E. Hobbie systems is dramatic; the living mass of organisms is more than 150 times greater per square meter of sediment than of water, and the activity rates (e.g., respiration) reflect the same ratio. In spite of the relatively high sediment activity, most of the detritus pool, nearly 4 kg C m~\ is not being broken down. Instead, the food for the biota comes from recently formed detritus (about 0.02 kg C m Mn Figure 1 -5). In the water column, there is a similar large quantity of carbon, the dissolved organic carbon (DOC), that consists of a large pool of inactive carbon and a much smaller pool of rapidly-cycling carbon. Some of this rapidly-cycling pool of DOC comes from the sediment, as the mass and activity of bacteria is quite high considering the low primary production of the planktonic algae. The detritus, algae, and bacteria support a large standing crop of zooplankton grazers, a crop that is much larger than algae alone could support. Actually, the relationship between the zooplankton and detritus may be more complicated than this. We observed that from year to year the amount of planktonic algae and bacteria remained about the same (5 to 10 ng C liter "^) but the amount of detritus fluctuated from 300 to 1400 Mg C liter ~\ Zooplankton production was highest in years when the average amount of detritus was lowest and vice versa. This could be cause- and-effect but it is impossible to tell if the high detritus loads prevented the zooplankton from harvesting very much of the nutritious algae and bacteria (blocking) or if the high numbers of zooplankton removed the detritus. It is also possible that the zooplankton excreted enough phosphorus back to the water to increase the phytoplankton production. Carbon dioxide moves rapidly from the water into the air. In fact, an amount of dissolved CO 2 equal to that in the water is replaced each day. The flux of CO 2, appears to balance the primary production but in spite of the intensive study, we could not say whether or not annual respiration equaled photosynthesis. At best, our measurements were only within 20% of the true value and an accumulation of only 10% of the total primary production each year would easily account for the organic sediments of the pond. (Ten percent each year is 5 cm of sediment in 400 years.) In the pond ecosystem it was obvious that grazing food chains are unimportant relative to the detritus food chain (Table 1-1). In the sediments, the detritus is either eaten directly by animals or is attacked first by microbes. Our evidence for direct utilization comes mostly from studies of the energy requirements of the chironomid larvae (the "animals" in Figure 1-5). At the rate of particle ingestion that we measured, the larvae had to be digesting mostly detritus. Previous workers postulated that the animals were obtaining enough energy by stripping the microbes from the detrital particles. We actually measured the quantity of bacterial and algal biomass that is included in the detritus and found it to be only 0.06% of the total carbon (Figure 1-5). This amount of carbon is 0.3% of the organic carbon requirement of the larvae. Although these animals may select microbe-rich particles or locations for feeding, they Major Findings Carnivorous C i 1 i a t e s U. 1 mg C Bac terivorous Ciliates 0.2 mg C Zoof lagellates 35 mg C T 0.05 mg C T Bac teria 1500 mg C 15 mg C ■'-'0.5 mg C Algivorous Ciliates 1 mg C 0.15 mg C Algae 700 mg C MicromeCazoa 35 mg C FIGURE 1-6. Carbon flow through the protozoans and micrometazoans of the sediment of a tundra pond. Units are mg C m'^ day~^ for the fluxes and mg C m'^ for the standing crop. would have to be extraordinarily selective to meet their energy requirements from bacteria and algae alone. The bacteria which break down detritus are (along with the benthic algae) the base of another food chain of protozoans and micrometazoans such as nematodes. The feeding rates and production of these animals have not previously been studied in the field. Here, these animals grazed only 1 to 2% of the bacteria and algae per day (Figure 1-6). This seems small yet represents 20% of the bacterial production and 5% of the algal production each day; thus, the small animals may control the bacteria to some degree. The protozoans and bacteria interact in other ways as well. It has long been known that decomposition proceeds faster, and bacteria are more active, when grazing animals are present. One hypothesis has been that nutrients were rapidly released by the grazers and that this release allowed higher microbial activity. This hypothesis was tested in an experiment which investigated the rate of cycling of phosphorus-32 in small flasks containing Carex, bacteria, and one species of protozoa (Barsdate et al. 1974). When the protozoan was present, the bacterial biomass was lower, the bacteria were more active, and phosphorus was taken up faster by the bacteria (1 .67 vs. 0.25 pg P cell ~ ' hr " '), than when the protozoan was not present. Yet, only a few percent of the phosphorus actually cycled through the protozoans. Thus, direct release of phosphorus by the protozoans did not affect the bacteria and the hypothesis was disproved. It is possible that the bacteria are kept in a phase of rapid growth by the grazing and that it is this rapid grovyth that is responsible for the faster decomposition. 10 J. E. Hobbie Flux of Phosphorus Phosphorus enters the ponds in rainfall (8 ng P liter"') and in overland flow. The quantities entering a pond are small and are equaled by the losses. We did measure an annual loss of 0.7 mg P m^^ but this amount is minuscule relative to the 25,000 mg P m "^ found in the top 10 cm of sediments. In spite of the large amounts of phosphorus present, the concentrations of inorganic phosphorus in the waters of the pond and in the interstitial water of the sediments is always extremely low, between 0.00005 r 0.4: Phytoplankton 0.027 0.015 0.033 0.0075 Sestonic Bacteria 0.14 0.13 0.074 Zooplankcon 1.4 t 0.084 0.07 O.U Sestonic Detritus 0.17 0.42 r 1.0 0.42| I 0.42 Woter 0.11 0.72 1.1 1.2 0.73 1.2 Benthic Algae 17 0.007 3 oTZT _1_ 3.G Benthic liacteria 56 -1 0.002fi L_ 7.2 Benthic Animals 22 0.47 -a (N o i T 11 _i_ 1.2 0.39 Sediment 10 cm 5 cm Sediment and Detritus 10000 ^ -1 FIGURE 1-7. Phosphorus flow diagram in a tundra pond for 12 July 1971. Units are mg P m'^ and mg P m'^ day'\ Fluxes were measured whenever possible (Prentki 1976) or were based on the carbon flux data (Figure 1-5). Major Findings 11 0.001 and 0.002 mg P liter'. This is the form of phosphorus that is available to algae and higher plants; their primary production can be enhanced over several weeks by adding phosphorus to a pond. Even in the fertilized ponds, however, the concentrations of inorganic phosphorus rapidly decline to 0.001 to 0.002 mg P liter '. Why are these phosphorus concentrations so low and why are there such small changes over the summer? A part of the answer is that the dissolved reactive phosphorus (DRP) cycles very rapidly in the ponds (Figure 1-7). For example, there is 0.14 mg P m~^ in the water on the day illustrated in the figure while the bacteria and algae take up 5.8 mg P m ^ day " \ At the same time, there is also a transport of 0.73 mg P m ^ into the DRP pool of the interstitial water; the DRP thus turns over 50 times per day in the ponds. During the rapid turnover of the small amount of DRP in the water, the large quantities of phosphorus in the sediment turn over very slowly and actually buffer the whole system. The other part of the answer lies in the chemical properties of the sediment. When DRP enters the pond, it quickly moves to the sediment where much of it is sorbed onto a hydrous iron complex. The concentration of DRP and the release rate of the sorbed phosphorus are controlled by a chemical equilibrium; ponds with different amounts of iron and inorganic phosphorus in the surface sediments will have different 10.0 ouu ' i 1 1 ' / 1 1 v 8 O "o. 600 t— — 4. a. o / ° c 1.0 o> o a. O rt/ 'Z- „:.400 o/ - o / — 3 0. o / ■o « 0 / 0 o S 200 0/ o / ° ° qI o °/o / , 1 1 1 1 1 - £ 0.1 8 0 2 4 6 DRP, fj.q P 1 ite r-' FIGURE 1-8. Oxalate extractable phosphorus in the sediments of five tundra ponds of similar origin plotted against the dis- solved reactive phosphate (DRP) of the overlying water column. 400 500 Phosphate Sorption Index 600 FIGURE 1-9. Algal photosyn- thesis in the water column of a series of tundra ponds plotted against the phosphate sorption index of the underlying sediments (9 August 1973). (Data are from Prentki 1976.) 12 J. E. Hobbie amounts of DRP in the water. In a series of intensively-studied ponds (Figure 1-8), the concentration of DRP in the water column could even be predicted from a single measurement of the sediment phosphorus that could be extracted with oxalate. In the same ponds, the oxalate- extractable phosphorus appeared to be directly related to the photosynthesis rate of planktonic algae (Figure 1-9). Thus, we conclude Air FIGURE 1-10. Nitrogen standing crop and fluxes in the water and sedi- ments of a tundra pond on 12 July 1971. Units are mg N nr- for the con- centrations and mg N nr' day' for the fluxes. The amounts and flux rates were measured whenever possible; some are calculated from the data in Figure 1-5. Major Findings 13 that chemical reactions in the surface sediments, especially those reactions involving iron, set the concentration of DRP in the water and in this way control the productivity of the ponds. Flux of Nitrogen The main inputs of nitrogen to the pond came from the rain water (1 1.5 mg inorganic N m ' yr ') and from nitrogen fixation (28 mg N m ^ yr"'). The ponds appear to accumulate some nitrogen each year but the total, somewhat less than 80 mg N m^^ yr"\ is very small compared to the 38,400 mg N m~^ stored in the top 5 cm of sediment. In the water column, ammonia was more abundant than nitrate, 20 to 40 Mg NH3 N liter ' vs. 2 to 13 Mg NO3-N liter \ Uptake in the plankton was slow (Figure 1-10) so that turnover times ranged from 30 to 100 days for the inorganic nitrogen. Measurements with ''NH3 indicated that the rate of supply of ammonia from within the water column was high enough to replace the NH 3 in 6 to 48 hr. As expected, we found no evidence of nitrogen limitation upon the primary production of the algae in the pond. When the uptake of nitrogen was used to calculate primary productivity, by taking the ratio of C uptake to N uptake as 100 to 12, the results exactly matched the ^"C primary productivity measurements. In the sediments, the interstitial water contained high amounts of ammonia except where plant roots were present. For example, there were 0.7 to 2.7 mg NH3-N liter"' in the sediments in the plant-free center of the pond but only 0.01 to 0.08 mg NH3-N liter"' inside a Carex bed at the pond edge. Based upon production calculations, the Carex may turn over all the ammonia each day. Despite this relatively high rate of removal, the Carex appears not to be limited by nitrogen concentrations. The only evidence for a limitation by nitrogen was that nitrogen fixation by sediment algae began when the ponds were continually fertilized with phosphorus. It is likely that the algae were phosphorus- limited; when excess amounts of P were added, the uptake rate of both nutrients increased and eventually the N became limiting. At this point, the blue-green algae gained a competitive advantage by fixing nitrogen. Effects of the Arctic Environment The annual primary production of the ponds is low, but this is largely a result of the short ice-free season. When compared to the daily production of other ecosystems the ponds are reasonably productive. Thus, food supply is adequate in spite of the low temperatures. There is, of 14 J. E. Hobbie course, a general slow-down of metabolism because of the cold temperatures and this will affect all ecosystem processes from decomposition to predation. Yet, low temperatures (2 to 8°C) are also found in all temperate ponds; in fact, a majority of months will be cold- water months. The unique properties of arctic ponds are: (1) they never warm, (2) they are frozen for 8 1/2 or 9 months of the year, (3) there is continuous light from late April until mid-August (although the intensity does vary greatly over 24 hours). The smallest life forms of the ponds do not seem to have any special adaptations to the Arctic. Bacteria, for example, are just as abundant in the Arctic as in the water and sediments of any temperate pond; their activity is low but it is about the same as for a temperate pond in the spring when the water is cold. Phytoplankton species are almost identical to the species found in temperate ponds in the spring. There is not even a reduction of species number as 105 species were found in the ponds. These same species, by the way, are found throughout the world and even reach the Antarctic. The physiology of the algae was also normal except that at low temperatures photosynthesis was strongly inhibited by high light levels. This is not adaptive; it may be a result of the low temperatures slowing the rate of repair of chlorophyll. It has been suggested that photosynthesis is less affected than is respiration by low temperatures and for this reason, biomass production would be very efficient in the Arctic. We were not able to measure algal respiration or biomass changes very well so we cannot say if this is true. Protozoans were also found to have the same species and total abundance as temperate pond communities. Some forms that do not have resistant resting stages might be absent, for so little is known of protozoan life history that this could not be detected. Thus, it was surprising to find Paramecium in the pond, for this species has never been known to form resistant cysts. Mttazoans are affected in. a number of ways by the arctic environment. The obvious way is by exclusion of some forms because of physiological limitations. Amphibia and sponges are absent while Hemiptera, Odonata and Megaloptera are rare. Ephemeroptera, Trichoptera, and Coleoptera are represented by only a few families or genera. Another way that organisms are affected is by exclusion from certain habitats. For example, fish are found in deep lakes but not in ponds or lakes shallower than 2 m; they are excluded by the 2-m-thick ice cover in arctic fresh water. This absence of fish allows large zooplankton to exist such as the large Daphnia middendorffiana which reaches 3 mm in length and the fairyshrimp which can be 20 mm long. These large animals, in turn, affect the species composition of the algae. We first noticed this effect when all the zooplankton were killed in a pond by the addition of oil. The same shift in the dominant algae, a replacement of the Rhodomonas by Uroglena, also occurred when we removed the zooplankton by net. Major Findings 15 Metazoans which are found in arctic ponds do not appear to have any particular adaptations to arctic conditions because the same forms are often found in temperate ponds as well. Rather, many of the adaptations or abilities they already possess that permit them to survive freezing or other stresses permit survival in the Arctic as well. For example, in temperate lakes, cladoceran zooplankton overwinter as diapausing eggs or embryos and the cyclopoid copepods overwinter as diapausing subadults (copepodids); the same species has a copepodid diapause stage in temperate lakes where it moves into the sediment for several months. They utilize the same methods for longer periods in the Arctic. Many of the arctic forms of midges have adults which have reduced wings. After hatching, the adults move about on the water surface rather than make the typical laying swarms. This behavior keeps them from being blown away by the constant winds. The life histories of many higher animals are affected by the low temperatures. Zooplankton grow so slowly that there is time for but a single generation per year in most species. This puts a very specific upper limit on annual production of zooplankton because the number surviving the winter may be the single most important factor determining annual production. Chironomids, which live in the top several centimeters of sediment where the annual average temperature in the summertime is only about 3°C, are most affected by the environment. They grow very slowly; one species of Chironomus takes 7 years to pass through the four instar stages. In spite of these difficult conditions for growth, the number of midge species found in the ponds, around 35, is not very different from the number in a temperate pond. Many of these species have been found only in the Arctic or only in the Barrow area. The sedge, Carex aquatilis, that grows in the pond is well adapted to a variety of habitats from dry meadows to the shallow water. Perhaps the most obvious adaptation to the arctic environment is its ability to begin growth when the air temperature is still close to 0°C. It also takes up phosphorus through the roots at these very low temperatures. Modeling A part of the original plan of the U.S. IBP was to construct predictive mathematical models of various ecosystems. This pond study offered several advantages for this approach; the ecosystem is somewhat simplified and physical factors (circulation and stratification) could be ignored; the active life of the pond organisms is completed in only 100 days each year; all the investigators worked on a common pond; controlled experiments could be carried out in replicate ponds and in the nearby laboratory. Our conclusion is that modeling is a very helpful approach but that the construction of a large, complicated predictive model is not possible with the existing gaps in our knowledge of aquatic ecology. 16 J. E. Hobbie Even in the pond ecosystem, the number of interactions is too great and most cannot be included in any model. We decided what to include on the basis of the carbon flow results (Figure 1-5). The modeling helped by forcing all the workers to sample just one pond and by forcing us to work on some processes which were virtually unknown (e.g., the detritus food web). It quickly became obvious that we simply did not understand many of the controls of the pond ecosystem. Despite this, the deterministic model which resulted did an excellent job of mimicking or simulating the annual cycles of carbon flow through the plankton and benthic systems. For example, the submodel of benthic algal photosynthesis (Figure 1-11) was developed with physiological data from 1973 but the simulation fits the 1971 and 1972 field data very well. Unfortunately, there were several assumptions that had to be included, in spite of our attempts to measure every parameter, and the results of the calculations were very sensitive to J L I I I I I I I I I L FIGURE 1-11. Model simulation (solid line) and measured estimates (circles) of the photosynthesis of epipelic algae (mg C m'^ hr'^). (Redrawn from Stanley 1974.) M I E u 9 E 20 40 Time, days Major Findings 17 bUU 1 1 ■Till D - 400 ^ / /sD^Xy—^^^N^ - ^ ^ — 200 «<\ _ SM ^ 1 1 1 1 - 60 FIGURE 1-12. Model simulation of epipelic algal biomass. The lines represent the deterministic model (D), the mean of nine runs of the stochastic version of the same model (SM), and the standard deviation of the stochastic mean (SD). (After Tiwari et al. 1978.) some of these. Therefore, the result shown in Figure 1-11 can be easily changed in an extremely drastic way by changing the algal respiration coefficient from 0.25 to 0.30. Even if we did have a good way of measuring algal respiration, in the field it would not distinguish between these two values. In a similar fashion, if the rate of maximum photosynthesis is changed from 0.05 to either 0.04 or 0.06, the algal biomass rapidly approaches zero. Yet we know that this maximum photosynthesis rate does change over the year in our pond. The next step was to see what would happen to the model if all the coefficients and parameters were varied slightly. This stochastic model is probably more realistic, as variability is a property of every biological measurement and interaction. Unfortunately, we did not have the data on the mean and standard deviation of every measurement that would be necessary to implement this approach. However, when we incorporated reasonable variability into the benthic model, the mean values (Figure 1- 12) were quite different from the deterministic model values. In only a few cases could the model be used to test hypotheses. For example, we did examine the hypothesis that the mixing of the sediments was an important control of the growth of benthic algae. This mixing could not be measured directly in the field or laboratory but the modeling exercise helped to put reasonable limits on this rate. Our conclusions from the modeling effort were that we knew too much about the pond ecosystem to be satisfied with fine-tuning a model 18 J. E. Hobbie which contained several untested assumptions. It was obvious that the interactions which controlled the results of the model were not necessarily the same ones acting to control the real ecosystem. In spite of our failure to construct a predictive model, the modeling exercise was very worthwhile and should be a part of every large-scale ecosystem project. Introduction and Site Description John E. Hobbie THE TUNDRA BIOME PROJECT Biome Studies The International Biological Program (IBP) was organized with the overall goal of discovering more about the biotic resources of the world through studies of the ecology of natural communities and of man himself. These studies took many forms in many countries but were generally small- scale efforts. It was decided that a part of the U.S. effort should be a series of large-scale, tightly coordinated studies of the ecology of a unit of the earth's surface which would represent a major ecological classification (e.g., desert, grasslands). These units, called biomes, should ideally encompass a watershed or similar area where terrestrial and freshwater ecosystems, and their interaction, could be investigated. The use of mathematical models of whole systems was to be a major tool for the investigations. Five biome studies were eventually established; one was the Tundra Biome. The site for the study, the flat coastal tundra near Barrow, Alaska, was suitable for terrestrial studies but was ideal for an aquatic study as ponds and lakes were abundant. Also, the well-equipped Naval Arctic Research Laboratory would provide logistic support and work space. The numerous ecological studies of this area over the preceding 25 years provided background information as well as a core of experienced scientists. The goal of the Tundra Biome study was to obtain a detailed understanding of the ecology of this site. The flows of carbon, nitrogen, and phosphorus were to be quantified and a mathematical model was to be constructed which would incorporate the data of these fluxes, their controls, and the interactions with the physical and chemical environment. It was hoped that the models could then be used in a predictive manner to investigate possible changes in the environment due to man or to natural alterations in the climate. 19 20 J. E. Hobbie The Tundra Aquatic Project In 1970, a modest, pre-IBP grant from NSF allowed the pond studies to begin. The goal of this grant was to follow the effects of oil and nutrient fertilization on ponds. Two projects, that of V. Alexander on phyto- plankton responses and that of R. Barsdate on nutrient and water chemistry, were started and the ponds chosen for whole-pond experimentation and for controls. One whole pond and several small subponds (plastic enclosures) were fertilized with P and N and one whole pond was treated with crude oil. In 1971, the IBP funding began and a complete range of aquatic projects was started. The emphasis during this year was on obtaining the most complete data possible for carbon, nitrogen, and phosphorus flow for both a pond and a lake (Ikroavik). The projects of Alexander and Barsdate were continued and additional projects added that dealt with zooplankton (R. Stross), bacteria, decomposition, and benthic algae (J. Hobbie), fish (J. Cameron), dissolved carbon, particulate carbon, and benthic respiration (M. Miller), macrobenthos (D. Bierle), protozoa (R. Dillon), and macrophytes (P. McRoy). Observations on the manipulated ponds were continued throughout the entire project. In 1972, the first modeling efforts began and a preliminary plankton model was developed. The field work at Barrow was oriented towards sub- pond experiments; treatments of nutrients (two concentrations), of added light, of darkness, and of higher temperatures were used. Because the outdated tracked vehicle continually broke down, travel to Ikroavik Lake became impossible and no more samples were taken. During 1973, emphasis was shifted to modeling, and simulation models of benthic, planktonic, and zooplankton systems were developed. The field work was devoted mostly to developing the constants and rates needed for the models. Several specialists were brought in for the summer to work on areas of research that were still poorly-understood: (T. Fenchel, protozoan ecology; D. Kangas, zooplankton respiration; S. Dodson, invertebrate predation.) The summer of 1974 was spent in preparing reports. Modus Operandi To reach the goals of the project we first used traditional limnological techniques to identify and measure the important pathways of carbon and energy flow in the tundra ponds. Next, the modeling was begun and used both to plan future experimental research on the processes and to evaluate the importance of proposed research. The philosophy we have followed in modeling is as follows: Introduction and Site Description 21 1. We must first understand as completely as possible the system to be modeled. 2. Only the important parts of a system, and their controls, can be modeled. 3. Wherever possible, constants, rates, and relationships included in the model must be measured and not taken from the literature. 4. The modeling exercise and the resultant simulations are regarded as tools to be used to further our understanding of how the ecosystem operates. 5. The modeling must be done by ecologists with the aid, if necessary, of a professional modeler rather than vice versa. LIMNOLOGY OF THE ARCTIC Circumpolar For this report, the Arctic is defined as the region north of the tree- limit which has a mean air temperature of less than 10°C during the warmest month. Until very recently, arctic limnology was entirely organism-oriented. This was in part due to the great difficulties of carrying out anything more than expedition-type collecting activities and in part due to the particular interests of the people involved. Modern studies of the physical, chemical, and limnological processes in lakes and ponds began after 1950 or so when research stations became established (e.g., at Disko Bay, Greenland, and Barrow, Alaska). As a result, the first review of arctic limnology (Rawson 1953) mentioned only seven papers that dealt with arctic lakes. The availability of these research stations, plus the realization that long-term studies were needed, led to a number of detailed investigations. In Scandinavia, the biology of Lapland lakes has been investigated by Ekman (1957), Holmgren (1968), and Nauwerck (1968). In Spitzbergen, water chemistry and zooplankton biology have been studied (Amren 1964a) and there are a number of reports of investigations of Greenland lakes (e.g., Hansen 1967, Holmquist 1959). Several investigators, such as McLaren (1964), Kennedy (1953), and Oliver (1964), worked in northern Canada. By far the greatest number of studies were made in northern Alaska (e.g., Livingstone et al. 1958, Hobbie 1964, Kalff 1967a, Stross and Kangas 1969, Carson and Hussey 1960) of the chemistry, biology, and physics of a wide range of lakes and ponds. The results from all these and other arctic studies are summarized in reviews by Livingstone (1963a), Kalff (1970), and Hobbie (1973). In brief, arctic lakes and ponds have the following characteristics: / 22 J. E. Hobbie 1. Arctic lakes and ponds seldom warm above 10°C and almost never stratify. 2. Arctic lakes and ponds shallower than 1.7 to 2.0 m usually freeze completely. 3. Ponds and lakes less than 2 m in depth do not contain fish. One consequence of the lack of predation is that the zooplanktonic crustaceans are almost all large species in ponds and shallow lakes. 4. The ice cover is 1 to 2 m thick and lasts for 8 or 9 months. 5. Arctic lakes and ponds usually contain low amounts of available nutrients and low total dissolved salts. However, as in the temperate regions, the total inorganic ion concentration is dif- ferent for drainage basins in different types of bedrock. 6. Oxygen is usually present in saturation concentrations in open waters but becomes depleted to some extent near the end of the under-ice period. In shallow lakes the exclusion of oxygen during the freezing of the ice may result in super-saturation (200%). 7. The biota of shallow freshwater lakes and of ponds are subjected to strong physiological stresses as the ions may be concentrated 30-fold during freeze-up, while the water immediately after the spring melt may resemble distilled water. 8. Only nannoplankton are found in arctic lakes and ponds. These usually bloom beneath the springtime ice of lakes but total primary production is low and lakes and ponds are oligotrophic. 9. With a few exceptions, each species of zooplankton has a dor- mant phase in its life cycle. 10. Fish are very slow-growing, but large fish may live for 40 years. 1 1. There are no benthic animals that graze on aquatic plants or that shred large organic particles or leaves. 12. The number of animal species is small and some groups — for example, sponges, Notonectidae, Corixidae, Gyrinidae, Dytisci- dae, and Amphibia — are rare or not present. 13. Decomposition rates are slow and large amounts of energy and nutrients are tied up in dead organic matter. It is obvious that this list of characteristics was obtained from largely descriptive studies. Recently, two IBP projects were carried out that were designed to add experimental studies to the descriptive observations in , order to gain an understanding of controlling factors and environmental interactions. These projects, one at Resolute Bay, Canada, and one at Barrow, Alaska, were located near airfields and studied only a single lake (Char Lake) or a small group of ponds in a single area (Barrow). Some of the results of the Char Lake project have been published. A general description (Rigler 1972) indicates that the lake is ice-covered until early August, has a moss cover over 30% of its bottom, and has low Introduction and Site Description 23 quantities of nutrients. Phytoplankton began to increase beneath the ice cover in February and reached a peak in May (Kalff et al. 1972). Morgan and Kalff (1972) found a maximum of 2 x 10 ' bacteria ml ' with peaks of glucose uptake in July and October. Zooplankton had low populations with Limnocalanus macrurus as the dominant form (Roff and Carter 1972). Most of the population hatched, grew, reproduced, and died between December and October, although a few adults were present during the entire winter. Finally, the long period of ice cover allowed the lake to be used as a sealed vessel respirometer to measure respiration of the ecosystem by changes in oxygen concentration (Welch 1974). Northern Alaska It was possible to carry out the IBP aquatic program only because of the experience and information provided by previous research in northern Alaska. Thus, even before the IBP aquatic project began, we knew such things as the primary productivity of the phytoplankton, the basic cycles of water chemistry, and the life cycles of many of the zooplankton species. There are seven types of freshwater habitats in northern Alaska: deep lakes, shallow lakes, ponds, large rivers, small rivers, streams, and springs. The deep lakes, located in the mountains, were formed mostly behind end moraines that dam narrow valleys. These lakes are rather rare and may number only 20 or 30. Shallow lakes, very abundant (many thousands) on the flat coastal plain, were formed mainly by melting of the ice-rich permafrost. These are only a few meters deep and many will freeze to the bottom each winter. The area of these lakes can be large, with lengths reaching up to 10 km or so. Ponds are extremely abundant (tens or hundreds of thousands) in the coastal plain region, particularly in the old lake beds. Here, the growth of ice wedges has pushed up networks of small ridges that contain small (50 m on a side), shallow (10 to 50 cm) ponds. Most of the limnological investigations have been carried out on lakes and ponds; little is known about the flowing water systems. However, there are a number of large rivers in northern Alaska and parts of these rivers are deep enough to allow fish to survive. Small rivers and streams, in contrast, cease flowing completely each fall. Because of the flat landscape and small amount of total precipitation, the drainage is poorly developed in the Barrow area and sizable amounts of flow occur only during the melting period. The final habitat, springs, occurs only in the mountain and foothill area. Although some ten or twenty springs exist, they are a very minor part of the entire aquatic scene. The fact that they flow year-round, however, allows a rich fauna to develop and illustrates both the potential production of arctic water and, by contrast, the strong stresses on intermittent streams. 24 J. E. Hobbie Deep mountain lakes were first investigated by Livingstone et al. (1958) who pointed out that most of the thermal, chemical, and biological events in these lakes were similar to those of oligotrophic temperate lakes. They thought that the major effect of the arctic environment was on the physiographic process affecting lake origins, sedimentation rates, and input from the drainage basins. An intensive study of two other deep lakes (Hobbie 1961, 1962, 1964) revealed that most of the yearly primary productivity of the plankton occurred beneath the ice cover in late spring and early summer. Later, both the light regime and the algal species and biomass responsible for this early season bloom were investigated in detail by Holmgren, Kalff, and Hobbie (reported in Hobbie 1973). It was found that when the snow depth was less than 10 cm, great amounts of light penetrated the ice. The light, plus the non-turbulent conditions beneath the ice, allowed large numbers of flagellates and diatoms to develop. After the ice left the lakes, conditions were poor for algal growth. Most of the research on shallow lakes has centered on the question of the origin and development of the oriented lakes of the coastal plain. These lakes originate when permafrost melts and the soil subsides (Rex 1961) and receive their orientation from wind-driven currents which differentially erode the ends of the elongated lakes (Carson and Hussey 1960). Scattered bits of information that exist on Ikroavik Lake, near Barrow, show it to have low numbers of algae, nutrients (Prescott 1953), and benthic animals (Livingstone et al. 1958). The whitefish population has also been described (Wohlschlag 1953). Another well-studied shallow lake is Imikpuk, which is not an oriented lake and which lies close to the Arctic Ocean. Chemistry of the lake has been reported by Howard and Prescott (1973) and Boyd (1959); the primary productivity by Howard and Prescott (1973) and Kalff (1967b); the zooplankton by Comita (1956) and Edmondson (1955); and the microbiology by Boyd and Boyd (1963). All of the pond limnology in northern Alaska has been done near Barrow. The most extensive investigations were of the chemistry and plankton productivity (Kalff 1965, 1967a, 1971). In these ponds, phosphate, nitrate, ammonia, trace elements, and growth factors all stimulated photosynthesis at various times. Kalff concluded that nutrient deficiency did exist and that plankton productivity was extremely low (around 1 g C m"^ yr~'). Another series of studies dealt with repro- ductive cycles and controls of zooplankton (Stross and Kangas 1969). The only large rivers studied in northern Alaska have been the Colville (Kinney et al. 1972) and the Sagavanirktok (Carlson et al. 1974). These studies were mostly concerned with water chemistry but zooplankton (Reed 1962), fish (McCart and Craig 1971), and discharge (Arnborg et al. 1966) have also been looked at. The rivers contain little plankton and a scanty bottom fauna. Introduction and Site Description 25 Smaller streams are also little known and the research has been restricted to observational limnology. The best-studied area is at Cape Thompson (68''N, 165''W) where the discharge (Likes 1966) and biology (Watson et al. 1966a) on Ogoturuk Creek were investigated. Near Barrow, Brown et al. (1968) measured the hydrology of a small watershed (1.6 km^) over four summers and Lewellen (1972) reported on flow and chemical data from three other Barrow area watersheds. Springs are present only in the foothills and mountains of arctic Alaska. One spring, Shublik Spring on the Canning Rivt., has been sampled by Kalff and Hobbie (unpublished, quoted in Hobbie 1973). It flows year-round at 4.0 to S.S'C, and contains a fantastic abundance of insects as well as a dwarf char (McCart and Craig 1973). There are other reports that cover several aquatic haoitats. Hydrology was reviewed by Dingman (1973), Kalff (1968), and Barsdate and Matson (1966). GEOGRAPHY AND GEOMORPHOLOGY Geographical Setting Northern Alaska, all of which lies north of the tree limit, is cut off from the rest of the state by the east-west running Brooks Range, an extension of the Rocky Mountain System. North of the mountains, which have an area of 136,200 km^ lie the Arctic Foothills (100,800 km^) and between the foothills and the Arctic Ocean lies the Arctic Coastal Plain which contains 70,900 km^ (Walker 1973). Barrow lies on the northern tip of the coastal plain, some 1 75 km from the foothills (Figure 2- 1 ). Near Barrow, the flat coastal plain is covered either by large lakes, shallow ponds, or old drained lake basins (Figure 2-2). In places, freshwater lakes and ponds cover up to 40% of the surface. Despite the abundance of water, streams are small and most flow only during the spring melt. The remainder of the area is covered by grasses, sedges, mosses, and lichens. Usually, the standing dead stems and leaves of the grasses and sedges dominate the scene and color the tundra brown. The Naval Arctic Research Laboratory is on the coast of the Chukchi Sea, 10 km from Point Barrow, and the town of Barrow lies 5 km further southwest (Figure 2-1). Research on tundra ecology has been carried out at this laboratory since 1947, while the National Weather Service has operated a first-class station at Barrow since 1920. The pond research site (71°18'N and 156''42'W) is halfway between the laboratory and the town and 2 km inland. A shallow lake, Imikpuk, lies adjacent to the laboratory and to the ocean, and a large lake, Ikroavik, lies 7 km south. 26 J. E. Hobbie FIGURE 2-1. 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 and the study ponds in the lower left corner. (Photograph by CRREL.) Geomorphology The Arctic Coastal Plain consists of unconsolidated silty sand and gravel of Quaternary age (Gubik Formation) deposited in a shallow sea (Black 1964). The uppermost section at Barrow was deposited and reworked over the past 35,000 years (summarized in Brown and Sellmann 1973). Radiocarbon dates and composition analyses of peat suggest that tundra existed in the Barrow area for as long as 14,000 years (Brown 1965). Based on a number of radiocarbon dates, it is believed that most of the soils and surficial features of the present land surface are not older than 8,000 to 10,000 years and perhaps are considerably younger. Mean annual air temperatures on the North Slope of Alaska are below freezing; thus, there is continuous permafrost (perennially frozen ground) beneath the entire area. At Barrow, the frozen layer is 400 m thick (Brown and Sellmann 1973) but there is a layer of soil 25 to 100 cm thick that does thaw each summer. The depth of thaw is influenced by the type Introduction and Site Description 27 ^^:k V '.V ^- i. \ ^^'>^"n>*» ^2 • -.-I -» FIGURE 2-2. Aerial photograph of Tundra Biome site. Arrow indicates Pond C. of vegetation, amount of insulating plant litter, and type of soil. For example, beneath a thick vegetation mat the depth of thaw may be only a few centimeters while coarse-textured, south-facing materials may thaw to 28 J. E. Hobbie a depth of 100 cm (Walker 1973). At the Barrow site, the maximum depth of thaw is around 25 cm below the grassy tundra while in the sediments of the ponds it may vary from 20 to 50 cm. Large rivers and lakes of this region may be underlain by extensive thawed areas. Brewer (1958) found 60 m of thawed material beneath Imikpuk Lake. The presence of permafrost has important biological effects. Because the permafrost is impervious, the water cannot drain away, and the low- lying soils are saturated. Roots are restricted to the upper, thawed layer of soil which limits the total quantity of nutrients available. Finally, nutrients and energy are removed from circulation either when the permafrost level rises or when soil and sediments accumulate and become part of the permafrost. The upper layers of permafrost on the coastal plain contain large quantities of ice. One form of this occurs as interstitially segregated ice (up to 80% of the top 3 or 4 m of permafrost) (Sellmann and Brown 1965). When this melts, due to disturbance of the plant cover or to heat transfer by flowing water, a depression is formed that may result in a pond. Another form of ice occurs when water runs into cracks formed by the winter contraction of the frozen tundra. The resulting buried ice takes the form of ice wedges that can range from a few centimeters to 8 m in width. Over many years, the wedges grow and eventually a network of ice wedges is formed. Sometimes these wedges are expressed on the surface as polygonal ground (Figure 2-3) caused by heaving or other surface processes that form troughs and ridges. Typically, these polygons may be 20 to 50 m across; polygonal ground covers almost the entire coastal plain. In the early stages of growth of polygonal ground, the polygons are low-centered and often contain small ponds. The water changes the insulating properties of the surface and also traps heat so that the upper layers of the permafrost thaw, subsidence occurs as the ice melts, and a basin up to 0.5 m in depth is formed. These ponds frequently coalesce and may form a lake. Eventually, the lake may grow enough that a drainage divide is breached. Then the lake may drain and the polygonal ground start to form again; the whole process has been called the thaw-lake cycle (Britton 1957). Many of the larger thaw lakes of the coastal plain display a striking elliptical shape with an elongated north-south axis. The exact reason for the orientation has been the subject of a number of studies and theories (Black and Barksdale 1949, Livingstone 1963a, Carson and Hussey 1960), but it is evident that differential erosion is still occurring today. For example, Lewellen (1972) measured a rate of elongation of 1.3 m per year in Twin Lakes. In similar lakes, Hussey and his co-workers measured currents at the ends of the lakes of up to 61 cm per second, which Livingstone (1963a) believed to be adequate to account for the elongation of the lake basins. However, Walker (1973) believes that the precise mechanism of elongation is still unexplained. Introduction and Site Description 29 Approx. Scale FIGURE 2-3. Aerial photo showing polygonal ground near Barrow, Alaska. Study ponds are labeled. Another type of pond, called a trough pond, is formed when an ice wedge melts. Around Barrow, melting is often caused by destruction of the insulating vegetation by tracked vehicles. * I I ■I 1 U < 1) M 6Xj >. ..M 03 't oo - t —1 (N o m OO lO H O < fN -H r- t~- ^ Q > o Z • r^ IT) o r^ <^ m 0 W o on sill o ON ON UJ li-j z °o Z ^ w lO o O o ^ 00 . a: "5 K ^ c O <-> 00 t-l (U ,c cd ^ c o re Z u o 1-1 3 O 30 em > < 0\ VD ON OS oo On ON ON o On ON ON r- ON E ON o 00 O o O 00 en On o o CNl ON q 00 00 1 p M ^^ 3 5^^ == ^ "■ ° O ni 3 4, ON CO >o r- (N ■t 0\ 1-H 00 '^ (N 00 ■^ 00 CO 4 o in o