Restoration of Lakes, Streams, Floodplains, and Bogs in Europe
Wetlands: Ecology, Conservation and Management Volume 3
Series Editor:
Max Finlayson Institute for Land, Water and Society Charles Sturt University Albury, NSW, Australia e-mail:
[email protected]
Aims & Scope: The recognition that wetlands provide many values for people and are important foci for conservation worldwide has led to an increasing amount of research and management activity. This has resulted in an increased demand for high quality publications that outline both the value of wetlands and the many management steps necessary to ensure that they are maintained and even restored. Recent research and management activities in support of conservation and sustainable development provide a strong basis for the book series. The series presents current analyses of the many problems afflicting wetlands as well as assessments of their conservation status. Current research is described by leading academics and scientists from the biological and social sciences. Leading practitioners and managers provide analyses based on their vast experience. The series provides an avenue for describing and explaining the functioning and processes that support the many wonderful and valuable wetland habitats, such as swamps, lagoons and marshes, and their species, such as waterbirds, plants and fish, as well as the most recent research directions. Proposals cover current research, conservation and management issues from around the world and provide the reader with new and relevant perspectives on wetland issues.
For other titles published in this series, go to www.springer.com/series/7215
Martina Eiseltová Editor
Restoration of Lakes, Streams, Floodplains, and Bogs in Europe Principles and Case Studies
Editor Martina Eiseltová Environment and Wetland Centre Public Association, Prague and Crop Research Institute Prague Czech Republic
[email protected]
ISBN 978-90-481-9264-9 e-ISBN 978-90-481-9265-6 DOI 10.1007/978-90-481-9265-6 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: 2010930512 © Springer Science+Business Media B.V. 2010 No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Preface
In my life I have had many wonderful opportunities to meet and work with top scientists – people who have devoted their professional life to research in wetland ecology and applied their scientific knowledge to practical wetland conservation, including wetland restoration. For me this has revealed a new world – a passion for understanding the play-rules of nature, for discovering its wisdom and divining the relationships between the members of an ecological community and their multiple feedback loops. It also instilled in me that wetland restoration has to be based on deep understanding of natural processes, understanding the position and functioning of wetlands in a wider context of the landscape and that one can learn from both successful and less successful restoration activities, if they are well described. A big challenge with us today, and which indeed sometimes appears almost insurmountable, is to overcome the sectoral divide in water resource management. Due to our unsustainable management of water resources, water (and especially clean water) is becoming a scarce commodity and hence there arises a conflict: between environmentalists, who require sufficient water supply for the biota and nature, and human society’s overall consumption of water resources – whether it be the use of water for electricity production, agriculture irrigation, industry use or supply of drinking water. The conflict is ever growing. Current scientific and administrative structures in society tend to keep these sectors apart – and in a state of competition rather than cooperation. When it comes to land/water links the situation is also highly complicated as the land and water management is again in separate hands and the understanding of land and water resource management interactions not widely adopted. My recent job in the agricultural field has revealed to me that the prevailing response of agricultural research to the threat of climate change and increasingly widespread droughts has been plant breeding or even gene manipulation in order to develop crops that will maintain high yields with reduced consumption of water. But lower water uptake means lower evapotranspiration and hence reduced humidity and hotter and drier local climates, and thus only aggravating the overall problem of water shortages. Another outcome of climate change is also the greater fluctuations in dry and wet periods, i.e. droughts interchanging with floods, as observed in recent years. All this, one may conclude, calls for the re-creation and return of more wetlands and water-saturated soils into our landscapes. v
vi
Preface
The irreplaceable role of wetlands in landscape functioning has become ever much clearer, while the destruction and damage of wetland ecosystems has become ever more widespread. Increasingly, wetland ecologists began to turn their attention to wetland restoration, and in the past 40 years an extensive experience in wetland restoration has been gained. This is a field that is rapidly developing, bringing more experience and allowing fine tuning to some early restoration approaches. As is stressed in this book, wetland restoration is not usually a simple or straightforward task. It often requires a sound scientific understanding of general wetland ecology and – as wetlands are not isolated systems – also an understanding of their position and functional role in the broader context of the landscape. Generally speaking, wetland restoration requires cooperation between different sectors – and a communication between a diverse number of disciplines such as: hydrology, geology, wetland ecology, botany, and zoology; and more recently, socio-economy can also help by offering new aspects of wetland functions and their benefits to humans. An understanding of wetland development is essential too, as wetlands are not static ecosystems but are gradually evolving. For wetland ecosystem restoration to be successful in the long-term – to be sustainable – space and time factors must also be taken into account. All this means that wetland restoration cannot follow a standard prescription, step-by-step; however, the experience that has been gained in other restoration projects, and presented in the form of case studies, can provide a valuable source of information and guidance. Valuable lessons can be learned from well-described restoration projects; at least, those that were completed and their achievements evaluated with the help of a well-designed monitoring programme. The origin of this book lies in an integrated wetland management training programme launched by Wetlands International in 1992. From 1992 to 1998, I had the pleasure to coordinate a series of international training courses addressing different aspects of restoration of various wetland types – shallow lakes and fishponds, streams and their floodplains, and peatbogs. Later on, further training courses were organised under the auspices of the Ramsar Convention (an international treaty on wetlands) through the Czech National Ramsar Committee, and also since 1997 in cooperation with the UNESCO Man and the Biosphere Programme. The focus of these courses was to bring course participants up-to-date with not only the current available theoretical knowledge and advances in wetland ecology required for successful wetland restoration but also with the experience gained from practical wetland restoration projects implemented in the field. Furthermore, the aim was to allow the exchange of expertise between scientists and wetland managers, between ecologists and engineers, and above all to bridge the gap between the various sectors involved in land and water resource management. Out of this experience was born this book. Bridging the gap between ecologists and engineers is the big challenge in natural resource management: to start an equal dialogue between ecologists trying to understand natural processes and who accept a certain level of unpredictability when dealing with ecosystem management, and engineers who, used to the power of calculations, expect predictable and precise outcomes from their interventions within natural systems. The recent disastrous floods experienced almost every-
Preface
vii
where around Europe have hopefully taught us an important lesson – nature cannot be so easily ‘ordered about’ and ‘precise’ calculations can be wrong. Thus the dialogue that tries to close our gaps in knowledge between the experts of different fields should be based on observations of natural processes, searches for variability and diversity in solutions, an acceptance of a certain level of flexibility and adaptability to changing conditions, and the ever-fluctuating network of nature itself – the web of life. The information, knowledge and experience collected in this book should provide a valuable overview of the key causes and processes that lead up to ecological degradation and give suggestions for possible solutions. The first chapter of this book is about water – the water cycle and its role in the landscape. Its main message is the need for sustainable and coordinated land and water use. Arguments that there is an urgent need to return more water and natural vegetation to the landscape as well as criteria for the assessment of landscape sustainability are presented. The chapters that follow – on wetland evolution, development of aquatic vegetation, the trophic interactions and the food chain in water bodies – give an insight into the functioning of shallow water bodies and wetlands and their development within the landscape influenced by human interventions. The following chapters (Chapters 6–17) are then focused on individual restoration projects implemented in the field, presenting a spectrum of restoration activities addressing different types of wetland ecosystems – shallow lakes, streams and their floodplains, and bogs – that may serve as guidelines and provide useful information and inspiration for future projects. My hope is that the book will become a valuable resource material to experts from many different fields and help unite their views on the irreplaceable role of water and wetlands in the landscape. Prague, December 2009
Martina Eiseltová
In memory of Jaroslav Hrbácˇ ek, an outstanding Czech limnologist who sadly passed away a few days before this book was published.
ix
Acknowledgments
The concepts presented in this book are the results of many years of experience with wetland restoration in different parts of temperate Europe – an experience gained by many outstanding wetland ecologists and excellent teachers. My first acknowledgment must therefore go to the authors of individual chapters who kindly agreed to contribute to this book. Many of them served as lecturers on training courses in wetland management and restoration that have been organised since 1992 – first by the International Waterfowl and Wetlands Research Bureau/Wetlands International from the UK and later by the Wetland Training Centre, located in Třeboň, Czech Republic. During these training courses a nice team of people got together and created the core contributors to the present publication. Many others joined later. Amongst these was Sake van der Schaaf, to whom I am indebted for help with reviewing the manuscripts dealing with the restoration of mires. The final strength to complete this book I gained during a recent study tour to vast areas of wetlands in Belarus, Lithuania and Poland – just seeing the sheer dimension of these still well-preserved wetlands gave me a glimpse of how the landscape elsewhere in the temperate zone might have looked like in the past – before human intervention with the hydrological cycle, drainage and reclamation of land for agriculture and urban settlements took its toll. Walking bare-footed through the mats of Sphagnum was such a refreshing experience – one which should be offered to as many people as possible. There is no doubt that the immense spirit of wetlands would turn people on to protecting these invaluable habitats once they have experienced their power. My thanks here go to the nature itself and people who strive to protect her. This book is a result of almost 20 years devoted to the promotion of scientific knowledge in wetland management, restoration and conservation through international training courses. During that time I met many wonderful people both in the role of course lecturers and course participants. The exchange of knowledge, experience and ideas was always enriching and without these encounters this book would not come into the world. I would like to dedicate this book to three outstanding wetland scientists I have met and worked with throughout the years, Sven Björk from Sweden, Willy Ripl from Germany and Jan Květ from the Czech Republic. They have always been extremely xi
xii
Acknowledgments
kind and offered much needed support. Furthermore, I am very grateful to Max Finlayson and Jan Pokorný for giving me valuable advice and guidance in the early years of my professional life. My final thanks are reserved for my family, my husband Steve Ridgill who kindly joined me in the effort of preparing this book for publication and has painstakingly read all the contributions and improved the English of many, and our children Philip and Nathalie who gave us the space and time to work. Prague, December 2009
Martina Eiseltová
Contents
1 Criteria for Sustainable Restoration of the Landscape.......................... Wilhelm Ripl and Martina Eiseltová
1
2 The Evolution of Lakes and Wetlands..................................................... Sven Björk
25
3 Development of Aquatic Macrophytes in Shallow Lakes and Ponds.......................................................................... Jan Pokorný and Sven Björk 4 Food Web and Trophic Interaction and Development........................... Jaroslav Hrbáček 5 Principles, Planning and Accomplishment of Lake Restoration Projects..................................................................... Sven Björk 6 Restoration of Eutrophic Lakes by Sediment Treatment....................... Wilhelm Ripl 7 Restoration of Eutrophic Lakes by Phosphorus Precipitation, with a Case Study on Lake Gross-Glienicker.................. Klaus-Dieter Wolter
37 45
71 77
85
8 Restoration of Lakes Through Sediment Removal, with Case Studies from Lakes Trummen, Sweden and Vajgar, Czech Republic............................................................................. 101 Sven Björk, Jan Pokorný, and Václav Hauser 9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control: The Case Study of Lake Hornborga, Sweden..................................................................... 123 Sven Björk xiii
xiv
Contents
10 The Stream and Beyond: Reinstating Natural Functions in Streams and Their Floodplains........................................ 145 Bent Lauge Madsen 11 Floodplain Restoration of Large European Rivers, with Examples from the Rhine and the Danube...................... 185 Erika Schneider 12 Restoration of Streams in the Agricultural Landscape........................ 225 Lena B.-M. Vought and Jean O. Lacoursière 13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands: A Quantitative Assessment...................................................................... 243 S. van der Schaaf, M.J. van der Ploeg, S.H. Vuurens, and M.M.J. ten Heggeler 14 Self-Recovery of Cut-over Bogs: Summary from Case Studies........... 265 Elve Lode, Lars Lundin, and Mati Ilomets 15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands................................................. 285 Hilde B.M. Tomassen, Alfons J.P. Smolders, Sake van der Schaaf, Leon P.M. Lamers, and Jan G.M. Roelofs 16 Restoration of Drained Mires in the Šumava National Park, Czech Republic............................................................... 331 Ivana Bufková, František Stíbal, and Eva Mikulášková 17 Local and Global Impacts of Mire Drainage: An Impetus for Hydrology Restoration: Yelnia Mire, Belarus............ 355 Alexander Kozulin, Sergey Zuyonok, and Viacheslav Rakovich Index.................................................................................................................. 367
Contributors
Sven Björk Department of Ecology, Limnology, University of Lund, SE-223 62, Lund, Sweden
[email protected] Ivana Bufková Administration of the Šumava National Park, Sušická 399, 341 92, Kašperské Hory, Czech Republic
[email protected] Martina Eiseltová Environment and Wetland Centre, Public Association, U Křížku 8, 140 00, Prague 4, Czech Republic and Crop Research Institute, Drnovská 507, 161 06, Prague 6, Czech Republic Václav Hauser ENVI Ltd., Dukelská 145, 379 01 Třeboň, Czech Republic Jaroslav Hrbáček Biology Centre of the AS CR, v.v.i., Institute of Hydrobiology, Na Sádkách 7, 370 05 České Budějovice, Czech Republic, Corresponding address: Hekrova 820, 149 00, Praha 4, Czech Republic
[email protected] Mati Ilomets Institute of Ecology, Tallinn University, Uus-Sadama 5, 10120, Tallinn, Estonia
[email protected] Alexander Kozulin Institute of Zoology of the National Academy of Sciences of Belarus, Akademicheskaya Street 27, 220072, Minsk, Belarus,
[email protected] Jean O. Lacoursière Sustainable Water Management Group, Kristianstad University, 291 88, Kristianstad, Sweden xv
xvi
Contributors
Leon P.M. Lamers Department of Aquatic Ecology and Environmental Biology, Radboud University Nijmegen, P.O. Box 9010, 6500, GL Nijmegen, The Netherlands Elve Lode Department of Soil and Environment, Swedish University of Agricultural Sciences, P.O. Box 7001, SE-75007, Uppsala, Sweden
[email protected] or
[email protected] Lars Lundin Department of Soil and Environment, Swedish University of Agricultural Sciences, P.O. Box 7001, SE-75007, Uppsala, Sweden
[email protected] Bent Lauge Madsen National Agency of Forest and Nature, Watercastle Old School Research Station, 38 Kirkensgaardvej, DK-7620, Lemvig, Denmark
[email protected] Eva Mikulášková Administration of the Šumava National Park, Sušická 399, 341 92, Kašperské Hory, Czech Republic Jan Pokorný ENKI, p.b.c., Dukelská 145, 379 01, Třeboň, Czech Republic
[email protected] Rakovich Viacheslav Institute for Nature Management of the National Academy of Sciences of Belarus, Scoryny Street 10, 220114, Minsk, Belarus Wilhelm Ripl Technical University of Berlin, Nuthestrasse 4A, D-14513 Teltow, Germany
[email protected] Jan G.M. Roelofs Department of Aquatic Ecology and Environmental Biology, Radboud University Nijmegen, P.O. Box 9010, 6500, GL Nijmegen, The Netherlands Erika Schneider KIT Karlsruhe Institute for Technology – University of Land Baden-Württemberg, Chair of the WWF-Institute for Floodplains Ecology, Josefstrasse 1, 76437, Rastatt, Germany
[email protected]
Contributors
xvii
Alfons J.P. Smolders B-WARE Research Centre, Radboud University Nijmegen, P.O. Box 6558, 6503, GB Nijmegen, The Netherlands František Stíbal Administration of the Šumava, National Park, Sušická 399, 341 92, Kašperské Hory, Czech Republic M.M.J. ten Heggeler Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700, AA Wageningen, The Netherlands Hilde B.M. Tomassen B-WARE Research Centre, Radboud University Nijmegen, P.O. Box 6558, 6503, GB Nijmegen, The Netherlands
[email protected] M.J. van der Ploeg Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700, AA Wageningen, The Netherlands Sake van der Schaaf Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700 AA Wageningen, The Netherlands
[email protected] Lena B.-M. Vought Sustainable Water Management Group, Kristianstad University, 291 88, Kristianstad, Sweden
[email protected] S.H. Vuurens Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700, AA Wageningen, The Netherlands Klaus-Dieter Wolter Kolberger Street 3, D-65191, Wiesbaden, Germany
[email protected] Sergey Zuyonok National Union for Bird Conservation of Belarus, Makaenka Street 8-313, 220023 Minsk, Belarus
Chapter 1
Criteria for Sustainable Restoration of the Landscape Wilhelm Ripl and Martina Eiseltová
Abstract Water, matter and energy are the three basic requirements for any ecosystem to thrive. Studies of natural processes in a central European virgin forest have brought us an understanding of how nature closes the cycles of water and matter and evenly dissipates the incoming solar energy that runs processes. As a result, climatic events, such as precipitation, runoff and temperature, are evenly distributed in time and irreversible matter losses remain low. By minimising matter losses nature prolongs its life-span, i.e. enhances its sustainability. When compared to agricultural landscapes, we can reveal the main mistakes of human interference with natural processes that lead to the opening of cycles, bringing about high irreversible matter losses. Investigations have shown that areal matter losses measured in agricultural catchments in Germany are some 50–100 times higher than those from unmanaged land in a virgin forest. As matter losses are mainly connected to water run off, every disturbance to the hydrological regime has a vital impact on landscape sustainability. Extensive drainage, including that of wetlands and transformation of rivers into drainage channels, has such a negative impact. This chapter brings forward the argument that for greater landscape sustainability it is essential to restore and return more wetlands and natural vegetation cover to the landscape and restore natural dynamics to rivers and streams. Criteria suitable to assess the sustainability of land use are proposed as being: solar energy dissipation and water and matter recycling within the smallest delimited area, such as a catchment or sub-catchment.
W. Ripl (*) Technical University of Berlin, Nuthestrasse 4A, D-14513 Teltow, Germany e-mail:
[email protected] M. Eiseltová Environment and Wetland Centre, Public Association, U Křížku 8, 140 00, Prague 4, Czech Republic and Crop Research Institute, Drnovská 507, 161 06, Prague 6, Czech Republic M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_1, © Springer Science+Business Media B.V. 2010
1
2
W. Ripl and M. Eiseltová
Keywords Dissolved matter losses • Energy dissipation • Sustainable land management • Virgin forest • Water and matter cycles
1.1 Introduction Humans have greatly modified the natural forms of most large rivers – constraining them into straightened river channels and cutting them off from their floodplains by impoundments. The lowering of the groundwater table and the drainage of a large number of wetlands have led to a serious impoverishment of the landscape with respect to surface water bodies and water-saturated soils. Whilst the restoration of individual wetland sites has been attempted for more than 40 years now, there are hardly any restored wetland sites that would be sustainable in the long term without further intervention – because of the negative impacts caused by unsustainable land use within their catchments. Years of experience in wetland restoration have shown that wetlands function in the context of the wider landscape and that for any lake, river or wetland restoration to be sustainable in the long term, restoration must start in the catchment. Land use within the catchment must be optimised in such a way that the determining processes acting both on and within the catchment, namely – the dissipation of solar energy, and the recycling of water and matter – are maximised. If this is not the case, then any restorative measures undertaken just within the immediate borders of the wetland itself will be of relative short duration only – for example, a lake treated for increased eutrophication will quickly deteriorate again if the input of nutrients from the catchment – the probable source of the eutrophication in the first place – has not been reduced. Understanding natural processes is an important pre-requisite to implement sustainable land management and/or restoration of the landscape. Observation of nature and learning from natural communities of plants, animals and microorganisms brings us an understanding how nature utilises its resources in a sustainable way (in nature, what is ‘waste’ for one organism becomes a resource for another, i.e. the processes are coupled together and ‘wastes’ are kept to a minimum). Reuse and recycling are the rules of nature. This knowledge can then help us to design criteria essential for assessing whether our management of nature and the landscape is sustainable or not. It should be stressed that sustainable management of natural resources is our only possible way to sustain life on our planet. In the next few paragraphs we will explore the most important advances of science that have helped us to understand the self-organising process of nature and the sustainable conditions of living systems. For a long time in history, due to the limitations of classical physics, the way how to achieve sustainable conditions was unclear. With classical thermodynamics, which necessarily considered an approximation to equilibrium conditions, only processes that went from structured to unstructured conditions were thought to be possible. A breakthrough to the understanding of living systems was offered by
1 Criteria for Sustainable Restoration of the Landscape
3
Schrödinger who came up with his ‘order from disorder’ theory using the help of non-equilibrium thermodynamics. Schrödinger described the self-organisation of living systems, taking energy from outside, i.e. being open systems with regards to energy flow, and using it to produce, within themselves, a more ordered (structured) state (Schrödinger 1944, in Schneider and Kay 1994). It was Schrödinger who recognised that ‘living systems operate in a world of energy and material fluxes’ far from equilibrium (Schneider and Kay 1994). Another important contribution to the description of self-organising systems was the theory of ‘dissipative structures’ by Ilya Prigogine. Prigogine discovered that under dynamic conditions, far from equilibrium, a kind of self-organisation is possible (Prigogine 1980; Prigogine and Stengers 1984). Structuring is enabled by the sorting of material under dynamic conditions – carried out at the interface between two different states, i.e. solid/liquid or gas. For example, sand grains of different sizes on a beach, under the influence of a rolling wave of water, undergo acceleration and then fall under the influence of gravity, and end up sorted on the beach in waves – as ripples in the sand. Similarly, sand dunes are the result of sorting by the wind. Equally, life and living structures are the result of dynamic, structured processes rather than of random ones. This is important to have in mind when addressing landscape sustainability. Yet another concept important when dealing with sustainability is the concept of the so-called r- and K-strategies introduced by Mac Arthur and Wilson (1969) that shows the importance of material re-cycling – as space becomes a limiting factor then recycling (the coupling of processes) becomes increasingly critical. At the beginning of any ecosystem development – say an ecosystem developing on bare land – the first plants to colonise the place are so-called ‘r-strategists’, plants with high reproduction rates, fast growth and high productivity. These pioneer plants quickly colonise the open terrain and their populations increase rapidly while they are unrestricted by a shortage of nutrients or lack of space. Such a phase of high productivity is also called a pioneer phase or r-phase where energy dissipative productivity is the aim and organisms are selected accordingly. The rapid expansion phase of the ‘r-strategists’, however, is soon replaced by their fast disappearance as they are outcompeted by ‘K-strategy’ organisms. ‘K-strategists’ are better suited for survival in a crowded or limited environment, though net productivity becomes lower and biomass grows more slowly. ‘K-strategy’ is characterised by team work: functionally diverse organisms (such as decomposers, consumers and producers) achieve local recycling, and are therefore suited for a sustained development under the conditions of local natural resource limitation. A similar development to that in nature can also be found in the development of our human societies. Initially, we were surrounded by abundant natural resources – be it fossil fuels, extensive forests, clean water and air, and fertile soils. Accordingly, we have adopted the development mode of ‘r-strategists’. This has, however, led to severe limitations and to the overexploitation of natural resources in many parts of the world. It is, therefore, necessary to accept that we are now in a transition period to a new phase. Non-renewable resources are quickly becoming depleted, initially extensive primary forests have been largely destroyed, clean water has become
4
W. Ripl and M. Eiseltová
scarce and soil fertility largely reduced. Further survival and development of our societies will require a change from a production-oriented economy to a recyclingoriented communication-and-socialisation-related economy in order to achieve sustainability. Above we have briefly described the scientific progress that was necessary to gain understanding how nature, through the process of self-organisation, tends towards enhancing sustainability. In the following text we will give a more detailed description of processes that lead to sustainable development of ecosystems and also show how humans interfere with this sustainability by the prevailing current management of land and water. Some general rules for landscape restoration are proposed.
1.2 Self-organising Natural Systems – Studied in a Virgin Forest Investigations were carried out in a small untouched virgin forest in Austria (Rothwald virgin forest of 3.5 km2, altitude between 900 and 1,900 m asl, see map – Fig. 1.1) between February 1999 and December 2004. Water flow and the coupled matter flow processes were monitored by automatic devices – temperature probes,
Fig. 1.1 Physical map of Rothwald virgin forest, Austria. The studied area is delineated by black line, black dots show the location of temperature sondes
1 Criteria for Sustainable Restoration of the Landscape
5
pressure and water level gauges with high time resolution. The aim was to reveal the most important natural processes in its spatiotemporal distributions leading to sustainable structures.
1.2.1 Water and Matter Cycling in the Virgin Forest The data obtained in the virgin forest showed very clearly the unique feedback-control character of unmanaged areas. The most amazing property of this site was the very damped and even distribution of climatic events such as precipitation, runoff and temperature in time. The temperature amplitudes between day and night almost never exceeded 8–9°C during the summer (Ripl et al. 2004). Despite the relatively high precipitation in this area (over 1,000 mm year−1) the runoff from the virgin forest was very low and restricted mainly to the period of snow melt (February until May), with water very low in ionic content (see below). Since, in the studied area of the virgin forest, there were no periods of overheating or areas that would dry out, the conclusion was that very short water cycles with a frequency of 1 day or less prevail. In water-saturated conditions of the virgin forest, the amount of dead organic matter and debris is much higher than in managed forests due to the fact that in a virgin forest nothing is removed or harvested from the area, and the decomposition of organic matter is much slower due to the water-saturated conditions. The debris layer is 2–4 times higher in the Rothwald virgin forest than in the neighbouring managed forest (Splechtna K. personal communication, 2000) The water-retaining capacity is so high that water runoff within and from the virgin forest during summer took place only when rain events exceeded more than 35–50 mm a day (Ripl et al. 2004). The water was mostly absorbed in the fibrous debris layer at the surface of the soil. The high water-holding capacity was further increased by the frequent occurrence of large holes created by the uprooting of fallen dead trees. These holes become successively filled with litter, keeping the water-holding capacity of the site at a very high level. The high debris content at the soil surface and the abundance of holes retaining water lead to the permanently high water-retention capacity in the virgin forest. Evapotranspiration from the virgin forest is maximised since the debris layer never dries out. Thus decomposition of debris is to a large extent controlled by the low redox-conditions due to the relatively small solubility of oxygen in water. During the period of study, the soil, close to the soil surface, was moist, and in many cases fairly water-logged. During summer no observable overheated areas occurred at any time within the boundaries of the virgin forest; temperatures were quite evenly distributed and excessive wind events were very scarce. Despite the climatic change observed elsewhere, in the virgin forest the snow cover lasted every year from about mid-November until almost the end of May. The water analyses of melted snow samples showed extremely low conductivity values. From 17 samples a median of 0.6 mS m−1 at 20°C and a pH of 6.27 were obtained (Table 1.1). These results indicate that there was a much quicker turnover of water
0.01
0.02
0.01 0.01 0.01 * 0.09 0.01
0.02 0.05
13.02.01 0.41
22.01.01 0.47
1.19 0.37 0.26 0.94 1.45 0.52
02.03.01 19.03.01 31.03.01 15.04.01 11.02.01 21.01.01
13.06.01 0.78 12.06.01 0.99
*
0.02 0.02 0.09 0.00 0.01 0.01 0.01
0.09 0.00 0.01 0.03 16 Alkalinity mmol l−1
0.05
1.45 0.26 0.60 0.72 17 Conductivity mS m−1 20ºC 0.91 0.98 0.31 0.95 0.57 0.53 0.60
12.02.01 0.91
24.05.99 15.09.00 15.09.00 03.03.01 18.03.01 30.03.01 20.01.01
Date
No data available
Schnee 402 Schnee 15 cm Schnee Oberfl. Schnee Altholz Schnee Altholz Schnee Altholz Schnee bei 1000j. Tanne Schnee bei 1000j. Tanne Schnee Langboden Schnee Mitte Rotmösel Schnee Mösern Schnee Mösern Schnee Mösern Schnee Mösern Schnee Sonde Schnee Sonde Landböden Schneegrube Schneegruben lacke
Max Min Median MW n Sampling site
0.07 0.02 2.55 0.01 0.03 0.03 0.03
2.55 0.01 0.03 0.19 17 Ptot mg l−1
0.03 0.01 0.01 0.01 0.04 0.02
5.97 0.24 6.04 0.03
6.14 6.33 6.51 6.80 6.20 6.54
6.74 0.06
6.41 0.01
6.37 0.04
6.09 5.54 7.22 * 6.16 6.60 6.01
7.22 4.73 6.27 6.49 16 pH
0.20 0.02
0.64 0.19 0.14 0.37 0.21 0.20
0.09
0.14
0.21
0.64 0.01 0.20 0.22 17 (NO3 + NO2)-N mg l−1 0.18 0.02 0.01 0.48 0.26 0.22 0.26
0.54 0.21
0.00 0.00 0.01 * 0.23 0.00
0.00
0.15
0.25
0.11 0.04 0.25 0.00 0.01 0.01 0.00
0.54 0.00 0.01 0.11 16 NH4-N mg l−1
Table 1.1 Chemical composition of melted snow from Rothwald virgin forest
0.36 0.97
0.51 0.25 0.74 0.71 0.61 0.57
0.52
0.62
0.58
0.41 0.42 0.43 0.59 0.51 0.53 0.49
0.97 0.25 0.53 0.55 17 N-anorg mg l−1
2.19 0.54
1.44 0.36 0.04 0.65 0.69 0.01
0.01
0.12
0.61
1.04 0.74 23.60 0.71 0.93 0.49 0.37
23.60 0.01 0.61 1.97 17 Ntot mg l−1
1.05 1.03
0.81 0.18 0.16 0.91 0.55 0.34
0.36
0.43
0.54
1.02 1.18 1.20 0.49 0.39 0.27 0.37
1.20 0.16 0.49 0.60 17 Cl mg l−1
1.38 0.61
0.72 0.05 0.07 1.07 0.87 0.62
0.25
0.12
0.33
2.00 2.85 2.79 0.87 0.20 0.11 0.77
2.85 0.05 0.62 0.80 17 SO4 mg l−1
0.30 0.70
0.73 0.30 0.26 0.77 1.82 0.73
0.18
0.36
0.73
0.40 0.89 1.37 0.55 0.26 0.55 0.55
1.82 0.18 0.55 0.65 17 Ca mg l−1
0.06 0.11
0.06 0.02 0.02 0.23 0.12 0.05
0.01
0.30
0.13
0.10 0.14 0.25 0.03 0.03 0.03 0.08
0.25 0.01 0.06 0.08 17 Mg mg l−1
0.49 0.48
0.49 0.03 0.04 * 0.30 0.09
0.56
0.30
0.19
1.40 0.57 1.09 0.16 0.53 0.03 0.08
1.09 0.03 0.30 0.34 16 Na mg l−1
0.20 0.71
0.20 0.03 0.04 0.21 0.29 0.04
0.01
0.14
0.34
0.09 0.37 0.68 0.07 0.05 0.08 0.07
0.71 0.01 0.14 0.21 17 K mg l−1
0.27 0.78
0.04 0.02 0.02 * 0.23 0.02
0.01
0.01
0.06
0.12 0.06 5.06 0.04 0.04 0.01 0.01
5.06 0.01 0.04 0.42 16 Fe mg l−1
0.01 0.03
0.05 0.07 0.10 * 0.08 0.08
0.00
0.00
0.19
0.02 0.00 0.26 0.07 0.01 0.02 0.06
0.26 0.00 0.06 0.06 16 Mn mg l−1
0.00 0.00
* 0.03 0.00 * 0.00 0.00
0.04
0.00
0.00
0.00 0.00 0.35 * 0.06 0.05 0.00
0.35 0.00 0.00 0.04 14 SiO2 mg l−1
0.35 0.96
0.49 0.17 0.69 0.69 0.61 0.52
0.50
0.59
0.55
0.40 0.42 0.43 0.57 0.49 0.50 0.44
0.96 0.17 0.50 0.53 17 NO3-N mg l−1
1 Criteria for Sustainable Restoration of the Landscape
7
evaporated from the virgin forest in relation to precipitation brought from longer distances away, as such precipitation water would usually have approximately ten times higher conductivity. At the same time nutrients and minerals were being kept in place, bound within the debris and biomass of the decomposing organisms such as bacteria and fungi. The losses from sites within the virgin forest were thus minimised by the uncoupled water and matter cycles.
1.2.2 The Play Rules of Nature The study of the processes in the virgin forest has revealed how nature arranges itself to reduce losses and optimise production. The intricate strategies of nature can be demonstrated by tree seedlings benefiting from old fallen trees where they find optimum conditions for their growth. In an unmanaged forest, the renewal of trees usually takes place on the upper sides of trunks of old fallen trees (Fig. 1.2). Here the period without snow cover during which the seedlings can grow is prolonged and also the tree seedlings benefit from the supply of nutrients and minerals made available from the decomposing dead organic matter of the fallen trunks. The rapid changes occurring in the oxidising and reducing conditions on the upper sides of tree trunks – which also means a higher level of microbial activity – offer far better conditions and an improved nutrient supply for the growth of the seedlings. Ecosystems in general are controlled in a feedback mode by their physical limitations in interaction with the water balance. The water table oscillations within the
Fig. 1.2 Tree seedlings find best conditions for their development on old fallen tree trunks
8
W. Ripl and M. Eiseltová
debris layer are mainly controlled by plants through evapotranspiration. While solar energy, water and nutrients are the prerequisites of primary production, water uptake is controlled by the activity of plants’ roots and by evapotranspiration. Root processes and activity regulate nutrient availability through decomposition of organic matter (see below) and thus primary production. It is therefore most likely that the vegetation cover controls, by means of water uptake and evapotranspiration, the irreversible matter losses, soil fertility and sustainability of the environment. The water table oscillations within the root zones regulate the redox conditions and thereby the decomposition of debris and transformation of humic substances into mineralised compounds of ionic state that enable plant growth and plant production. Loss-free, localised recycling of matter and water within a minimised area ensures environmental stability with respect to both plant growth and energy dissipation controlled by evapotranspiration and condensation of water – fully utilising the capillary structures of fibrous debris. Driven by the sun’s irradiation water is cycling continuously and is a key in energy dissipation and cycling of matter. As water has a great capacity for carrying energy (i.e. high heat capacity), it makes it a very efficient heater or cooler. When water is evapotranspired by vegetation or evaporated from surfaces to the atmosphere, i.e. it changes from the liquid state to the water vapour, energy is stored in the form of latent heat in the water vapour and the site is cooled down. At the time and place of water condensation, energy is released and the site warmed up. Without water, the energy of the sun’s radiation would be transformed into sensible heat and the site overheated in the day time and far cooler at night (the conditions known from the desert areas where, due to the lack of water, the differences between the day and night temperature may exceed 50°C). The attenuation of an energy pulse to a mean level which can be seen as reduced temperature amplitudes is a result of energy dissipation by water (Fig. 1.3). The dissipation or attenuation of the daily solar energy pulse received by the Earth is essential to smooth the temperature gradients between day and night thus creating conditions for high species diversity as only very few species are adapted to big differences in day and night temperatures.
1.2.3 The Dissipative-Ecological-Unit The concept of the minimum unit system, termed the dissipative-ecological-unit (DEU), which has all the elements required for the efficient dissipation of energy, and water and matter recycling, has been introduced by Ripl and Hildmann (2000); and it depicts the strong association between the water retention capacity and detritus accumulation (Fig. 1.4). Steadily increasing resource stability of the DEUs is achieved by reducing water percolation through the soils to the groundwater and
1 Criteria for Sustainable Restoration of the Landscape
9
Fig. 1.3 Evapotranspiration and condensation of water vapour play an important role in the attenuation of solar energy pulse, thus reducing temperature amplitudes
instead increasing and short-circuiting the local water cycling within ecosystems by enhancing evapotranspiration. An indication of the different efficiencies of various habitats (DEUs) to attenuate the daily energy pulse has been shown by high time-resolution temperature recording (every 20 min) in three different ecosystems – woodland, grassland and arable field (Ripl et al. 1995, 1996). The calculated standard deviation clearly shows that the daily amplitudes of temperatures measured in meadows and arable fields are far higher than those in the woodland. Lower efficiencies of agricultural areas as compared to wooded areas were pointed out already at the end of nineteenth century by Müttrich (1890) who also showed the differentiation between different tree species (Fig. 1.5). The concept of the dissipative-ecological-unit is further important to demonstrate how nature, when not disturbed by sudden changes in climatic conditions, tends to close cycles of matter, i.e. run an efficient local resource economy, and maintain relatively even temperature and moisture conditions, i.e. environmental conditions without big fluctuations that best suit a vast majority of organisms. In contrast, human society seems to have a long way to go to realise that nutrients and base cations, and also water are becoming a limited resource that must be recycled. Our intensive land use pattern with no or little consideration of its impact on water and matter cycles leads to a higher frequency of large floods but also is a cause of the land drying out, soil fertility failing, and clean water becoming a scarce resource (see also Kravčík et al. 2007; Ripl and Eiseltová 2009).
10
Fig. 1.4 The dissipative-ecological-unit
W. Ripl and M. Eiseltová
1 Criteria for Sustainable Restoration of the Landscape
11
Fig. 1.5 Daily temperature reductions in three forest types compared to open field (based on data from Müttrich 1890)
1.3 The Stör River Catchment as an Example of Human Interference with Natural Processes 1.3.1 Site Description and Monitoring Programme An intensive monitoring programme was set up during 1991–1994 in the Stör River catchment (an area of 1,155 km2, in Schleswig-Holstein, NW Germany) in order to gain a detailed understanding of the processes responsible for the very high matter losses from this predominantly agricultural catchment. The land (at an altitude of 6–90 m asl) is dominated by agriculture (72%, 50% of which is arable and 22% covered by meadows and pastures) with some forestry (15%). The parameters measured at various points throughout the catchment included: total runoff (discharge) and dissolved chemical load (using both chemical analysis and conductivity measured in streams) from catchment and sub-catchments; and temperature measurements (both at a micro-habitat level and with the use of satellite information). The data from high time-resolution (every 20 min) ground measurements
12
W. Ripl and M. Eiseltová
were evaluated in connection to data of relatively high spatial resolution from satellite imagery. A detailed description of measurements performed and methods used is given in Ripl and Hildmann (2000). The data obtained from ground measurements, i.e. dissolved matter flow, run-off and 24-h temperature course were coupled with the data from satellite imagery showing the spatial temperature distribution of different land-use types within the catchment. This approach proved to be suitable to identify sites and land use types with the highest losses of dissolved matter (Ripl et al. 1995, 1996).
1.3.2 Matter Losses The average losses of dissolved mineral ions measured within the Stör River catchment were alarmingly high, about 1,050 kg of mineral salts ha−1 year−1 (excluding NaCl). Of this, the highest amount contributed base cations, Ca2+ 263 kg, K+ 27 kg and Mg2+ 19 kg (Ripl and Hildmann 2000). In comparison, the losses of nitrogen and phosphorus were relatively low – about 21 kg ha−1 year−1 for N and 0.6 kg ha−1 year−1 for P. Furthermore, atmospheric input counterbalances the losses of nitrogen (Ripl and Hildmann 2000). The calcium input in fertilisers over the Stör River catchment is only 80–90 kg ha−1 year−1 (Statistisches Landesamt Schleswig-Holstein 1989, cited in Ripl and Hildmann 2000) which means that calcium is being depleted from the soils as the losses cannot be sufficiently quickly compensated by rock weathering.
1.3.3 Disturbance to the Matter Flow and Temperature Damping In most agricultural areas, the damping mechanism of nature has been disturbed by our management, a management which has little understanding of the exact mechanisms and organism assemblages that regulate the decomposition of organic matter to get optimum yield under sustainable conditions. The destruction and replacement of short-circuited cycles within natural vegetation cover, the drainage of vast areas of farming land and the straightening and deepening of rivers and streams to transport water as fast as possible to the sea, have caused our landscape to be highly overheated on summer days – i.e. daily temperature attenuation has been reduced and the dissipation of solar energy in space and time has been largely degraded. The results from the Stör project have shown that the overheated areas usually correspond to the areas of the highest matter losses (as long as those areas still have relatively abundant soluble matter) (Ripl et al. 1995). The drainage of agricultural land which has been practised elsewhere for more than 40 years has resulted in the reduced fertility and reduced water-retention capacity of soils. High water-table oscillations, land surface overheating, increased mineralisation of organic matter and subsequent matter losses (with respect to nutrients and minerals) are the consequences of such agricultural practices. Natural systems are usually highly complex systems and any intervention by humans to
1 Criteria for Sustainable Restoration of the Landscape
13
their structure may cause substantial change in their functioning. Any intervention should therefore be considered from a holistic viewpoint.
1.3.4 Disturbance of the Natural Water Flow Dynamics The dynamics of rivers and streams is directly linked to the dynamics of the runoff and catchment topography. Any changes in the slope, substrate or river cross-sectional profile, as well as different retention capacities of the adjacent landscape have an influence on the acceleration and/or delay of the water flow. The hydraulic engineer’s measure of a river’s efficiency is the hydraulic radius which expresses the capacity of a river’s profile to carry water, i.e. the ratio of cross-sectional area to wetted perimeter. The greater the hydraulic radius the lower the resistance, the faster the flow and the more ‘efficient’ a river. From the thermodynamic point of view, however, the river’s efficiency to dissipate the energy of water is more important. The form factor expresses the reverse of the hydraulic radius and was defined in relation to a semi-circular river cross-section (the least resistant form to water movement) – higher the value of form factor broader and more shallow the river section, hence higher the resistance to water movement and the greater the dissipation of the energy of water. The plant form factor was later added to incorporate the surface area of submersed vegetation to the wetted perimeter – each species being designated a certain value for a given density. The development and distribution of morphological features found in running waters are the result of energy dissipation process. Dynamic changes in stream morphology, such as the stream bed profile, are determined by the dynamic potential of the runoff from the land. High fluctuations in catchment runoff cause temporary but extensive shifts of sediment material. A more regular runoff leads to a spatial gradient with respect to the turbulence distribution and results in a grading of material in increasingly smaller shifts of material. In a self-structured river bed water movement will be ‘smoothed out’ through the development of localised current gradients. Transport of material (suspended and bed load) is consequently slowed down. The emergence of morphological structures in flowing waters (e.g. rifflepool system, plant surfaces) can be considered as a feedback, optimisation, process that tends to maximise the delay of material being transported and thus helps to reduce matter losses from the landscape. It can be regarded as a process of selfoptimisation by the attenuation of energy. Present day river management preventing bank erosion through implementation of bank enforcement measures and constantly removing river sediments to enhance the streams discharge efficiency prevents the river from developing a dissipative structure that would correspond to the dynamics of the catchment runoff. This leads to highly uneven discharges in so disturbed rivers. The two contrasting hydrographs from the Stör River catchment – Osterau and Dosenbek streams (Fig. 1.6) – show the impact of agricultural land use on the stream’s discharge regime (Dosenbek) compared to the stream that flows through forest and meadows (Osterau). Accentuated by the reduced water retention capacity of soils within the river catchments, the streams are
14
W. Ripl and M. Eiseltová
Discharge hydrograph of the Osterau 4,5 4
Discharge m3/s
3,5 3 2,5 2 1,5 1 0,5 0 01.11.92 31.12.92 01.03.93 30.04.93 29.06.93 28.08.93 01.12.92 30.01.93 31.03.93 30.05.93 29.07.93 27.09.93
Date Discharge hydrograph of the Dosenbek 1 0,9 0,8
Discharge m3/s
0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 01.11.92 31.12.92 01.03.93 30.04.93 29.06.93 28.08.93 01.12.92 30.01.93 31.03.93 30.05.93 29.07.93 27.09.93
Date Fig. 1.6 Hydrographs from two rivers of the Stör river catchment (Osterau and Dosenbek)
1 Criteria for Sustainable Restoration of the Landscape
15
prone to floods at times of high precipitation or very low discharges at times of prolonged droughts. The hydrological regime has been substantially impaired by human activities and the consequences are clearly visible: high fluctuations in stream water discharge, high irreversible losses of life-supporting matter (nutrients and base cations), reduced soil fertility, extreme floods but also droughts. We can say that the quality of land management is best viewed in the river. The quantity and quality of water in the river reflect all actions taking place in the entire catchment. As Falkenmark et al. (1999, p. 33) stressed: ‘If there is an ailing river, a sick landscape may be the cause.’ This statement also indirectly suggests that restoration should start on the land and not in the river.
1.4 Restoring Landscape Sustainability 1.4.1 The Rules for River Authorities and Water Managers From the above it is clear, that any river management measure that stops or hinders the river’s self-structuring process and development of vegetation within or along the river bed, leads to the increased losses of matter and reduces the overall efficiency with respect to energy dissipation of the system. However, the question remains as to what extent restored in-channel structures can cope with changes in the runoff and highly accelerated matter losses from the catchments. Attempts to restore streams and rivers to their ‘natural state’ using only in-channel structural modifications are bound to failure (see also Section 10.3.1). With strongly irregular peaks in runoff from the catchment, large quantities of both dissolved and suspended matter enter the river. These inputs are usually outside the control span of the river processes themselves, i.e. the addition of retention surfaces, for example, in form of added stones for the development of periphyton, etc. are not sufficient to reduce the high matter losses coming from agricultural catchments. Every river restoration process should, therefore, start on land as it is the vegetation cover (with its evapotranspiration) within the catchment that will best help to smooth out the runoff and encourage more regular hydrograph. A general rule for river bed restoration itself would be to give river more space and freedom to return to a more natural state that would best correspond with the river’s discharge pattern. More scientific background and practical experience with stream restoration is given in Chapter 10.
1.4.2 The Rules for Land Managers Two contrasting situations have been described above – the optimised situation in the virgin forest of Rothwald – an example of sustainable system, and the humandisturbed agricultural system of Stör where climate-damping elements have been mostly removed and the system is far from being sustainable.
16
W. Ripl and M. Eiseltová
From the above, it is clear that sustainable land management should be a management that maintains high evapotranspiration in the landscape. Farmers and other land managers should seek to maximise evapotranspiration from their land. Maximising evapotranspiration brings about minimised losses of matter via rivers to the sea – losses that are irreversible for the farmers (and society). In drained landscapes, water is transported to the groundwater as fast as possible, and large areas of top soils are turning into unsaturated soil zones and drying out. With the subsequent lowering of the water table, decomposition of soil organic matter is maximised under aerobic conditions. Strong acids are produced by the oxidation of sulphurous and nitrogen-containing compounds in the organic matter, leading to easily dissolvable salts in the soil. In autumn (when the evapotranspiration is lower), the water table rises and water pressure increases again and the increased runoff takes the dissolved salts to the rivers. The result is a steady reduction of soil fertility and an ever-increasing need for fertilising, and above all, the loss of sustainable development for the whole of society. The limiting elements in agriculture are water and nutrients. The recycling of water and nutrients is therefore essential and has to be locally optimised. Water converted into water vapour is distilled, cleaned and coming back to the fields when fields are kept cooler than surrounding areas by cooling due to evaporation enhanced by plants with high surface to volume ratio (leaf index). This makes it necessary to maximise evapotranspiration and thereby minimise losses of matter to the rivers. On farmland we should produce as much of distilled water, energy and food as we need for local consumption and not to produce more for long-distance sales. The production should be done in a cyclic mode and the cycles should be as shortcircuited as possible. The aim should be to minimise the difference between the temperatures of day and night and to reduce the concentrations of plant usable matter in the rivers and transport systems. Figures 1.7–1.9 show three different states of landscape development: (a) landscape with intact functionality which depicts natural landscape development that tends towards sustainability (Fig. 1.7); (b) landscape degraded by humankind where natural resources such as soil, water and vegetation are overexploited (Fig. 1.8); and finally (c) landscape where the criteria of sustainable management are implemented (Fig. 1.9). As we stressed already, to restore landscape sustainability it is essential to return more vegetation into the catchments and restore water retention capacity of soils in order to recreate the necessary ‘cooling spots’ that are essential to efficiently dissipate the solar energy through evapotranspiration. This will help close water and matter cycles, and thus minimise the transport of irreversible matter losses and attenuate the hydrograph. The principal measures that should contribute towards restoring the cooling function by increased evapotranspiration and the re-creation of closed water and matter cycles are as follows: 1. As a priority upper parts of catchments should be covered with well-established vegetation cover (such as mixed forests with little or no management at nearzero net production) as these areas are particularly sensitive to leaching of soils and overheating.
1 Criteria for Sustainable Restoration of the Landscape
17
Fig. 1.7 Schematic diagram of a landscape with intact functionality, short water and matter cycles and hence sustainable in a long-term
Fig. 1.8 Schematic diagram of a landscape degraded by humankind, landscape with high matter losses, high ground water level fluctuations and overheating speeding up the landscape’s ageing
18
W. Ripl and M. Eiseltová
Fig. 1.9 Schematic diagram of a landscape where the criteria of sustainable management are implemented
2. In water-source areas, springs and at the confluence of rivers, wetlands (such as the original fens and bogs) should be restored. Wetlands in these areas are particularly important as hydrological buffers to even out irregular peaks in discharge hydrographs, thus reducing the possibility of flood events and retaining higher water flows in summer. 3. Riparian zones along rivers should be restored and managed as buffer/retention zones in which surface and subsurface water flow from the catchment will be retarded by evaporation. Wetlands, such as reedbeds and floodplain forests, along water courses help to retain part of the dissolved material by precipitation and absorption by vegetation. Biomass from the floodplain can also be harvested and used as raw material to produce silage, energy, building material, etc. The retained bases and nutrients can be used to improve soil fertility in higher parts of the catchment, thus helping to close the matter cycles.
1.5 Assessment of Catchment Sustainability – Monitoring Approach As we described above, in nature the system is closing the cycles in smaller and smaller parts of the landscape and thereby increasing its diversity and reducing transport and sedimentation in lakes and rivers. The natural development of larger catchments during evolution was successively covering the whole catchments with
1 Criteria for Sustainable Restoration of the Landscape
19
vegetation and water cycles were short-circuited and matter cycles practically closed. The reduced sedimentation rates in lakes with the development of a natural vegetation cover have been described by Ripl (1995) using the data from Digerfeldt (1972). Based on the current scientific knowledge, the criteria for assessing landscape sustainability proposed are: the efficiency of an ecosystem to recycle water and matter and its efficiency to dissipate energy. Then the achievements of individual land managers (as farmers or foresters) with respect to sustainable management of their land can easily be evaluated by: 1. Continuously measuring conductivity (see below) and flow-rates in streams as a basis for estimating closed or open processes in the catchment or sub-catchment. 2. Regular evaluation of satellite pictures with regards to temperature distribution over the catchments. The damped temperature of the landscape surface is a result of short-circuited water cycle and a prerequisite for localised matter cycles and hence minimised matter losses. Conductivity is a suitable parameter to estimate dissolved load from catchments or sub-catchments as it is relatively easy to monitor on a regular basis and it shows a rather close correlation with total mineral salt concentrations (Fig. 1.10).
Fig. 1.10 Correlation between total mineral salts and conductivity measured in the Stör River catchment
20
W. Ripl and M. Eiseltová
Evaluation of the thermal channel of Landsat satellite pictures showing the relative temperature deviations demonstrates the impact of largely de-watered areas, such as urban areas and areas of drained agricultural land, on energy dissipation (Ripl et al. 1995, 1996). Such a monitoring approach was successfully tested in the Stör river catchment and Rothwald virgin forest (both described above; see also Ripl et al. 1995, 2004; Ripl and Hildmann 2000) and also in three small sub-catchments in South Bohemia (Pechar et al. 2005).
1.6 More Responsible Towards the Future 1.6.1 Recapitulation of the Current State of Knowledge The progress in scientific understanding of natural processes made in the past 50 years enables us to make informed decisions and implement sustainable management of our landscapes. The observed escalating problems, such as the critical state as with regards to surface as well as groundwater pollution, increased fluctuations in river discharges bringing about severe floods and prolonged droughts, depleting natural resources, lost or severely reduced soil fertility, and humans’ interference with the climate, all call for urgent change in land use. The complex research and monitoring programmes undertaken in Rothwald virgin forest and Stör river catchment have offered us a reference for sustainable (Rothwald) and non-sustainable (Stör) landscapes. In most agricultural landscapes the sole criteria of maximum production and maximum profit were adopted discarding the criteria of long-term sustainability. We have shown that the destruction of natural vegetation cover over large areas and large-scale drainage that has been practiced have had severe consequences in terms of reduced evapotranspiration and hence solar energy dissipation, and increased losses of matter to the groundwater and rivers. In the Stör River catchment the losses measured reached about 1,500 kg ha−1 year−1 of dissolved inorganic matter while the area of the virgin forest showed losses in the order of 5–15 kg ha−1 year−1 (mainly with the snow melt water in the spring). There was practically no runoff observed from the virgin forest during summer as all water went to evapotranspiration. The debris layer in the virgin forest acts as a permanent moisture-containing sponge buffering the needs for evaporable water and is practically impenetrable for the water flow to the groundwater. This shows that the resource economy with respect to bases and nutrients is optimised by localised recycling of matter needed for the plant growth. The frequency of water cycle in the virgin forest is approximately 1 day compared to an average water cycle elsewhere of 10–12 days. The high evapotranspiration resulted also in a practically perfect temperature damping and even temperature distribution. The secret of sustainable development of the landscape is: high
1 Criteria for Sustainable Restoration of the Landscape
21
evaporation, local cycles, and maximum temperature damping – all leading to a prolonged phase of a landscape being able to carry vegetation and support life as we know it at present.
1.6.2 Radical Changes in Landscape Management Needed We have provided evidence that manipulation with the vegetation cover has an adverse impact on water quality and affect water partitioning into evapotranspiration and runoff. The data obtained have confirmed that to achieve sustainable land use we need to decouple the water flow from the matter flows as the elevated losses of matter are connected to water runoff. Furthermore, the highest matter losses originate in ‘overheated spots’ where there is faster degradation of organic matter, the retention of nutrients is low and nutrient leaching is therefore enhanced. The overheated spots correspond to areas of reduced evapotranspiration, i.e. areas where natural vegetation cover was destroyed, replaced by crop monocultures, and, in most cases, the land was also drained. It is not, however, only the arable land that is prone to elevated leaching of nutrients and soil erosion, but equally harmful erosion problems are known from grasslands exposed to overgrazing. The result too is the disturbance of the water cycle and largely increased oscillations of the water table. From all the above, it is evident that to achieve sustainable land use and sustainable landscape development we need to reduce irreversible losses of matter to a minimum while maintaining high primary productivity. We need to keep water in the landscape and support its short-circuited cycling. Management of water is a key in the endeavour to restore sustainable landscapes. Land and water management have to aim at closing water and matter cycles. Responsibility for water management should be in hands of land managers. The limitation of plant growth by water, occurring almost everywhere, can only be overcome by intelligent distribution of crop production and natural vegetation over the landscape (see Fig. 1.9). This is probably the only approach to repair the highly degraded water cycle and its function in climate optimisation, and preventing further landscape degradation and drying up. Energy, water and land management should be integrated. The land managers, including farmers, should be responsible for ensuring sufficient quantity and quality of water and also take care of producing energy for local needs (i.e. renewable energy produced, for example, from fast-growing trees). Intelligent distribution of crop (food) production and energy production over the land, supporting maximum evapotranspiration will help close water and matter cycles and reduce the humaninduced climate change. Sustainable land management shall be awarded and incentives paid to land managers according to their achievements in closing the matter cycles thus reducing water pollution, and closing the water cycles which would benefit the local climate. Incentives should be awarded on the basis of real
22
W. Ripl and M. Eiseltová
achievements – temperature damping and minimised losses of matter from the land – this would be regularly evaluated using the information from the monitoring programme described in Section 1.5.
1.6.3 Responsible Politicians A similar development to that in nature can also be found in the development of our human societies. Initially, we were surrounded by abundant natural resources – be it fossil fuels, extensive forests, clean water and air, and fertile soils. Accordingly, we have adopted the development mode of ‘r-strategists’. This has, however, led to severe limitations and to the overexploitation of natural resources in many parts of the world. It is, therefore, necessary to accept that we are now in a transition period to a new phase. Non-renewable resources are quickly becoming depleted, initially extensive primary forests have been largely destroyed, clean water has become scarce and soil fertility largely reduced. Further survival and development of our societies will require a change from a production-oriented economy to a recycling-oriented communication-and-socialisation-related economy in order to achieve sustainability. Sustainable management is never a random management. Instead any wise land management must be a locally-adapted management that is continually improved by feedback mechanism. This is also the reason that all laws (demanding universality) concerning agriculture and nature protection have the disadvantage of not reacting to and adopting the feedback mechanisms of natural processes. What we need to do better is to improve and to regulate the local processes at the interfaces between the living green and the substrate soil, putting the responsibility for sustainable land management in hands of farmers and other land managers and paying them accordingly. If sustainable development is the ultimate goal of our society, than the quality of governments and politicians should be evaluated based on their responsible management of three basic resources – the tools to be used are education, science and framework laws to improve and manage. The three basic resources are: 1. The landscape interacting with the atmosphere with respect to spatiotemporal distributions (providing water, energy, vegetation, food, daily needs) 2. The energy dissipative properties of the landscape (evapotranspiration-condensation of water, dissolution-precipitation of soil-solids, production-respiration of organic matter) partitioning solar energy and structuring matter in the managed area 3. The intelligence of mankind with respect to managing spatiotemporal phases of landuse (space) and energy-partition (time) in order to improve efficiency (closing cycles) and thereby minimising openness (losses) and randomness Our policies must be changed to aim, as a top priority, at land and water use that is sustainable in the long-term. Our politicians must be guided in their decisions by long-term benefits to society rather than aiming at short-term gains.
1 Criteria for Sustainable Restoration of the Landscape
23
1.6.4 Implications for Climate Change The above measures need to be implemented in order to achieve sustainability of land and water management and to counteract climate change. Our focus must turn from CO2 and industry to land and water management, and the conservation of living structures thus providing the atmosphere, climate, water, energy, food and soil. Water vapour with an extremely short turnover and its distribution in the atmosphere is a far more important proxy for climate in comparison to CO2 in the atmosphere. ‘IPCC science’ is providing the whole world with the wrong key for climate-change problems. It is high time to change our views if we want to counteract global warming and support sustainable ecosystems. We have provided the evidence that the lack of water in the landscape and its highly uneven distribution in time (in the case of the Stör River catchment) is the cause of landscape overheating. By contrast, the much more even distribution of temperature that was found in the case of the Rothwald forest was due to the watersaturated conditions in the forest. Water (including water vapour) is a highly efficient distributor of heat – both cooling and heating – and can ensure the even distribution of temperature and hence have a positive impact on local climate. The existence of natural ‘cooling structures’, i.e. natural vegetation with its evapotranspiration function, within the landscape is therefore most important, and such a distribution of land use should be implemented as to provide a more even distribution of temperatures.
References Digerfeldt G (1972) The post-glacial development of Lake Trummen. Regional vegetation history, water level changes and palaeolimnology. Folia Limnologica Scandinavica 16:104 Falkenmark M, Andersson L, Castensson R, Sundblad K (1999) Water – a reflection of land use. Options for counteracting land and water mismanagement. Swedish Natural Science Research Council, Stockholm, 128 pp Kravčík M, Pokorný J, Kohutiar J, Kováč M, Tóth E (2007) Water for the recovery of the climate – a new water paradigm. Ludia a voda, 122pp. Available from: https://www.waterparadigm.org Mac Arthur RH, Wilson EO (1969) The theory of Island Biogeography. Princeton University Press, Stockholm, 203 pp Müttrich A (1890) Jahresbericht über die Beobachtungsergebnisse de rim Königreich Preussen, des Königreichs Württemburg, des Herzogthum Braunschweig, der thüuringischen Staaten, der Reichslande und dem Landesdirectorium der Provinz Hannover eingerichteten forstlichmeteorologischen Stationen. Fünfzehnter Jahrgang: Das Jahr 1889, Berlin (in German) Pechar L, Procházka J, Hais M, Pecharová E, Eiseltová M, Bodlák L, Sulcová J, Kröpfelová L (2005) Assessment of landscape efficiency in matter retention in submontane agricultural catchments of the Czech Republic. In: Dunne EJ, Reddy KR, Carton OT (eds) Nutrient management in agricultural watersheds – a wetlands solution. Wageningen Academic Publishers, The Netherlands, pp 34–38 Prigogine I (1980) From being to becoming: time and complexity in the physical sciences. Freeman, San Francisco, CA, 272 pp
24
W. Ripl and M. Eiseltová
Prigogine I, Stengers I (1984) Order out of chaos. Bantam Books, New York, 349 pp Ripl W (1995) Management of water cycle and energy flow for ecosystem control – the energy-transport-reaction (etr) model. Ecol Model 78:61–76 Ripl W, Eiseltová M (2009) Sustainable land management by restoration of short water cycles and prevention of irreversible matter losses from topsoils. Plant Soil Environ 55(9):404–410 Ripl W, Hildmann C (2000) Dissolved load transported by rivers as an indicator of landscape sustainability. Ecol Eng 14:373–387 Ripl W, Hildmann C, Janssen T, Gerlach I, Heller S, Ridgill S (1995) Sustainable redevelopment of a river and its catchment: the Stör River project. In: Eiseltová M, Biggs J (eds) Restoration of stream ecosystems – an integrated catchment approach. IWRB Publishing, Slimbridge, pp 76–112, Publ. No. 37 Ripl W, Janssen T, Hildmann C, Otto I (1996) Entwicklung eines Land-Gewässer Bewirtschaftungskonzeptes zur Senkung von Stoffverlusten an Gewässer (Stör-Projekt I und II). In Zusammenarbeit mit F. Trillitzsch, Backhaus R, Blume H-P, Widmoser P (eds) Im Auftrag des Bundesministeriums für Bildung, Wissenschaft, Forschung und Technologie (BMBF) und des Landesamtes für Wasserhaushalt und Küsten Schleswig-Holstein. Förderkennzeichen 0339310A und 0339538. Endbericht. Technische Universität Berlin, Fachgebiet Limnologie. 203 pp. + Anhang. http://www.aquaterra-berlin.de/index.php/ component/content/category/37.html Ripl W, Splechtna K, Brande A, Wolter KD, Janssen T, Ripl W jun, Ohmeyer C (2004) Funktionale Landschaftsanalyse im Albert Rothschild Wildnisgebiet Rothwald. Im Auftrag von LIL (Verein zur Förderung der Landentwicklung und intakter Lebensräume) NÖ Landesregierung, Österreich, Final Report, 154 pp Schneider ED, Kay JJ (1994) Life as a manifestation of a second law of thermodynamics. Math Comput Model 19(6–8):25–48
Chapter 2
The Evolution of Lakes and Wetlands Sven Björk
Abstract When dealing with lake restoration, we should bear in mind that inland water ecosystems are not static units but subject to continuous evolution. Palaeolimnological studies have helped us to understand the development of lakes and their catchments. In particular, the development of northern European lakes has been studied very thoroughly. It has been revealed that the initially high productivity of lakes (shown by sediment growth rate) was due to the leaching of nutrients from the nutrient-rich moraine after the last deglaciation. With time, however, lake productivity dropped as the supply of nutrients from the catchment area diminished. This reduction depended partly on the decreased leaching and partly on the development of terrestrial vegetation as it accumulated and recycled nutrients. In southern Sweden, the current sediment growth rate is about 0.2 mm per year in more-orless-intact, shallow oligotrophic lakes, and about 0.5–1.0 mm per year in shallow eutrophic lakes. If a lake becomes polluted by the discharge of nutrient-rich sewage, the sediment growth rate can increase to about 10 mm per year. The ageing of lakes, and their potential terrestrialisation, depends on the balance between production and decomposition of organic matter. In northern latitudes, the break-down of organic matter in cold and oxygen-free sediment and peat is much slower than in warmer waters in the south where mineralisation processes take place at higher rates and over a longer period of the year. It is therefore much harder to prevent lakes in the north from being terrestrialised. Keywords Lake evolution • Northern European lakes • Palaeolimnology • Sedimentation rates • Terrestrialisation
S. Björk (*) Department of Ecology, Limnology, University of Lund, SE-223 62 Lund, Sweden e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_2, © Springer Science+Business Media B.V. 2010
25
26
S. Björk
2.1 Postglacial Evolution of Lakes and Wetlands Inland water ecosystems are not static units but subject to continuous evolution. The speed of these changes is high in shallow, productive accumulation basins and extremely slow in deep, large, oligotrophic lakes. Palaeolimnological studies on ecological succession clearly demonstrate how important it is to take the time-factor into account when dealing with the restoration of systems that have been degraded and which then possess all the characteristics of rapidly ageing wetlands. From the perspective of geological time, shallow lakes and inland wetlands are short-lived ecological units which become filled up with inorganic and organic material. During, and immediately after, the latest deglaciation of northern Europe, layers of minerogenic matter (i.e. mineral in origin) were deposited in depressions. Coarse particle fractions settled close to the shore beside the mouths of feeder streams, while clay and other fine particles settled further lakewards. The landscape of northern Europe has been characterised by its original richness of lake and wetland basins produced by glacial action. By comparison, the region south of the glaciated area is poor in lakes. In northern Europe, former lake and wetland basins were cleaned of sediment and peat by glaciers; after deglaciation, starting from scratch, as it were, lakes developed in shallow moraine depressions with a minerogenic bottom. South of the glaciated area, tundra conditions had prevailed during the ice age; consequently, as a result of the frost erosion – cryoturbation – the topsoil layers resembled the moraine of the deglaciated region, as they were also largely unleached. Also, the processes in these soils, including the leaching of nutrients and the influence of the successively developing vegetation, corresponded to that found in the morainecovered, formerly-glaciated part of Europe. The postglacial (Holocene) relations between catchment areas and lakes are therefore comparable in both regions. Because the developmental history of northern European lakes is so well known, the following description refers to these systems (cf. Digerfeldt 1972; Digerfeldt and Håkansson 1993). As to the origin of lake basins see, for example, Hutchinson (1957).
2.2 Development of Northern European Lakes After deglaciation, the fresh moraine was generally rich in nutrients. These were subsequently leached and transported by water to the lakes, where they supported a high production of algae. In the sediment, the transition from minerogenic to successively more organogenic strata (i.e. organic in origin) reflects this course of events. In several lakes, the lowest and oldest organogenic sediment consists of algal gyttja deposited during a short period of high primary productivity in the lake but equally a low supply of organic matter from the surrounding areas. Even in regions which now have a typically oligotrophic character, the oldest layers can consist of algal gyttja and lake marl, sediment types in sharp contrast with the present-day leached and unproductive character of the lake surroundings.
2 The Evolution of Lakes and Wetlands
27
After this initial phase of high productivity, lake productivity dropped as the supply of nutrients from the catchment area diminished. This reduction depended in part on the decreased leaching and partly on the development of terrestrial vegetation accumulating and cycling nutrients. However, during the Holocene of the last 10,000 years, changes in temperature, precipitation and vegetation cover have caused variations in the productivity of lakes. Along with the climatic changes, the water level of lakes has also changed, and this, in turn, meant either a lakeward expansion of the littoral macrophytic vegetation or its landward retreat and reduction (Digerfeldt 1972). At the same time, the sedimentation limit, i.e. the highest level up to which organic particles settle below the shoreline, has been dislocated downwards (due to erosion) during periods of low water and upwards (because of deposition) during high water periods. In any individual lake, the location of the actual sedimentation limit is lower along those shores exposed to winds than along more sheltered shores. As prevailing winds are westerly in large parts of Europe during the ice-free season, the littoral zone is washed clean of organic particles down to a greater depth along eastern, wind-exposed shores than along sheltered, western shores where particles settle in more shallow, calmer water.
2.2.1 Growth Rate of the Sediment The rate at which sediment thickness increases is dependent on the productivity of the lake itself, on the supply of matter from the catchment area, and the efficiency of the mineralisation in the lake ecosystem. Intact lakes, within previously glaciated regions, had the highest productivity and sediment growth rate during the first phase of their evolution (cf. Digerfeldt 1972; Digerfeldt and Håkansson 1993). After that, these processes were reduced at the same time as the leaching and supply of nutrients from the surroundings decreased (Fig. 2.1). The thickness of sediment varies from lake to lake as well as within the individual lake depending on the topography of its minerogenic bottom. The deposition of organic matter ‘flattens’ the bottom such that the thickest sediment is found in the deepest depressions of the minerogenic bottom. The normal, average thickness of organogenic sediment in, for example, a south Scandinavian lake with an age of 12,000–13,000 years is about 5 m. In this region, the current sediment growth rate is about 0.2 mm or less per year in a more-or-less intact, shallow oligotrophic lake, and about 0.5–1.0 mm per year in a shallow eutrophic lake. If a lake becomes polluted by the discharge of nutrient-rich sewage, the balance between production and mineralisation is disturbed. Organic matter accumulates and the sediment growth rate can increase to about 10 mm per year. Thereby, the speed of ageing of a shallow lake is multiplied – as a result of the addition of an internal nutrient supply from the sediment added to that of the increased external loading (i.e. from the catchment area).
28
S. Björk
Deposition [mm year−1] 2.1 mm
1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
1.2 mm
1312
11
rec. poll. SA2 SA1 2.0
10
9 SB2
8 SB1
3.0
7
6 AT2
4.0
AT1
5 BO2 5.0
4
3
2
BO1
1 PB
6.0
Number of sediment sample Zones Depth (m)
Radiocarbon age B.P. 1,000 2,000 3,000
fitted polynomial radiocarbon dating (observed value)
4,000 5,000 6,000 7,000 8,000 9,000 10,000 11,000 Key Rec. poll. = recent pollution, SA = Sub-Atlantic period, SB = Sub-Boreal period, AT = Atlantic period, BO = Boreal period, PB = Pre-Boreal Period, B.P. = before present.
Fig. 2.1 Lake Trummen, Sweden. Upper Diagram: Rate of sediment deposition. (The indicated rates during the Preboreal (PB) and Early Boreal (BO 1) periods are mean values.) Lower diagram: Radiocarbon dating of the sediment (From Digerfeldt 1972)
2 The Evolution of Lakes and Wetlands
29
2.2.2 Primary Productivity and Vegetation Succession in Lakes and Wetlands In a naturally ageing lake, there is a successive decrease in the primary productivity based on plankton. However, a very dramatic change in the productivity takes place when the lake has become so shallow that it is possible for peat-forming macrophytes – such as the common reed (Phragmites), bulrush (Schoenoplectus) and cattail (Typha) – to colonise the organogenic bottom (Fig. 2.2). The plankton is then replaced by communities of microorganisms (periphyton) that develop in the water on the stems and leaves of emergent, floating-leaved and submerged macrophytes. The most productive phase in the whole evolutionary history of a lake is the period when the shallow lake has just been transformed into a wetland overgrown by emergent vegetation. The reasons for this sudden increase in productivity are as follows: • There has been a continuous supply of nutrients from the catchment area to the lake which has acted as a trap for these elements. • There is never a shortage of water in a wetland. • The perennial, highly-productive emergent macrophytes are well adapted to an efficient utilisation of the environmental conditions in all three media: sediment, water and air. Macrophytes, rooted in the sediment, can thus make use of the accumulated, nutrient-rich resources which until now, i.e., during the previous stages of ecosystem development, have not been available for deeply-rooted plants. Wetlands overgrown by plants such as Phragmites and other perennial, emergent macrophytes constitute the most productive ecosystems of any at the same latitude. In northern latitudes, the break-down of organic matter in cold and oxygen-free sediment and peat is much slower than in waters of warmer latitudes where mineralisation processes take place at higher rates and over a longer period of the year. This makes it especially troublesome to preserve northern, highly-productive wetlands characterised by high rates of accumulation of coarse detritus produced by the emergent macrophytes. The general ecological succession and terrestrialisation process includes various stages where Typha, Schoenoplectus and Phragmites are replaced by Carex, and the formation of peat above the sediment prepares for the invasion of Salix and Betula. A lowering of the water level in shallow lakes brings about a tremendous speeding up of the ageing process (Fig. 2.3).
2.2.3 Palaeolimnological Studies The composition of a lake’s sediment directly reflects the organism communities which have inhabited a lake and its catchment area over the millenia. Provided the sediment growth rate is known, it is possible, by means of chemical analysis, to
30
S. Björk
1. Lake with deposition of organogenic sediment (gyttja) on the minerogenic bottom.
2. The lake filled up with gyttja overgrown by peat-forming emergent macrophytes.
3. The macrophyte peat covered by sphagnum peat. During a period with dry climate the bog is colonised by Pinus.
4. The former lake completely transformed to a raised bog with fast-growing Sphagnum during a wet climate period. The bog surrounded by a fen (lagg), i.e. the zone influenced by water from the mineral soil. The raised bog totally dependent on precipitation and airborn matter (ombrogenous bog).
A
B
C
D
E
F
Key A = moraine, B = sand-slit, C = silt-clay, D=gyttja, E = peat of emergent macrophytes, F = Sphagnum peat.
Fig. 2.2 The transformation of a lake from an open water ecosystem to a raised bog
2 The Evolution of Lakes and Wetlands
31
Fig. 2.3 The effect of water level lowering in a shallow lake on the quantitative development of macrophytic vegetation. The location of the littoral barricade indicated (cf. Fig. 2.6)
estimate the supply and calculate the deposition of different elements during the various phases of development. By means of such palaeolimnological investigations, using a long series of analytical methods, the developmental history of both the lake and its catchment area can be reconstructed. The presence of frustules of diatoms, cysts and other remnants of algae, remains of insects and cladocerans, seeds and other diaspores, roots, etc., allows an ecological interpretation to be made of the conditions prevailing when various layers of sediment and peat were deposited. Pollen analysis, dating of sediments by determination of the content of radioisotopes of carbon (14C), caesium (l37Cs) and lead (210Pb), and palaeomagnetic investigations, are used to determine the age of sediment layers. When the age is known, the sediment growth rate and the accumulation of different elements per unit of time can be determined (cf. Digerfeldt 1972). This means that both natural and man-induced leaching and transport processes from the catchment area to the lake can be revealed by studying the sediment’s ‘archives’ (Fig. 2.4). Palaeolimnological investigations are often an essential pre-requisite for the design of restoration plans because the aim is usually to try to re-create ecosystem qualities which have been lost. Past and present relationships between the catchment area and the lake have to be elucidated and used as guidelines in
32
S. Björk 3.4 kg km−2 yr−1
5.4 m
1.5 kg km−2 yr−1
4.0 m
0.9 mm yr−1 −2
0.4 mm yr−1 −1
44 mg m yr
Early Boreal 9000 BP
20 mg m−2 yr−1
Late Atlantic 5800 BP
1.5 kg km−2 yr−1
2.5 m
2.0 m
0.2 mm yr−1 −2
8 mm yr−1 −1
20 mg m yr
Early Sub-Atlantic 1400 BP
100 kg km−2 yr−1
1,300 mg m−2 yr−1
Pollution Period 1940-1965
Fig. 2.4 Lake Trummen, Sweden. The transport of phosphorus (indicated by arrows) from the catchment area for deposition in the sediment during four periods. The figures within the lake denote: lake depth, rate of deposition, and the annual deposition of phosphorus (compiled on the basis of data from Digerfeldt (1972), from Björk et al. 1972)
restoration planning. With a collapsed ecosystem as the patient, and with a holistic approach in space and time as a necessity, the pre-project investigations need to be organised as team-work in order to gain an understanding of the system as a whole.
2 The Evolution of Lakes and Wetlands
33
2.2.4 Important Physical Processes The influence of man on lakes and wetlands often eliminates important natural agents such as ice movements and seasonal water level fluctuations. The temperature changes that took place during the various Holocene climate periods have involved variations in the development of lake ice cover. A general feature is that ice can have an influence on shorelines both when it covers the whole lake and when the ice breaks up. In cold winter periods, especially at night when the ice cools down and contracts, tension cracks appear in the ice cover. When cracking, a sound like thunder is heard – the ‘bellowing’ of Scandinavian lakes. The cracks are immediately filled with water which then freezes. In warmer periods, as on sunny days, the volume and area of the ice cover increases and a very strong pressure is exerted against the shorelines, which are subject to ice-pressing. With the changes in winter temperatures, this process is repeated continually and means a successive increase in ice pressure over the winter. The erosive zone of the shore (Fig. 2.5) which is influenced by the ice, can be cleaned of vegetation, bottom material (including big boulders) can be forced landwards, and building constructions demolished. In regions covered by coarse moraine deposits, barricades of boulders along the shores have been built up by ice-pressing (Fig. 2.6). These barricades were probably mainly formed in the early sub-Atlantic period (about 2,000 years before present), when the ice-pressing is considered to have been stronger than at present. In a very large number of Scandinavian lakes, water levels nowadays reach far below the barricades, indicating that the lakes have been lowered. However, in intact lakes, the water level is still found at the base of the barricade (Fig. 2.3).
cm 0
50
Water surface
Erosive zone
100 Zone without organic sediment 150
200
Sand Gyttja composed of fine detritus Diatom ochre Clay
Litorella - Lobelia
Bottom of organic sediment
Litorella - Lobelia Lobelia - Litorella
Fig. 2.5 Schematic distribution of vegetation in different littoral zones in a Lobelia lake (Lake Fiolen, Sweden). The upper portion of the minerogenic zone is under the erosive influence of ice movements. The minerogenic zone is lakewards followed by the organogenic bottom (From Thunmark 1931)
34
S. Björk
Fig. 2.6 Lake Fiolen, Sweden (summer, above; winter, below). Littoral barricade of boulders pushed together by the ice (photo: Einar Naumann 1917)
Besides ice-pressing, another type of ice movement can occur, which is also important in the physical sculpting of the shore. This is ice-pushing, brought about by ice-floes in connection with the break-up of the ice. When a sudden break-up of the ice is caused by strong winds, the ice-floes are pushed against the shore and piled up along the wind-exposed side of the lake. In this way, the shore is also cleaned of vegetation and trees are de-barked on the lakeward side. Whenever possible, in the restoration and management of lakes and wetlands, the conditions for the formation of qualitatively good ice in order to make use of ice movements, as well as of water level fluctuations, should be re-created. The utilisation of ice movements is a cheap way to protect lakes from the overgrowth and heavy detritus production of macrophytes, and to preserve open shores suitable for wading birds, etc.
2 The Evolution of Lakes and Wetlands
35
References Björk S et al (1972) Ecosystem studies in connection with the restoration of lakes. Verh Internat Verein Limnol 18:379–387 Digerfeldt G (1972) The post-glacial development of Lake Trummen. Regional vegetation history, water level changes and palaeolimnology. Folia limnologica scandinavica 16, 104 pp Digerfeldt G, Håkansson H (1993) The Holocene paleolimnology of Lake Sämbosjön, Southwestern Sweden. J Paleolimnol 8:189–210 Hutchinson GE (1957) A treatise on limnology, vol I. Geography, physics and chemistry. Wiley, New York, 1015 pp Thunmark S (1931) Der See Fiolen und seine Vegetation. (Lake Fiolen and its vegetation). Acta phytogeogr suec 2:198 (in German)
Chapter 3
Development of Aquatic Macrophytes in Shallow Lakes and Ponds Jan Pokorný and Sven Björk
Abstract The distribution of macrophytes in shallow lakes and ponds is subject to zonation. A typical zonation of the littoral of Scandinavian lakes under oligotrophic conditions include, from the lake surface to the lowest part of the littoral, helophytes, nymphaeids, elodeids and isoetids. With increasing pollution and eutrophication of lakes leading to increased turbidity and reduced transparency, first to disappear are the isoetids followed by the elodeids. At the same time, i.e. with increasing nutrient levels, the quantity of emergent and floating-leaved plants increases. Further increases in nutrient levels result in system instability, excessive development of chlorococcal algae, blue-greens or filamentous algae, extreme values of oxygen concentration and high pH, oxygen depletion at the bottom, and, in extreme situations, even to the release of free ammonia – the conditions which may lead to fish kills as experienced in fish ponds of Bohemia. Keywords Littoral zonation • Eutrophication • Plant biomass • Diversity • Fish stock • pH changes • Oxygen concentration • Algae • Blue-greens
3.1 Zonation of Macrophytes in Oligotrophic Lakes A typical distribution of macrophytes in the littoral zone of cold, temperate water bodies is shown in Fig. 3.1. Such zonation occurs providing that the littoral topography, bottom conditions, exposure to wind and waves, light penetration in the water, grazing pressure, etc., allow the development of a complete constellation of life forms of aquatic macrophytes. Based on the occurrence of different life forms, the littoral zone can be divided into a number of sub-zones. The terminology given
J. Pokorný (*) ENKI, p.b.c., Dukelská 145, 379 01, Třeboň, Czech Republic e-mail:
[email protected] S. Björk Department of Ecology, Limnology, University of Lund, SE-223 62, Lund, Sweden M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_3, © Springer Science+Business Media B.V. 2010
37
38
J. Pokorný and S. Björk
Supralittoral
Spring high water level
Eulittoral
Summer low water level
Upper
Helophytes
Lower
Nymphaeids
Upper
Elodeids
Lower
Isoetids
Sublittoral
Elittoral
Profundal
Fig. 3.1 Schematic illustration of the littoral zonation of macrophytic vegetation in an oligotrophic lake
in Fig. 3.1 is used to characterise the littoral sections of lakes in Scandinavia. The groups of macrophyte life forms include: 1. The hyperhydates (helophytes) – emergent plants such as the graminids, Phragmites and Typha, and the herbids, Alisma and Cicuta 2. The ephydates – floating-leaved plants represented by spirodelids (Spirodela, Lemna) and nymphaeids such as Nymphaea and Potamogeton with floating leaves 3. The hyphydates – the group of submerged plants including riccids (Riccia), elodeids (i.e. plants with long shoots like Elodea, Myriophyllum and Potamogeton without floating leaves), isoetids (submerged plants with short shoots, exemplified by Isoëtes and Littorella), and muscids (submerged mosses) The uppermost zone of the littoral, the eulittoral (Fig. 3.1), is delimited by the highest level of the spring high water and the lowest level of the summer low water. The remaining division into sub-zones, as illustrated in Fig. 3.1, is based on the vertical distribution of hyperhydates, nymphaeids, elodeids and isoetids. Thus the lakeward limit of the upper sublittoral coincides with the extension of hyperhydates (helophytes) and that of the lower sublittoral with the deepest occurrence of nymphaeids. In the same way, the upper elittoral ends with the deepest distribution of elodeids and the lower elittoral with that of isoetids and mosses. Below these littoral zones inhabited by macrophytes comes the zone called the profundal. Depending on environmental conditions, the aforementioned types of plant life forms can be present or missing in lake ecosystems of different character and the
3 Development of Aquatic Macrophytes in Shallow Lakes and Ponds
39
deepest level to which macrophytes grow can vary tremendously. In a typical, South-Scandinavian, clear-water lake, macrophytes occur down to ca. 5 m. In some Eifel maar lakes in Germany, Nitella flexilis grows at a depth of more than 20 m (Melzer 1992) and in a Bavarian lake, meadows of Chara are found at a depth of 15 m (Melzer 1976). The most diverse vegetation is found in the upper littoral sub-zones, with a mixture of life forms. As the water depth increases, the hyperhydates, the nymphaeids, the elodeids and isoetids disappear one after the other, as the limits of their ecological demands become exceeded. A complete spectrum of macrophyte life forms as shown in Fig. 3.1, is typically found in oligotrophic lakes with clear water (Lobelia lakes), allowing the development of dense carpets of isoetids in the deep-water lower elittoral. In brown-water (humic) lakes the levels of the sub-zones are displaced upwards. With pollution and increased nutrient concentration, leading to turbidity caused by plankton as well as to overgrowth by periphyton, first the isoetids and then the elodeids disappear, i.e. the lakes lose their lower and upper elittoral zones as structural ingredients and functional sections of their ecosystems. At the same time, the quantity of emergent and floating-leaved plants typically increases. Further increase in nutrient concentrations results in ecosystems characterised by instability leading to sudden changes and ecosystem collapse. The development to this stage, and the interrelations among ecological factors in such systems, are exemplified by the heavily fertilised fish ponds of Bohemia.
3.2 Structure and Function of Fish Ponds Under Heavy Human Impact With an increase in nutrient input, stands of aquatic plants (macrophytes, chlorococcal and blue-green algae) become denser, their biomass per unit area increases, the vertical distribution of this biomass changes, and a shortage of carbon dioxide (CO2) in the water occurs as pH increases during the diurnal cycle of plant photosynthetic activity. Competition for light and CO2 are assumed to be one of the crucial processes determining the development and succession in submersed vegetation habitats in shallow ponds with increased eutrophication. Figure 3.2 shows vertical profiles of biomass distribution, light extinction, pH and dissolved oxygen concentration under different trophic conditions or human impact (fish stock). This scheme is based on works of De Nie (1987), Pokorný et al. (1990), Pokorný and Ondok (1991), and does not include emergent wetland plants like sedges, common reed, etc. Below, the ecological conditions in fish ponds at different trophic stages are described as schematically illustrated in Fig. 3.2a–g: A. Oligotrophic stage In oligotrophic water bodies, the growth of macrophytes is limited by lack of nutrients. Likewise, the lack of nutrients dissolved in water limits the growth of algae. Transparency of water is high and only the plants which are able to take
40
J. Pokorný and S. Björk
Fig. 3.2 Development of aquatic vegetation 1 – from oligotrophic stage (a) over mildly eutrophic (b), eutrophic (c, d) to hypertrophic (e, f, g). 2 – biomass distribution in vertical profile, 3 – extinction profile, E0 = PhAR irradiance at water surface, Ez = PhAR irradiance at the depth z, 4, 5 – pH and oxygen concentrations in vertical profile (- - -) night minimum values, (_______) day maximum values
nutrients through their roots from the bottom sediment can grow. Plant biomass is accumulated near or on the lake bottom. Values of pH and oxygen concentration do not change during day and night. Similarly, there are no differences in oxygen concentrations and pH values in the vertical profile, the values of oxygen concentrations being about air saturation level. Such situations may be found in “Lobelia lakes” (Littorela sp., Lobelia sp. and Isoëtes sp.). These conditions were also common in some Bohemian fish ponds – before the fertilising of these fish ponds and the increased transport of nutrients from the catchment began.
3 Development of Aquatic Macrophytes in Shallow Lakes and Ponds
41
B. Long-lasting oligotrophic to mesotrophic stage In more fertile water bodies, the aquatic vegetation consists of a relatively high number of species. Water transparency is 2 m or more and the plant biomass is evenly distributed from water surface to the bottom. The tips of submersed macrophyte shoots only rarely reach the water surface, some of them developing floating leaves. Species diversity of the periphyton on macrophytes and of the benthos near their roots is high. Nutrient concentrations in the water remain low during the whole season and the development of planktonic algae is therefore limited. Shading and carbon dioxide uptake by the periphyton do not seriously limit the growth of submersed macrophytes as the periphyton is intensely grazed by molluscs, insect larvae and fish. The concentration of oxygen in the water is about air saturation level and no marked differences are observed between the day and night or surface and bottom oxygen concentrations. The photosynthetic uptake of carbon dioxide does not have any strong effect on water pH – neither diurnal changes nor differences in the vertical profile are seen in pH. At this stage, the aquatic primary production is essentially limited by a shortage of mineral nutrients. Nowadays, fish ponds of this kind are rare, not only because fish pond management is more intensive, but also because of high levels of nutrient input from the catchment areas. C. Initial stage of progressive eutrophication Higher nutrient input in the water bodies by, for example, manuring and direct fertiliser application (as in the case of fish pond management), and agricultural runoff, brings about a more vigorous growth of macrophytes and their stands become denser. Plants grow fast and their biomass accumulates at the water surface. The young, green parts of macrophytes shade the deeper water. Water transparency decreases and even in water bodies shallower than 1 m, irradiance at the bottom may not attain the light compensation point for photosynthesis (less than 3 W m−2). While photosynthesis prevails at the water surface, respiration prevails at the bottom – steep gradients of oxygen concentrations and pH values develop during daylight hours when no mixing occurs due to temperature differences between water surface and bottom. As nutrient loading continues, species diversity of aquatic plants decreases but the stand biomass rises. This stage is also characterised by a mass development of periphyton which suppresses the growth of some macrophytes. The respiration rate of the whole community is higher than at previous stages and therefore a lack of oxygen often occurs near the bottom. The rapid decomposition of organic matter at the bottom results in low oxygen concentrations and the release of nutrients from the bottom sediment which became anaerobic. Internal loading of the water body starts to play an important role. D, E, F. Eutrophic to hypertrophic stage At high nutrient loading, the fish stock plays an important role in the development of water vegetation. At lower fish stock (seasonal mean live biomass of fish under ca. 350 kg ha−1), large Daphnia (filter-feeders of phytoplankton) are not completely
42
J. Pokorný and S. Björk
consumed by fish. The high feeding pressure by Daphnia (cladocerans) prevents the growth of algae and water transparency is kept high, despite the high nutrient level. A high transparency and high nutrient levels cannot last long and results in: (a) The growth of macrophytes (D) in shallower waters. At high nutrient levels, macrophytes are very often invaded by filamentous algae. Filamentous algae, namely Cladophora sp., use inorganic carbon effectively even at low concentrations, grow fast and develop dense mats throughout the water column. When the filamentous algae reach the water surface (E), they completely shade the water column, pH increases to values above 11 (Eiseltová and Pokorný 1994) and those macrophytes which served as the substrate for filamentous algal growth die off. (b) the development of Aphanizomenon blooms in deeper waters, as large colonies of Aphanizomenon cells are not consumed by filter-feeders so effectively and therefore remain, whereas smaller chlorococcal algae are consumed. At higher fish stock levels (seasonal mean live biomass of fish above ca. 400 kg ha−1) large Daphnia numbers are reduced and chlorococcal algae develop dense populations of several hundreds micrograms chlorophyll a l−1. These dense populations of chlorococcal algae (F) bring about low light transparency (Secchi disc less than 0.4 m) which prevents the development of blue-green algae like Aphanizomenon and Microcystis later in the season when higher temperatures are suitable for their growth. However, later in the summer, other species of blue-greens may occur in the phytoplankton (Limnothrix, Planktothrix). Fish managers use the number of large Daphnia as an indicator of fish stock feeding pressure and are able to prevent water blooms of blue-greens at high nutrient levels by adjusting the fish stock. This stimulates the development of dense populations of small chlorococcal algae (Faina R personal communication, 1994). Such a manipulation of the food web is possible in fish ponds where the fish stock is under control. However, also small fish like Pseudorasbora parva and Carassius auratus that are difficult to control feed on large Daphnia and effectively reduce the number of large zooplankton. When such fish are present a lower biomass of carp does not necessarily result in a recovery of large zooplankton species. G. Development of Oscillatoria species Spring development of filamentous blue-green algae, such as Oscillatoria, at the sediment surface results in a high pH in the whole water column and the subsequent release of free ammonia (NH3) responsible for fish-kills (more details see Pokorný and Květ 2004).
3.3 Conclusion To summarise, the development and succession of submersed vegetation indicates the level of trophic conditions in shallow lakes and ponds, as illustrated in Fig. 3.2. If eutrophy increases, the biomass of submersed vegetation (submersed macrophytes,
3 Development of Aquatic Macrophytes in Shallow Lakes and Ponds
43
chlorococcal and blue-green algae) increases and the increased photosynthesis leads to extreme values of oxygen concentration and high pH, and subsequent undesirable changes in water chemistry. This leads to unfavourable conditions for many aquatic organisms and degradation of ecosystem structure and function. Considerable reduction of nutrient loads, both external and internal, is essential in order to re-create properly functioning ecosystems which are sustainable in the long term.
References De Nie HW (1987) The decrease of aquatic vegetation in Europe and its consequences for fish populations. EIFAC/CECP Occasional paper No. 19, 52 pp Eiseltová M, Pokorný J (1994) Filamentous algae in fish ponds of the Trebon Biosphere Reserve – ecophysiological study. Vegetatio 113:155–170 Melzer A (1976) Makrophytische Wasserpflanzen als Indikatoren des Gewässerzustandes oberbayerischer Seen. Diss Bot 34:1–195. J. Cramer, Vaduz Melzer A (1992) Submersed macrophytes. In: Scharf, BW and Björk S (eds) Limnology of Eifel maar lakes. Ergebnisse der Limnologie 38:223–237 Pokorný J, Ondok P (1991) Macrophyte photosynthesis and aquatic environment. Rozpravy Academia, 4:117, Praha. Pokorný J, Kvet J, Ondok P (1990) Functioning of the plant component in densely stocked fishpond. Bull Ecol 21:44–48 Pokorný J, Květ J (2004) Aquatic plants and lake ecosystems. In: O′Sullivan PE and Reynolds CS (eds) The Lakes Handbook Vol. 1 Limnology and limnetic ecology. Blackwell Science and Blackwell Publ. Malden, Oxford, Carlton, pp 309–340
Chapter 4
Food Web and Trophic Interaction and Development Jaroslav Hrbáček
Abstract The size-efficiency hypothesis predicts a decrease of phytoplankton biomass with increased size of zooplankton. To verify the extent of phytoplankton changes with small changes of zooplankton due to changes in the fish stock, three approaches to the estimation of size of zooplankton were developed: the average length of zooplankton, the percentage of cladocerans >0.71 mm in the total cladoceran biomass, and the length of the average biomass. The percentage of cladocerans >0.71 mm shows a statistically-significant correlation of their seasonal averages to the seasonal averages of chlorophyll concentration and, furthermore, a steeper long-term increase in chlorophyll than that of total phosphorus at a decrease of percentage of cladocerans >0.71 mm. The newly-developed method of estimating the length of average biomass from size classes of biomasses is more objective than that of the percentage of cladocerans >0.71 mm and includes both Cladocera and Copepoda. The values of the old and new method show highly-significant correlations. Both methods showed a long-term decrease of the size of zooplankton in two reservoirs over decades not related to the changes in biomass. The reason for this decrease and that of the lower efficiency of transport of energy in reservoirs as compared to fish ponds are discussed as well as weaknesses of the so-called biological indication. Keywords Biological indication • Energy transfer • Chlorophyll concentration • Top-down effect • Zooplankton size
J. Hrbáček (*) Biology Centre of the AS CR, v.v.i., Institute of Hydrobiology, Na Sádkách 7, 370 05 České Budějovice, Czech Republic e-mail:
[email protected] and Hekrova 820, 149 00, Praha 4, Czech Republic
M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_4, © Springer Science+Business Media B.V. 2010
45
46
J. Hrbáček
4.1 Development of Limnology and Fish Ecology The development of limnology and fish ecology had already substantially diverged by the end of the nineteenth century – and even today the two disciplines continue to largely ignore vital aspects of the approach of each other. The limnological approach is best documented by the following sentence: ‘While various reciprocal relations exist between the plankton and the non-plankton animals, there is reason to believe that if all non-plankton animals were removed from a lake and kept out, the plankton, with possibly some minor modifications, would continue to exist’ (Welch 1952, p. 277). The earlier paper by Lindemann (1942), which is often considered as the start of the ecosystem approach in limnology, demonstrates the dynamic relationships within an ecosystem using data from a bog lake containing no fish. However, already in 1939, Ivlev had shown a very high efficiency of energy transfer from primary production of phytoplankton to fish production. Unfortunately this paper was disregarded, but from the high efficiency of the energy transfer in the food chain, an intensive pressure of predation, a ‘top-down’ effect in todays’ terms, could have been predicted. In the ecology of fishes, investigations usually tacitly assume that the feeding activity of fish influences only the biomass (or quantity) of their food organisms. What is usually not considered is that changes in fish feeding activity due to changes in fish stock could change the competitive interactions, and thus the species composition and therefore also the size spectrum of their food. The different approaches and divergent paths in the development of limnology and fish ecology also influenced newly-evolving practical disciplines such as the management of water bodies.
4.2 Historical Background to Management of Lakes and Reservoirs The term lake (or reservoir) management usually does not include fish management (Henderson-Sellers 1979; Mitsch and Jørgensen 1989). This is despite the fact that the well-being of fish, especially predatory ones, is the best continuous indicator of an adequate management of water quality for human use. On the other hand, a lot of experience exists with fish management in man-made water bodies. In countries such as the Czech Republic, the area of man-made stagnant water bodies is much larger than that of natural ones. For this reason Czech limnology has not been so concerned with the classical objects of limnology, i.e. lakes, but with man-made water bodies, i.e. fish ponds and reservoirs. Many fish ponds in the Czech Republic were established in the Middle Ages, some having surface areas larger than a hundred hectares. With the construction of ponds new water bodies with their biota were formed, or in today’s terminology, new ecosystems established and managed. Since the middle of the nineteenth
4 Food Web and Trophic Interaction and Development
47
century the management of these artificial fish ponds has increasingly intensified and, as a result, their productivity of fish has exponentially increased (Hrbáček 1969a). Equally, the first reservoirs, for supplying water to mills, were built in the Middle Ages, some of which were also managed for fish production. Larger reservoirs for power generation and those for water supply are of more recent origin; the main features of their limnology are summarised by, for example, Hrbáček (1984). The above is to indicate that the management of freshwater ecosystems, mainly fish, has a long tradition and that due to the necessity to manage man-made water bodies, a more comprehensive understanding of the interactions of all components of the biota became necessary. In the last quarter of the twentieth century, efforts to decrease the undesirable development of algae and blue-green algae in lakes and reservoirs have followed two approaches: to decrease the concentration of the limiting nutrients, and to increase the activity of filtrators and browsers. For the latter approach the term biomanipulation or more correctly food web management is used. The success of food web management depends both on the extent of practical experience and theoretical understanding of the interactions within the ecosystem of stagnant waters. In my opinion neither is fully satisfactory at present and therefore it is impossible to present a concise set of instructions to safely achieve the goals of food web management. Below are presented the results of investigations and theoretical considerations of a group of Czech limnologists on these topics.
4.3 Nutrients–Phytoplankton Interactions The fundamental question which has to be addressed from the point of view of theoretical understanding as well as practical management is: “Which compound or element is limiting for the development of phytoplankton?” The term development includes both qualitative (species composition) and quantitative (total biomass of algae in space and time) aspects. Preferably, the frequency and biomass of individual algal species should be measured but this is extremely time-consuming. Therefore, the chlorophyll a concentration is often estimated and used as a substitute for the determination of the biomass or biovolume of phytoplankton; the advantage being that chlorophyll concentration is simple to determine and directly indicates the active component of phytoplankton. Moreover, recent developments in instrumentation enable the determination in situ of chlorophyll concentrations of individual groups of phytoplankton. The disadvantage is that it does not strictly correlate with biomass because the percentage of chlorophyll a in the biomass of algae varies (within certain limits) in different species and under different light conditions. All the available methods for its determination only consider certain static aspects. The dynamic aspect, namely how much organic matter is produced by the algae, is more difficult to assess. The course of events during the year fluctuates notoriously both within and between years. A cold May in one year can be followed by a warm one in the next.
48
J. Hrbáček
Seasonal averages are usually robust quantitative parameters due to the effects of autocorrelation (we have hot days only in summer) and compensation (a hot May can be followed by a cold June), more so than their coefficients of variation would suggest. Estimation of the seasonal average of chlorophyll a concentration has been used to establish the relation between certain parameters of algae and cyanobacteria (blue-green algae) on the one hand and phosphorus on the other. Dillon and Rigler (1974) correlated the concentration of total phosphorus during the spring circulation and average concentration of chlorophyll a within the growing season (Fig. 4.1). Thus, from the highly-significant correlation of data from lakes which differed in total phosphorus by several orders of magnitude, it follows that total phosphorus is the main factor influencing the seasonal average value of chlorophyll in these lakes. If one looks in greater detail, however, the picture is less satisfactory, i.e. a value of total phosphorus of 100 mg l−1 can correspond to values of chlorophyll ranging from 10 to 100 mg l−1. There are several reasons why the phosphorus-chlorophyll correlation is not closer than actually observed: 1000
Summer chlorophyII a (mg m−3)
100
10
1.0
0.1
100 10 Total phosphorus (mg m−3)
1000
Fig. 4.1 Relation of the seasonal average of chlorophyll a concentration to total phosphorus concentration in spring (From Dillon and Rigler 1974)
4 Food Web and Trophic Interaction and Development
49
1. Hydrological, e.g. during the growing season there occurs a considerable inflow of water with a concentration of total phosphorus sometimes very different from that of the spring circulation. This is especially important in reservoirs with average retention times lower than half a year – thus during the growing season a considerable water volume is replaced. In such cases, the correlation between the seasonal averages of both total phosphorus and chlorophyll a are usually used. One would expect that, in this case, the variation should be smaller because the value of total phosphorus includes the phosphorus compounds present in the phytoplankton and therefore these two variables are not entirely independent. Investigations, however, have shown that this expectation is not fulfilled (Hoyer and Jones 1983; Hrbáček et al. 1993) and that there is about the same residual variability as exists in the classical relationship. 2. Physiological, e.g. some other nutrients limit, at least temporarily, the development of algae. Limitation by nitrogen was found in those cases when the N/P ratio was below 15, and limitation by CO2 in cases where the pH was very high. 3. Methodological, e.g. the seasonal average value of chlorophyll is not comparable to that of the biomass of higher plants being related to the amount of nutrients present in the soil. Actual concentration of chlorophyll represents a dynamic equilibrium between the increase and decrease of algae. Increases are due to the growth and multiplication of algal cells which, in turn, is related to the actual concentration of nutrients. Decreases in the algae are due to consumption by filtrators or browsers, sedimentation and decay. Consumption of algae by filtrators is connected with the partial regeneration of nutrients due to catabolic processes in animals and in the euphotic zone these nutrients can be used by algae or cyanobacteria. 4. Statistical, e.g. sampling during the growing season is too infrequent such that extraordinary high values sampled may not be compensated by unsampled low values that are present, and vice versa. 5. Inherent variability, e.g. several other factors not mentioned above influence the phosphorus–chlorophyll correlation. This is analogous to the variability in other relations as, for example, the length-weight relation of fish.
4.4 Relation of Zooplankton and Benthos to Phytoplankton and Fish Stock In the early days of limnology, it was tacitly assumed that the development of algae is due only to the concentration of nutrients. The development of filtratory zooplankton and benthos animals is dependent on the development of algae. The development of fish, such as roach, bream and carp, was assumed to be influenced by the development of its invertebrate prey. Nowadays the shorthand term for this aspect of relations is ‘bottom-up’ control (resource limitation). However, it is evident that the zooplankton can influence, by their feeding activity, the development of algae, and equally, fish can influence the biomass and species composition of
50
J. Hrbáček
zooplankton and benthos. This relational aspect is now known as ‘top-down’ control (control by predators). From the published data it is evident that the selective effect of fish predation on zooplankton is much more pronounced than that on benthos. The reasons are intuitively assumed to be due to much greater protection of benthic organisms to fish predation, especially to that of most numerous small fish. Zoobenthos biomass per unit area, especially in shallow water bodies, frequently surpasses that of zooplankton during the growing season. On the other hand, its average turnover rate is lower than that of zooplankton. In fish ponds with an average depth of about one metre, it has been estimated that the zooplankton and zoobenthos in fish ponds contribute almost equally to the carp production (Lellák 1957; Kořínek et al. 1987). It may be reasonably assumed that the relative contribution of zoobenthos to fish production decreases as average depth increases due to both a decreasing temperature and a decreased amount of organic particles reaching the bottom. To assess the quantitative dynamic aspects of the aquatic biota including growth rates is a very complicated and time consuming task. The role of all important food web links will be presented below in the section on the transfer of energy in the food web. Studies of the relation of seasonal averages of zooplankton biomass to seasonal averages of chlorophyll concentration in various water bodies have indicated considerable residual variation. In backwaters and ponds, a relatively close relationship of the seasonal averages of zooplankton biomass to the concentration of nutrients has been found. Even extreme differences in the fish stock (Hrbáček 1969b) did not affect the relation of composite samples of zooplankton, taken at night from different areas and different strata, according to their proportion in the total volume of the water body, to the concentration of nutrients. The common sense view of a decrease of zooplankton biomass at higher fish stocks, as might be anticipated, could not be substantiated. The extremes in fish stock influenced the seasonal average decrease of oxygen concentration overnight. High fish stock produced a higher decrease i.e. induced a more intensive metabolism of the biota in open water at the same concentration of nutrients and biomass of zooplankton. In ponds without fish, or with only a few hundred fish per hectare, large Daphnia species such as Daphnia pulicaria prevailed in the zooplankton biomass. In water bodies with abundant fish (10,000–100,000 fish per hectare of all year-classes – a frequent size of fish stock in oxbows and reservoirs in central Europe) small Daphnia species, such as D. cucullata and dwarf strains of D. galeata, prevailed and especially Bosmina longirostris and Ceriodaphnia pulchella. In fish ponds with a very low fish stock (about 300 specimens of carp in their third year per hectare) with an average concentration of total phosphorus of 0.3 mg l−1, the spring values of chlorophyll decreased to a few mg l−1. Later on in the season, cyanobacteria developed water blooms the biomass of which was highly variable from year to year and from pond to pond. In fish ponds with a fish stock two or three times as high, higher concentrations of chlorophyll were found throughout the year (Fott et al. 1980). It could be therefore anticipated that in reservoirs with a total phosphorus concentration lower by one order of magnitude, a low fish stock may induce a lower biomass of phytoplankton than is the actual one at a high fish stock.
4 Food Web and Trophic Interaction and Development
51
4.5 Quantification of the Effect of Fish Predation on Zooplankton The high fish stock in reservoirs is manifested not only by the presence of small species in zooplankton but also by the slow growth rate of fish, which can be identified by the small distances between fish scale annuli. The literature on the interaction of forage fish-cladocerans-filterable phytoplankton, from Brooks and Dodson (1965) – through the later DeMelo et al. (1992) and Carpenter and Kitchell (1992) – to later authors, is based on comparisons of various habitats that have a considerably different fish stock. The main interest was to understand the mechanism of the depression of phytoplankton with possible practical applications. Shapiro et al. (1975) have introduced for this endeavour the term biomanipulation. The difficulty in the study of such changes of zooplankton-phytoplankton interaction in water bodies with intermediate fish stock was the estimation of the fish stock and quantification of its predation effect on zooplankton. The exact estimation of the fish stock is still the subject of intensive studies as is apparent from the sentence: ‘Obtaining a true picture of a fish stock is a hard challenge but it is a good goal for scientific development’ (Kubečka et al. 2009). Even with excellent data on fish stock it is very problematic how to evaluate the predation effect of individual size classes of individual fish species. Mills and Schiafone (1982), as well as Mills et al. (1987), have shown a statistically-significant correlation between the size of fish preying on zooplankton and size of zooplankton (Fig. 4.2). This type of approach to studies of food webs was infrequent in the last two decennia but has been recently strongly recommended by Brose et al. (2006). The method that was applied, namely the estimation of zooplankton mean length on the basis of measurements of the lengths of individuals from a representative sample of both zooplankton and fish and estimation of the biomass from the measured length and ascertained length – biomass relation of individual species, is very labour intensive – especially in the case of the zooplankton. To elaborate a method that would produce similar results but be less laborious, and therefore suitable for long-term measurements and hence monitoring, it was assumed that the percentage biomass of a group of larger individuals in the total biomass may be a reasonable parameter for this purpose. For the estimation of biomass the application of a biuret reagent (Blažka 1966) for protein determination was used. Contrary to a measurement of biomass by the estimation of dry weight, the estimation of proteins is not influenced by different larger organic non-living seston, because the proteins in debris are being intensively degraded. The method is simple, and determines the physiologically-most-important part of the biomass as does chlorophyll for phytoplankton biomass determination. The values of biuret protein N can be converted to values in biomass by a conversion factor of 8.16. Straškraba (1964) developed the method for the separation of Cladocera and Copepoda by their different adhesion to the surface layer in a separation flask. Instead of the average value of the length, the percentage of large cladocerans in the total cladoceran biomass has been determined. A sieve with an opening of 0.71 mm was selected as the criterion for separating the biomass of larger and smaller cladocerans. In this way it is possible to
52
J. Hrbáček 300
All zooplankton
Cl
Gr Cr
Mean total length of fish in gill nets (mm)
250
Mu
200
Re
PI
Mo
Bu
Po
Y = 160+83X
Hy
BI
r2 = 0.54
300
CI
Cladocera Gr
Cr
250 Mo Mu
200 Bi
Re
PI
Bu
Y = 168 +68X r2 = 0.60
Po
0.5
Hy
1.0
1.5
2.0
Mean body length of zooplankton [mm]
Fig. 4.2 Relation between the mean total length (in mm) of fish collected in gill nets and the mean body length of all crustaceoplankton and Cladocera (From Mills and Schiafone 1982)
obtain in about 24 h, with little in the way of equipment and laborious effort, the biomass both of cladocerans and copepods and the percentage of large cladocerans. The percentage of large cladocerans is independent of the method used for biomass determination and is found in individual samples, in addition to the predation pressure by fish being also influenced by rapid changes in the nutritional status of the zooplankton. In the case of there being a strong increase of food supply, the newborn of cladocerans became numerous and decrease the ratio in addition to the predation by fish, whereas decreasing the feeding conditions acts in the opposite way. During the growing season the extremes of feeding conditions of zooplankton are usually restricted to one per year and so this effect, within the seasonal average, is negligible. Kubečka (1989) and Seďa (1989) have shown a good correlation of the seasonal averages of percentage of large cladocerans to the biomass of the fish stock in longterm observations. In Slapy reservoir, for example, the percentage of large cladocerans in the years 1985–2001 (Fig. 4.3) decreased (Hrbáček et al. 2003) to less
4 Food Web and Trophic Interaction and Development
53
Fig. 4.3 Percentage of the biomass of cladocerans >0.71 mm in total cladoceran biomass in Slapy reservoir in 1985–2001 (full line = regression line; dotted line = distance weighted least square fit line) (From Hrbáček et al. 2003)
Fig. 4.4 Biomass of Copepoda (upper half and scale on the right) and Cladocera (lower half and scale on the left) in Slapy reservoir in the years 1964–2001 (full line = regression line; dotted line = distance weighted least square fit line) (From Hrbáček et al. 2003)
than one third (0.29). The biomass of both cladocerans and copepods in the same reservoir (Fig. 4.4) slowly but significantly increased during the years 1964–2001. The interesting point of this comparison is that the biomass of copepods is nearly the same as that of cladocerans.
54
J. Hrbáček
4.6 Effect of Different Fish Stock on the Seasonal Average of Chlorophyll Concentration in Reservoirs The seasonal average of chlorophyll concentration and the ratio of the biomass of Cladocera >0.71 in total cladoceran biomass were determined during several years in three reservoirs. In total, 26 seasonal observations were made (Fig. 4.5). The range of the percentage of large cladocerans varied from 1.5% to 45.3%, while the seasonal average of chlorophyll concentrations varied from 3.8 to 29.3 mg l−1, i.e. greater than one order of magnitude in the case of percentages and under one order of difference in chlorophyll. The range of the seasonal average of total phosphorus was from 12 to 38 mg l−1, i.e. a maximum value approximately threefold that of the minimum. From this relationship between the seasonal averages of chlorophyll concentration and percentage of large cladocerans, it follows that at percentages of large cladocerans above 40% (three seasonal observations), the seasonal average of chlorophyll concentrations was well below 10 mg l−1, whereas in all other cases it was above 10 mg l−1. When the seasonal average of chlorophyll concentration was linearly adjusted to the same concentration of total phosphorus, the general trend of an increase in chlorophyll concentration with decrease in percentage of large cladocerans nearly disappeared. However, the exceptionally low concentration of chlorophyll in the 50
Cladocera [percentage of biomass]
76 40
Vrchlice
75 78
79 80
30
Hubenov 82 81
83
77 83
79 80
20 87 10
80
89
82 83 10
79
82 88 81
86
85 0
Rímov
84 20
90
81 30
40
Chlorophyll a (µg I−1)
Fig. 4.5 The relationship between the seasonal average of the ratio of large Daphnia biomass to the total cladoceran biomass and the seasonal average of chlorophyll a concentration in three reservoirs. The numbers within symbols represent the last two digits of the respective years
4 Food Web and Trophic Interaction and Development
55
3 years with a percentage of large cladocerans above 40% remained (Hrbáček et al. 1993). These years also showed an exceptionally high biomass of zooplankton in relation to chlorophyll concentration (Hrbáček et al. 1994). In Slapy reservoir, a relative increase of chlorophyll of 1.55 times within the period 1975–2001 was found by Hrbáček et al. (2003), while the relative increase in total phosphorus was only 0.31. As mentioned earlier, in the period 1985–2001 the percentage of large cladocerans decreased nearly to a third. The above discussion on the interaction of phytoplankton-zooplankton-foraging fish concerned only seasonal averages of concentrations of total phosphorus, chlorophyll, biomass of zooplankton per unit area and the ratio of large cladoceran biomass. Hrbáček et al. (2003) have shown that in Slapy reservoir the biomass of copepods is about the same as that of Cladocera (Fig. 4.4) and this seems to be a typical figure for this type of reservoir, while in some pools the biomass of copepods can surpass the biomass of cladocerans (Hrbáček et al. 1994). Cyclopoid and calanoid copepods play an important role in exploiting the biomass of some larger algae not accessible to filtratory zooplankton (e.g. diatoms Asterionella and Aulacoseira). The organic matter of dead algae and cyanobacteria can be exploited by bacteria, which within the so-called microbial loop (Stockner and Porter 1988) are exploited by the larger protests, which in turn can be again used by both cladocerans and copepods. The above paragraphs have shown the advantages of using the percentage of cladocerans >0.71 as a measure of intensity of fish impact. On the other hand, this parameter is far from ideal. The use of a mesh size >0.71 mm is arbitrary. The procedure totally fails in samples which have no specimens >0.71 mm. It does not present any information on the influence of the predatory fish impact on copepods. A new approach has therefore been recently developed (Hrbáček 2008). The sample is sieved in sequence through 0.93, 0.71, 0.42, 0.20 mm sieves. The samples remaining on individual sieves are divided to cladoceran and copepod fractions by the difference in the adhesion to the surface and the protein N in individual sub samples determined colorimetrically using the biuret reaction. The total protein N retained on individual sieves is then obtained by adding the subsequent values of individual sub samples from the largest to smallest sieve. On the basis of this total, the percentage of individual samples retained on each sieve of individual size is obtained. A calculation of the size of the opening of the sieve through which half of the biomass should theoretically pass and half of which would be retained on the sieve is based on an interpolation of the data from the size of neighbouring size classes with a lower and higher percentage than 50% and the percentage values retained on these sieves. The procedure is shown diagramatically in Fig. 4.6. This value in mm can also be interpreted as the length of the average biomass and is highly-significantly correlated to the percentage value of Cladocera >0.71 mm especially at larger values. Figure 4.7 shows the seasonal variability of this new parameter in three water bodies with different levels of fish stock. Figure 4.8 presents the statistically significant decrease of the length of both Cladocera and Copepoda in Římov reservoir. The distance weighted least square fit line for Cladocera and Copepoda (Fig. 4.8) is only partly parallel in time periods
56
J. Hrbáček Samples from pool with winter kill, no fish in summer
% of cladocera passing selected mesh
100 80 60
50 40
me cla sh p do as ce sin ran g 5 bio 0% ma of ss
20 0 0,1
0,3
0,7
0,5
0,9
1,1
mesh size in mm
Fig. 4.6 Graph visualizing the procedure of the estimation of the length of average biomass = size of the mesh passing 50% of the biomass
mm Length of Aver.Biomass of Cladocerans 0.9 Pool T3 0.8 0.7
Hubenov
0.6 0.5 0.4
ec D
ov N
ct O
Se p
Au g
Period 2
Ju l
n Ju
M
ay
Period 1
Ap r
M
b Fe
n Ja
ar
Rímov
0.3
Fig. 4.7 Seasonal pattern of the length of average biomass of cladocerans in a pool without fish, in Hubenov reservoir with trout and perch stock, and in Římov reservoir with stock of stunted cyprinids and percids (period 1 is characterised by spawning of fish and large changes in the biomass of cladoceran populations, period 2 by the growth of fry
4 Food Web and Trophic Interaction and Development
length of average biomass [mm]
0.8
Cladocera: r = –0,228, p = 0,00008
0.7
57 distance weighted least sq.fit yearly averages
0.6 0.5 0.4
0.55
0.3
Copepoda: r = –0,605, p = 0,0028;
0.50 0.45 0.40 0.35
1985
1990
1995
2000
2005
0.30
Fig. 4.8 The length of average biomass of Cladocera and Copepoda in Římov reservoir in the years 1985–2006
above and below the regression line. From the graph it is also evident that the length of Cladocera is in most cases distinctly larger than that of Copepoda. It should be added that through the 0.2 mm-large openings of the plankton net used for collecting the samples, which also represents the lowest size class, most of the biomass of nauplii of copepods pass through, whereas the cladocerans are quantitatively retained. If it should be that the biomass of these nauplii would also be included in the total copepodan biomass, then the length of the average biomass of Copepoda would be even smaller than that presented in Fig. 4.8. In both reservoirs the length of the average biomass decreases – indicating an increased relative fish impact on zooplankton. The possibility of being able to predict the length of the average biomass to which this decreases points is difficult, as we do not have comparable data on zooplankton from water bodies with very low growth rates of non-predatory species as is the case, for example, in the Jordan reservoir (Kubečka and Bohm 1991). The reason for this decrease in the increased impact of fish on zooplankton is probably the decreasing ratio of biomasses of predatory to non-predatory fish. Mills and Schiafone (1982) considered the zooplankton size as an indicator of the predatory to ‘pan fish’ ratio (pan fish being a size of game fish smaller than the largest). The relationship of chlorophyll of phytoplankton on the one side and size of fish stock on the other to this new parameter in individual reservoirs have to be examined as has been the case in investigation of the relation of the ratio of Cladocera >0.71 mm to these parameters. For comparison with the Slapy reservoir (Fig. 4.4) the biomass of the zooplankton of Římov reservoir is presented in Fig. 4.9. In both reservoirs the biomass of Cladocera increases, but whereas the biomass of Copepods increases in Slapy it decreases in Římov reservoir. In Římov reservoir the total crustacean zooplankton biomass slightly decreases but not statistically. The distance weighted least square
58
J. Hrbáček
3 2
Distance weighted least sq. fit
Cladocera: r = 0,1639; p = 0,0134; y = 0,529+4,4376E-5*x Cladocera+Copepoda r = −0,0841; p = 0,21; y = 1,437 − 3,655E-5*x
1 4
0
Copepoda: r = −0,2595; p = 0,00008; 3 y = 0,9084-8,0924E-5*x
2 1 0 1990
1995
2000
2005
Fig. 4.9 The biomass of Cladocera and Copepoda in Římov reservoir in the years 1989–2006
fit line above and below the regression line of cladoceran and copepodan biomass is again only partly parallel. In this type of reservoirs, with steep and mostly rocky shorelines, the biomass of zooplankton is the main source of food and so is an indicator of possible increase or decrease of fish stock biomass. For comparing the parameters of different water bodies the averages over a certain period of years are used. Table 4.1 compares the averages of several parameters of zooplankton of Slapy and Římov reservoir for the period 1989–2006. The reservoirs are different in length, 42 km for Slapy against 12 km for Římov reservoir, in maximum depth, 55 m against 42 m, and the average inflow would fill Slapy reservoir in 1 month against 3 months for Římov. The water inflowing in the Římov reservoir is incorporated either to the surface or subsurface layer according to its temperature. In Slapy reservoir the inflowing water is from the outflow (from depth) of another even larger reservoir. Therefore its temperature over the warmer part of the year is considerably lower than that of the surface water of Slapy reservoir and therefore passes the reservoir at very deep layers with only a very slight mixing with the water of the surface and subsurface water. The average length of Cladocera in Slapy reservoir compared to Římov reservoir is larger by 3.65% and that of Copepoda by 3.03%. The biomass averages of Slapy and Římov reservoirs differ by 26.7% in Cladocera and 8.67% in copepods, though after logarithmic transformation the difference nearly disappears (0.11% and 1.03%). This indicates a very high similarity of basic zooplankton parameters from two reservoirs with very different morphologic and hydraulic parameters and, furthermore, an unparallel development of cladoceran and copepodan biomass in both reservoirs. The reason for the differences between the relationships of averages of logarithmically-transformed and non-transformed biomass data of both reservoirs is the extreme size distribution
4 Food Web and Trophic Interaction and Development
59
Table 4.1 Comparison of averages of Slapy and Římov Reservoir from the years 1989–2006. Clad = Cladocera, Cop =Copepoda, Cru = sum of the biomass of the Cladocera and Copepoda; Cl/Co = ratio of the cladoceran and copepodan biomass. LeBiCl and LeBiCo = length of average biomass of Cladocera and Copepoda Biomass (mg m−2) LnBiomass Ln Clad Cop Cru Clad Cop Cru Cl/Co Cl/Co LeBiCl LeBiCo Slapy 0.87 Římov 0.64 Difference 0.23 Difference 26.7 (%)
0.72 0.65 0.06 8.67
1.59 1.29 0.30 18.6
3.73 3.72 0.01 0.11
3.86 3.82 0.04 1.03
140
4.63 4.64 −0.01 −0.21
0.47 0.45 0.02 3.65
0.36 0.35 0.01 3.03
90 80
Slapy
70 No of obs
100 No of obs
4.47 4.51 −0.04 −0.81
100
120
80 60
60 50 40 30
40
20
20 0
1.40 1.76 −0.36 −26.1
10 −1
0
1 2 3 4 5 Cladocera biomass in mg* m2
6
7
0
−3 −2 −1 0 1 2 3 4 5 6 Cladocera logarithm of biomass
7
8
Fig. 4.10 Distribution of the biomass of Cladocera in Slapy reservoir in the years 1989–2006 in histograms of linear and logarithmic transformed data
of the biomass (Fig. 4.10), which after logarithmic transformation becomes closer to a normal distribution. This would indicate that the averages usually used for comparison purposes are sometimes not robust enough to counteract the effect of extreme distributions.
4.7 Transfer of Energy from Primary Production to Fish Production The approach presented up until now has been pragmatic, i.e. referring to observations and measurements which can be accomplished relatively easily, even if they do not fully agree with the requirements of methods used in productivity studies. A more rigorous approach is to measure, in addition to the static parameters such as concentration or biomass, the dynamic parameters such as uptake and regeneration of nutrients, growth rate, reproduction, mortality and production, and include them in a simulation model. There are, at least, three kinds of complications or drawbacks, two objective ones and one subjective one, which make this approach less productive for understanding the interactions within the aquatic biota than it would first appear:
60
J. Hrbáček
1. One objective complication is that some factors (e.g. the effect of nutrient concentration on the growth rate of algae, or the effect of food particles on filterfeeding zooplankton) can only be related to unit volume. This increases the number of estimations needed due to the reality that the activity of organisms is influenced by temperature and temperature is vertically, and within restricted limits also horizontally, stratified and some organisms undergo vertical, and partly also horizontal, migration. Furthermore, organisms are not distributed randomly, even within the same layer, but in more or less pronounced patches – especially in benthic organisms. Planktonic organisms can have a horizontal gradient with regard to being between open water and the shore, and also along the length of ‘fiord-like’ lakes or reservoirs. The horizontal gradients are sometimes visually demonstrated by the water blooms of cyanobacteria. To get a truly realistic picture, one would need to take so many measurements that their realisation would be outside the possibilities of even a large research team. It is therefore necessary not only to clump species with similar roles in the food web together, but also to include some assumptions the reality of which cannot be easily verified. 2. As a consequence of the necessary simplifications (mentioned above), a description of the energy flow does not have a predictive power able to evaluate what kind of changes will be induced by qualitative and/or quantitative changes of individual parts on this flow. In this respect, a more powerful idea is the concept of keystone species: by assessing how changes in their biomass or size will influence parameters, including the species composition of adjacent trophic levels. In the communities of standing waters, the fish stock has the role of key stone species and, in some parts of the year, also large copepod species. 3. The subjective difficulty is linked to the fact that the quantitative ecology of ecosystems is neither scientifically nor economically a very profitable endeavour for the researcher, as the chances of some fundamentally new facts or phenomena being found are very low. It is possible, of course, at least partially, to fulfil the aims of the dynamic approach. Primary production of the planktonic algae can be measured in fertile water bodies as an oxygen increase in light glass bottles suspended at several depths. Exposure for 24 h at different depths could, at the outset, be considered as ideal, but has the drawback that the value thus obtained is smaller than the sum of several shorter exposures; the flasks contain not only algae, but also bacteria and animals, their respiration decreasing the concentration of oxygen. The respiration of the whole community in the flasks can be assessed by measuring the oxygen decrease in dark bottles suspended close to the light ones; adding the value of the decrease in the dark bottle to the increase estimated in the light bottle gives a value for the so-called gross primary production. Due to the changes in frequency and distribution of algae, as well as variable illumination, these measurements have to be repeated every week, or at least every other week, to obtain a sum for the whole growing season. In fish ponds, the yield of the dominant fish species (carp) after draining the pond is closely related to the total production of fish, better expressed as net fish production. It is therefore necessary to convert one of these parameters so that we
4 Food Web and Trophic Interaction and Development
61
compare like with like – either both net or gross productions. For converting the gross to net primary production obtained by the above-described method it is necessary to subtract the respiration of algae from the gross primary production value. For this we do not have a direct measurement at our disposal but have to rely on an educated guess – not an easy task as the algae are present in different horizontal layers, thus living in different light and temperature conditions. In practice, a certain part, usually one third of the summary gross primary production, is subtracted. To convert the net fish production to gross production it is necessary to add the part of the production used by metabolism (respiration) – what would appear to be an easily measurable task. The difficulty is that, in addition to the effect of changing temperature during the growing season, we also have to take into account the effect of activity; so again we have to rely, within certain limits, on an educated guess. Thus there are limitations in any attempt to get a general picture of the flow of energy within a community. The outcome of this comparison is that, in fish ponds, the efficiency of transfer of energy from primary to fish production is about 5% (Kořínek et al. 1987). Assuming transfer by the food chain algae-filtrators-fish, this finding indicates an efficiency of about 20% from one level to the other. This is unexpectedly high as it is on the limit of the assumed physiological possibilities of the organisms involved. Furthermore, bacterial production and that of invertebrate predators also depend on the primary production. To estimate fish production in lakes and reservoirs it is necessary to estimate the numbers and biomass of individual year-classes of individual species present. The method applied is to catch, estimate the numbers and size of individual yearclasses (partly, later on, from samples of scales) for individual species, mark by fin-clipping and release and. later on, to catch again. From the proportion of the marked fish in a later catch, the probable number of fish in individual year-classes is estimated. From the procedure’s description it is apparent that only fish above a certain size, i.e. at least 2 years old, can survive the operation. Smaller fish can be estimated by other less elaborate procedures, e.g. by estimating their proportion to larger fish in catches with very dense seines. When this procedure is repeated over several years, it is possible to estimate the decrease in numbers of individual yearclasses from year to year, i.e. the mortality and recruitment. Thus having obtained these data, an estimate of total fish production is possible; however, this method is time consuming and hence the data so far collected are scarce. In reservoirs, the estimated efficiency of primary production to fish production is only about half that of fish ponds (Hrbáček 1984). This lower efficiency can be due to at least three causes: 1. The ambiguous feeding character of some year-classes of fish, such as perch, which can, in their third year, feed either on invertebrates or fish, thus adding an additional level in the food chain and lowering the efficiency of energy transfer. 2. The low growth rate of forage fish in reservoirs. 3. The more ‘diluted’ conditions by about tenfold of total phosphorus concentration in reservoirs as opposed to fishponds. The comparatively diluted phosphorus
62
J. Hrbáček
concentrations can mean a similarly ‘diluted’ biomass of food which induces a greater energy expenditure for fish searching for and accumulating food. Most likely in the future, with more data available, it will be possible to predict fish production from the concentration of total phosphorus – just as it is now possible to predict the value of chlorophyll concentration from total phosphorus. The chart (Fig. 4.11) of estimated energy flow through all important parts of the food web of the large shallow Rybinsk reservoir (about 300 km north of Moscow, total area 4,550 km2, average depth 5.6 m, retention time 274 days) shows an efficiency of transfer of primary production to fish production of approximately 4%, i.e. close to that of fish ponds. On the other hand, according to this chart, planktonic filtrators are dependent more on the production of bacteria than on the production of planktonic algae. This prolongation of the food link should again manifest itself in a lower efficiency; however, this is counterbalanced by the very high input of allochthonous organic matter used up by the production of bacteria. The chart also shows that in this shallow water body the estimated production of comparable levels is at least some five times higher in open water than on the bottom.
4.8 The Growth Rate and Yield of Fish Stock in Ponds The growth rate of fish is dependent on the pattern of annual temperature and the interaction between the fertility of the water body and its fish stock (the number of fish of individual year-classes per hectare). Carp (Cyprinus carpio) has been the preferred fish in pond fisheries in central Europe for centuries, though it is difficult to decide whether this was due to its taste, to the practices used in pond management (especially in the procedures used during fish harvesting and transport to market) or to its resistance to overcrowding due to its provenance from southern countries. In order to reach a market size of about 1.5 kg, in the second half of the nineteenth century carp was grown for about 6 years, whereas now it only takes 3 years (exceptionally two in warm areas and four in cold). In the first instance, the length of the period from egg to maturity is regulated by temperature (degree days), and then to marketable size, also by the density of the fish stock. Usually only a single year-class of carp is grown in each fish pond. For simplicity, a linear inverse relationship between the size of the fish stock and individual growth is normally assumed. The desired optimal number of fish to be reared in a pond is calculated from the pond’s long-term average fish harvest (which is related to pond fertility) and the desired increase in size during the growing season (Schaperclaus 1961). All these considerations concerning fishponds assume a very low use of additional artificial feeding. The above-mentioned linear relationship is, of course, applicable only within a restricted range – obviously one fish per hectare cannot fully utilise all the resources of the pond and grow to a size of several hundreds of kilograms representing the usual harvest. It has been found that at the stock which is close to the
4 Food Web and Trophic Interaction and Development Light 6 3.91 . 10 kJ
Allochthonous org. matter P = 5.275
Planktonic bacteria P = 1465
80
2nd trophic level
Planktonic protozoans P = 142 523
Detritus
3rd trophic level
138
50
209
Rotatorians + Calanoidea + Cladocera P = 184
63
117 4
17
Planktonic predators P = 26.8
8
13
4
Planktonophagous fish P = 46
25
4th trophic level
38
Predatory fish P = 13.4
0.54 1.55 Man
1.00 0.5
5th and 6th trophic level
production by the individual parts of the food web
0.8
Benthic predators P = 1.7
13 17
21 Fish fry P = 29
Benthic protozoans P = 21
8 29
84
63
13
Macrobenthos P = 33
46
4
Benthic bacteria P = 251
126
419
419
113
1004
Phytoplankton P = 2093
1549
4
Macrophytes P = 113
6
1.05 . 10
4270
251
5.4 . 10
1.7
138
Detritus
2.6. 106
1st trophic level
63
Benthophagous fish P = 11.3 4 21
10.5 Predatory fish P = 2.1
348 Benthic deposits Outflow
transfers within the food web
part of the production not used in the food web
Fig. 4.11 Schematic diagram of the quantitative trophic relationships within the Rybinsk reservoir ecosystem. Data from Sorokin in Kuzin and Shtegman (1972) in kcal m−2 year−1 transformed to kJ m2 year−1 using the factor 4,187 (Hrbáček 1984)
maximum growth rate of fish (mmax in bacterial or algal cultures), the maximum yield of fish per unit area is not reached. It should be mentioned that in fish ponds, except in cases of mortality due to disease or some negative environmental conditions such as low oxygen concentration or toxic substances, the yield represents
64
J. Hrbáček
the net production of fish as natural mortality is negligible. The maximum total yield or production of a fishpond is reached at two to three times a higher density of fish stock than the one at which maximum growth rate is reached. At still higher densities of fish stock the total yield decreases, of course, due to the fish needing to increase the intensity of searching for less accessible food. Usually it is assumed that this is due to the increased scarcity of food but in my opinion an assumption of lower accessibility is more adequate: this would include not only frequency but also size (for a large fish smaller food items are of lower importance at the same total biomass) and protection of the prey (e.g. the larvae of chironomids protected by a layer of mud). In fish-pond farming, natural fish recruitment is undesirable; from the above it is obvious why – it decreases not only the expected growth of individual fish but, in cases of intense recruitment, also the total yield.
4.9 Production and Yield of Fish in Reservoirs In lakes and reservoirs, the yield of fish does not depend only on fish production but also on the fishing effort (Bayley 1988). The fishing effort does not embrace evenly all year-classes of all fish species. Contrary to the situation in fish ponds, in these water bodies a part of the fish stock must remain over the winter to reproduce the next spring. Yield in these water bodies is much lower than fish production. The important finding in these estimations of fish production is that the first 2 year-classes, which are usually not caught by anglers, constitute a very important part of fish production (Fig. 4.11; Mills and Forney 1983; Hrbáček 1984). The proportion of the older year-classes, which constitute the main interest for anglers, is very low. It is obvious that it decreases as the growth rate diminishes, as a consequence of the high stocking level of fish. Thus the most effective way to improve the stock of game fish is not to increase their numbers by increasing recruitment – or even artificial rearing and planting of fry – but to increase the growth rate. As in fish ponds this can be achieved only by reducing the stock. Also, similarly as in fish ponds, the size of the impact of fish on zooplankton can be estimated by the size of the species forming the bulk of the zooplankton, i.e. by the earlier mentioned percentage of biomass of Daphnia retained on the 0.72 sieve to total cladoceran biomass and more recently by the length of the average biomass. In this way, it is also possible to monitor the effectiveness of management measures on the fish stock. From Fig. 4.5 and the Section 4.4 on the relation of zooplankton and benthos to phytoplankton and fish stock, it follows that increasing the proportion of large Daphnia in the total cladoceran biomass, or more exactly to increase the length of average biomass of Cladocera and Copepoda, will improve the water quality by reducing the development of algae. The other possibility to check the effectiveness of fish management is to monitor the growth rate of fish, especially that of game fish, by their scales (or other parts) from which it is possible to identify the age of fish or by assessment of average sizes of individual species.
4 Food Web and Trophic Interaction and Development
65
To reduce the numbers or biomass of stock of forage fish there are several possibilities: 1. To harvest both the small fish by old well-established methods (different kinds of nets and traps and especially by trawling in open water) 2. Intensive planting of predatory fish preferably of older year-classes 3. To decrease the recruitment of forage and coarse fish To my knowledge, up till now in Europe, no attempt to reduce the fish stock in lakes or reservoirs has been successful enough to reduce the forage fish to the extent that large Daphnia species (preferably D. pulicaria) become a considerable component of the zooplankton biomass during the whole year. At present, D. pulicaria is scarce in lakes and reservoirs in Europe – except for high mountain lakes lacking in fish. In North America, on the other hand, there are lowland lakes in which D. pulicaria and other large Daphnia species are leading component of the zooplankton. It may be speculated that this difference between Europe and North America is due to the different main stock of fish: Cyprinidae in Europe and Centrarchidae in North America. The lack of well-established comparisons shows, in my opinion, how little the interactions within the biota of lakes and reservoirs are studied – and therefore how little understood. Another phenomenon not fully understood is the spectacular increase in stock of predatory fish soon after reservoirs are put into operation (Hrbáček 1984) and the subsequent collapse within a few years. A water level rise in reservoirs during the growing season after a period of lowered levels for several weeks, during which some plant cover on dry shores has developed, simulates in many aspects the situation during spring floods in the inundation areas of rivers.
4.10 Historical Background to the Formation of the Biota of Standing Waters The phenomena and relationships described, and the conclusions derived therefrom, are based on biota as they exist at the present time. This, to a certain extent, is comparable to attempting to derive the meandering of a reservoir’s present shoreline from only the resistance of its different parts to erosion. For a better understanding of the function of the biota in reservoirs it might be useful to consider how the species composition of these biota came about. It was mentioned that these biota are relatively simple in the sense that the bulk of the biomass of individual levels of the food web is composed of only a few species. In zooplankton this may be perhaps expected – due to the relatively uniform physical environment. However, this explanation is probably not true, as in the ocean, in a single vertical haul of plankton net, the number of species can be 10- to 100-fold higher than in reservoirs or lakes, so probably the development of biota over geological time may have played a significant role.
66
J. Hrbáček
From a geological point of view, not only fish ponds and reservoirs but also most European and American lakes are of recent origin. Zooplankton, contrary to the benthic fauna, does not have endemics, not even in the geologically-oldest lakes. Most of the species present in the plankton of reservoirs and lakes are also present in pools and oxbow lakes. Bythotrephes longimanus is reported in the literature only from these lakes; A. Nauwerck (1965) has found this latter species also in pools in northern Europe, but I have not seen records of this crustacean from reservoirs. It is claimed that the low fish yield in reservoirs is due to the fact that the species of fish living there are derived from the stock of fish living in the riverine environment (Fernando and Holčík 1982). It is assumed that this fish stock is unable to fully exploit the plankton biomass in the central part of the water body. However, the range of the ratio of the size of cladocera to the size of fish shown in Fig. 4.2 does not support this assumption. The part of the statement concerning the origin of species can be perhaps enlarged so that by riverine ecosystem we mean not only habitats in the river itself but also in the pools, backwaters and oxbows connected with the river during a flood. In a pool near the river Lužnice, the coefficient of variation of total phosphorus of 72 samples is reported by Pechar et al. (1996) as 88.7. This indicates the restricted possibility of speciation to different levels of nutrient. The course of seasonal interaction between cladocerans and cyclopean copepods on the one side and fish on the other can be understood more properly if we take into account that this relationship developed in pools (Hrbáček et al.1994) under environmental pressure. Before man interacted with the biota of pools, cut-offs (oxbows) along the river and the river itself, their biota, especially the fish, were under the strong impact of vertebrate predators, both fish and warmblooded vertebrates (mammalian as well as avian). Man not only considerably decreased this effect by hunting vertebrates, but also changed the relation of predatory to non-predatory fish by a higher impact on predatory fishes. The biota in reservoirs do not have any impact from warm-blooded vertebrate predation, while angling exerts pressure preferably on predatory fish entirely neglecting small fish. In this way, forage fish have good conditions for an increase in numbers.
4.11 Species Diversity and Trophic Levels In recent years a major emphasis has been put on species diversity. One of the outcomes of the efforts in this direction has been the use of molecular methods to distinguish populations of different origin, and perhaps different requirements, which could not be distinguished on the basis of morphological comparisons. For example, populations of Daphnia pulicaria in Czech ponds and in some lakes in the Tatra Mountains, although morphologically extremely similar, have been found to be genetically so different that they can be considered as different species (Marková et al. 2007). Further, the role of hybrids of cladoceran species has been found to play a more important role, by promoting more abundant populations than might be anticipated; this indicates some advantages against maternal populations.
4 Food Web and Trophic Interaction and Development
67
The consequences of these recent findings on improving our knowledge of interactions in communities have yet to be seen. The relation of prevailing zooplankton species to biological indication of trophic state or predation pressure was found more vague as anticipated at the beginning (Hrbáček et al. 1961). Longmuir et al. (2007) did not find in 31 lakes in British Columbia significant correlation between the diversity of zooplankton, phytoplankton and bacteria. In addition, the physical factors that were associated with species composition in one trophic level were independent of those that were important for another. The authors speculate that this might be due to the broad diverse diets found in both the consumers and decomposers. Barnett and Beisner (2007) studied the biodiversity of crustacean zooplankton communities along a gradient of total phosphorus (TP), as well as the vertical heterogeneity and relative abundance of their phytoplankton resources, in 18 lakes in Quebec, Canada. Log of zooplankton species richness showed the previously-established tendency (not statistically significant) to a unimodal relationship with log TP. Log of functional diversity showed a significant negative correlation to TP; it also showed a significant curvilinear correlation to the log of vertical heterogeneity of cyanobacteria. Most of the other numerous regressions studied showed tendencies but only a few of them were statistically significant. In both these, as in most other similar, studies the possible effect of the different consumption of zooplankton by different fish, thus increasing or decreasing species diversity, has not been considered. It would appear that limnologists prefer to study objects under the microscope, and objects like fish or water plants are more frequently outside their direct interest. It is my opinion, therefore, that such limnologists do not contribute fundamentally new inputs to the quantification of the interactions important for promoting the knowledge necessary for a rational management of lakes and reservoirs.
4.12 Summary The requirements in reservoirs for good water quality management and that for good fishing are not necessarily contradictory. Both aspects can be improved by decreasing the total phosphorus concentration and by reducing the stock of foraging fish that usually develops in reservoirs after they come into operation. A reduction of the fish’s impact on zooplankton is manifested by an increase in the average length of cladocerans and copepods. The classic method to show this requires measuring the length of thousands individuals and is very time consuming. The proposed new procedure enables long-term monitoring: it is based on the determination of the distribution of biomasses in size classes separated by sieving and estimates the length of the average biomass using low-cost equipment and is much less labour intensive. A decrease in size of both Cladocera and Copepoda in two reservoirs over a period of nearly two decades indicates an increase in the predatory impact of fish on zooplankton; this is very likely due to an increase in the numbers
68
J. Hrbáček
of pan fish (smaller-sized game fish). A decrease in the growth rate of fish, by examining the scales of fish, could check these results. But this again is a laborious procedure; as an alternative, estimation of the average size of fish by measuring the mass and counting the numbers of several representative catches of individual fish species per growing season during a long term program could be also used for this purpose. An increase in both the size of zooplankton and individual fish species will induce a decrease in the biomass of algae. The reason why the biota of reservoirs are not self-regulatory, producing a growth rate of forage fish not far from the maximum, under given temperature conditions, is not fully understood. I speculate that the structure of the interaction between invertebrates, forage fish and their predators, that was developed in the conditions found in pools and cut-off meanders along rivers before man started to interfere with these relations, differ considerably from those presently found in reservoirs, particularly in terms of the development of fish and warm-blooded predators.
References Agraval A, Ackerly DD, Adler F, Arnold AE, Cáceres C, Doak DF, Post E, Hudson PJ, Maron J, Mooney KA, Power M, Schemske D, Stachowicz J, Straus S, Turner MG, Werner E (2007) Filling key gaps in population and community ecology. Front Ecol Environ 5(3):145–152 Barnett A, Beisner BE (2007) Zooplankton biodiversity and lake trophic state: explanations involving resource abundance and distribution. Ecology 88(7):1675–1686 Bayley PB (1988) Accounting for effort when comparing tropical fisheries in lakes, river floodplains, and lagoons. Limnol Oceanogr 33(4 part 2):963–972 Blažka P (1966) Bestimmung der Proteine in material aus Binnengewässern. Limnologica 4:387–396 Brooks JL, Dodson SJ (1965) Predation, body size and composition of plankton. Science 150:28–35 Brose U, Jonsson T, Berlow EL, Warren P, Banasen-Richter C, Besier LF, Blanchart TL, Brey T, Carpenter SR, Blandenier MFC, Cushing L, Dawah HA, Dell T. Edwards F, Harper-Smith S, Jacob U, Ledger ME, Martibey ND, Memmott J, Mintenbeck K, Pinnegar JK, Rall BC, Rayner TS, Reuman DC, Ruess L, Ulrich W, Williams RJ, Woodward G, Cohen JE (2006) Consumer resource body size relationship in natural food webs. Ecology 87(10):2411–2417 Carpenter SR, Kitchell JF (1992) Trophic cascade and biomanipulation: interface of research and management. Limnol Oceanogr 37(1):208–213 Colbourne JK, Crease TJ, Weider LJ, Hebert PDN, Dufresne F, Hobaek A (1998) Phylogenetics and evolution of a circumarctic species complex Cladocera: Daphnia pulex). Biol J Linnean Soc 65:347–365 DeMelo R, France R, McQueen DJ (1992) Biomanipulation: hit or myth? Limnol Oceanogr 37(1):192–207 Dillon PJ, Rigler FH (1974) The phosphorus–chlorophyll relationship in lakes. Limnol Oceanogr 19(5):767–773 Fernando CH, Holčík J (1982) The nature of fish communities: a factor influencing the fishery potential and yields of tropical lakes and reservoirs. Hydrobiologia 97:127–140 Fott J, Pechar L, Pražáková M (1980). Fish as a factor controlling water quality in ponds. In: Barica J, Mur LR (eds) Hypertrophic ecosystems. Dev Hydrobiol 2:255–261 Henderson-Sellers B (1979) Reservoirs. Macmillan, London and Basingstoke, 128 pp
4 Food Web and Trophic Interaction and Development
69
Hoyer MV, Jones JR (1983) Factors affecting the relation between phosphorus and chlorophyll a in Midwestern reservoirs. Can J Fish Aquat Sci 40:192–199 Hrbáček J (1969a) Relations between some environmental parameters and the fish yield as a basis for a predictive model. Verh Int Verein Limnol 17:1069–1081 Hrbáček J (1969b) Relation of productivity phenomena to the water quality criteria in ponds and reservoirs. Proceedings of the 4th International Conference held in Prague 1969, pp 717–724 Hrbáček J (1984) Ecosystems of European man-made lakes. In: Taub FB (ed) Lakes and reservoirs. Elsevier Science, Amsterdam, pp 267–290 Hrbáček J (1987) Systematics and biogeography of Daphnia species in the Northern temperate region. Peters RH, de Bernardi R (eds) ‘Daphnia.’ Mem Ist Ital Idrobiol 45:37–76 Hrbáček J (2008) A new procedure to estimate the size adaptation of zooplankton on fish predation. 48th Annual report. Biology Centre AS CR, v.v.i., Institute of Hydrobiology, České Budějovice Hrbáček J, Dvořáková M, Kořínek V, Procházková L (1961) Demonstration of the effect of the fish stock on the species composition of zooplankton and the intensity of metabolism of the whole plankton association. Verh Int Verein Limnol 14:192–195 Hrbáček J, Albertová O, Desortová B, Gottwaldová V, Komárková J, Kopáček J, Popovský J, Seďa J, Vyhnálek V (1993) Overshooting phenomenon of chlorophyll a concentrations and the year to year variation of the fish impact on zooplankton. Arch hydrobiol Beih Ergebn Limnol 40:175–184 Hrbáček J, Pechar L, Dufková V (1994) Anaerobic conditions in winter shape the seasonal succession of Copepoda and Cladocera in pools in forested inundations. Verh Internat Verein Limnol 25:1335–1336 Hrbáček J, Brandl Z, Straškraba M (2003) Do the long-term changes in zooplankton biomass indicate changes in fish stock? Hydrobiologia 504:203–213 Ivlev VS (1939) La production des poissons des étangs par rapport a ľ intensité. (The production of fish in ponds in relation to intensity). Bjul mosk obsh ispyt prir otd biol 48:29–38 (in French) Kořínek V, Fott J, Fuksa J, Lellák J, Pražáková M (1987) Carp ponds of central Europe. In: Michael RG (ed) Managed aquatic ecosystems. Elsevier Science, Amsterdam, pp 29–62 Kubečka J (1989) Popisný model průběhu početnosti, biomasy, racionu a produkce rybí obsádky v účelově obhospodařované vodárenské údolní nádrži. (Descriptive model of the course of frequency, biomass, ration and production of the fish stock and its use in a rationally managed reservoir). Manuscript of the thesis. Czechosl Acad Sci České Budějovice. 121pp. (in Czech) Kubečka J, Bohm M (1991) The fish fauna of the Jordan reservoir, one of the oldest man-made lakes in central Europe. J Fish Biol 38:935–950 Kubečka J, Amarasinghe US, Bonar SA, Hateley JA, Hickley P, Hohausová E, Matěna J, Peterka J, Suuronen P, Tereschenko V, Welcomme R, Winfield IJ (2009) The true picture of a lake or reservoir fish stock: a review of needs and progress. Fish Res 96(1):1–5 Lellák J (1957) Der Einfluss der Fresstätigkeit des Fishbestandes auf die Bodenfauna der Teiche. (The impact of feeding activity of fish on the aquatic fauna of the ponds). Zschr Fischrei 6 N.F. 8:621–633 (in German) Lindemann RL (1942) The trophic dynamic aspect of ecology. Ecology 23:399–418 Longmuir A, Shurin JB, Clasen J (2007) Independent gradients of producer, consumer, and microbial diversity in lake plankton. Ecology 88(7):1663–1674 Marková S, Dufresne F, Rees DJ, Černý M, Kotlík P (2007) Cryptic intercontinental colonization in water fleas Daphnia pulicaria inferred from phylogenetic analysis of mitochondrial DNA variation. Mol Phylogenet Evol 44:42–52 Mills EL, Forney JL (1983) Impact on Daphnia pulex of predation by young yellow perch in Oneida lake, New York. Trans Am Fish Soc 112:154–161 Mills EL, Schiafone A Jr (1982) Evaluation of fish communities through assessment of zooplankton populations and measures of lake productivity. N Am J Fish Manage 2:14–27
70
J. Hrbáček
Mills EL, Green D. Schiavone A. Jr (1987) Use of zooplankton size to assess the community structure of fish populations in freshwaters lakes. North American Journal of Fisheries Management 7:369–378 Mitsch WJ, Jørgensen SE (1989) Ecological engineering: an introduction to ecotechnology. Wiley, New York/Chichester/Brisbane/Toronto/Singapore, 472 pp Pechar L, Hrbáček J, Pithard D, Dvořák J (1996) Ecology of pools in the floodplain. In: Prach K, Jeník J, Large ARG (eds) SPB. Academic, Amsterdam, pp 209–226 Schäperclaus W (1961) Lehrbuch der Teichwirtschaft. Paul Parey, Berlin and Hamburk 582 pp Seďa J (1989) Populační dynamika perlooček Daphnia a Bosmina ve vodárenské nádrži se silným predačním tlakem ryb. (The dynamics of the populations of the cladoceran genera Daphnia and Bosmina in a reservoir with a high fish predation pressure). Manuscript of the Thesis. Czechosl Acad Sci. České Budějovice Seďa J, Dostálková I (1996) Live sieving of freshwater zooplankton: a technique for community size structure. J Plankton Res 18:513–520, 133 pp (in Czech) Seďa J, Kubecka J (1997) Long-term biomanipulation of Rimov reservoir (Czech Republic). Hydrobiologia 345:95–108 Shapiro J, Lamarra V, Lynch M (1975) Biomanipulation. An ecosystem approach to lake restoration, Contribution 143. Limnological Research Centre, University of Minnesota, Minneapolis, MN, 32 pp Sorokin (1972) In Kuzin BS, and Shtegman BK. Rybinskoe vodokhranilishe i yevo zhizn (The Rybinsk Reservoir and its life). Nauka, Leningrad, 360 pp (in Russian) Sterner WR, Elser JJ (2002) Ecological stoichiometry. Princeton University Press, Princeton/ Oxford, 439 pp Stockner JG, Porter KG (1988) Microbial food webs in freshwater planktonic ecosystems. In: Carpenter SR (ed) Complex interactions in lake communities. Springer, New York, pp 69–83 Straškraba M (1964) Preliminary results of a new method for the quantitative sorting of freshwater plankton into main groups. Limnol Oceanogr 33:587–594 Welch PC (1952) Limnology, 2nd edn. McGraw-Hill, New York/Toronto/London, 538 pp
Chapter 5
Principles, Planning and Accomplishment of Lake Restoration Projects Sven Björk
Abstract This chapter describes a procedure for accomplishing lake restoration projects, starting from the necessary pre-investigations of a lake’s character and condition, through the design of a restoration plan, and a monitoring programme designed so as to be able to evaluate whether the project goal has been achieved. Keywords Lake restoration • Limno-technical plan • Restoration objectives
5.1 Project Goals and Limnological Investigations An essential prerequisite when designing a restoration project is to set up restoration objectives for the given site. Arguments for carrying the project through to completion should be collected and the future status and use of the lake/wetland clearly determined. Limnological investigations, over at least a 1 year period, must provide the basis for the technical design and execution of a restoration project for most types of degraded systems. The analytical part of these studies includes the calculation of a hydrological budget and the recording of the present ecological status, involving investigations of: environmental conditions; communities of organisms; and ecological interrelationships. Amongst other things, data on the chemistry must be collected in order to estimate the input (external loading), availability in the lake (including internal loading) and output of nutrients. As regards the sediments, their horizontal distribution and thickness have to be mapped. The stratigraphy of the sediments should be examined in relation to soil type, and their physical and chemical conditions. It is important to elucidate the integrated role of the upper layer of the sediment in the degraded lake ecosystem concerned, as well as the conditions for the
S. Björk (*) Department of Ecology, Limnology, University of Lund, SE-223 62 Lund, Sweden e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_5, © Springer Science+Business Media B.V. 2010
71
72
S. Björk
release and precipitation of essential elements. These studies should be carried out in conjunction with the palaeolimnological investigations. If sediment has to be removed from a lake, it is advisable that the limnologist examines how the sediment behaves on drying and freezing, and asks questions about its possible utilisation for different purposes (soil conditioning, use in gardens, agriculture, etc.). The standard limnological pre-restoration programme should include investigations of plankton, macrophytic vegetation, bottom animals, fish and other vertebrate species, ornithological value of the site, etc., with special regard to rare and threatened species. Data on ecosystem productivity should also be obtained to make both quantitative and qualitative comparisons of the pre-and post-restoration system. Old photographs and maps, as well as records in archives concerning fishery and hunting, are often helpful to elucidate former conditions in the ecosystem. A limnological synthesis of the analytical data should explain the present ecological conditions, including the relationships between the catchment area and lake eco system, functional aspects and relationships within the system (trophic relations, production and degradation of organic matter, etc.) as well as the future development of the water body in the case that no restoration measures were to be taken.
5.2 The Limno-technical Plan Limnological planning of the restoration should describe in detail the measures necessary to achieve the agreed goals. As every ecosystem has its individual characteristics, the restoration plans have to be tailor-made for each specific project, and are most appropriately prepared by the limnologist in cooperation with the firm responsible for the technical side of the project. Each plan should include a detailed description of the technical procedure, costs involved and the monitoring programme needed to assess its effectiveness. Prognoses for the development of the lake after restoration should be included in the plan, and a follow-up programme to check environmental conditions in the restored lake should be designed. Both positive and negative results should be reported. Based on experience collected during the planning and execution of a number of lake restoration projects, the procedure of planning, designing, accomplishment and follow-up of a project can be schematically summarised as follows (specific considerations for acidified waters and ecosystems contaminated by poisonous substances are not included): I. Selection of ecosystem for restoration – what are the priorities 1. Degraded lakes close to towns and cities, drinking-water reservoirs, fish ponds 2. Degraded lakes/wetlands which used to be valuable waterfowl biotopes 3. Restoration of wetlands for improvement of self-purification: re-creation of nutrient traps and denitrification reactors 4. Integrated landscape water management (see Chapter 1) and in accordance with the EU Framework Directive 2000/60/EC in the field of water policy
5 Principles, Planning and Accomplishment of Lake Restoration Projects
73
II. Project goals 1. General improvement and re-creation of local, pleasant environmental conditions; aesthetic reasons 2. Re-creation of waters for water supply, swimming, canoeing, wind-surfing, etc. 3. Re-creation of waters and wetlands for waterfowl and aquatic biocoenosis in general 4. Education; before-during-and-after studies; access to aquatic biotopes for schools; inspirational demonstration projects; ecosystem structure and function - dependent on environmental changes 5. Re-creation of a landscape with sustainable production and living conditions (river basin management) III. Project planning 1. Pre-project investigations; the holistic approach in space and time A. The long-term development of the ecosystem; palaeolimnological investigations; the ecological conditions immediately before degradation; the status from a regio-limnological point of view B. Changes in the catchment area; present relationships between catchment area-lake/wetland; the external loading; what is the regio-limnological normal loading? Input and output of water and nutrients, calculation of budgets C. Internal ecosystem conditions; investigations during at least 1 year (a) The bottom – sediment, peat: horizontal distribution, stratigraphy and functional role in the ecosystem (seasonal changes) (b) The water – chemical and physical factors: nutrients, pH, transparency, etc., diurnal and seasonal patterns (c) Organism conditions: plankton, primary productivity, macrophytic vegetation, bottom fauna, vertebrates; changes in evapotranspiration conditions of importance for water-level fluctuations (d) Sediment growth rate in lakes, deposition rate of coarse detritus in wetlands (accumulation basins) 2. Project design and implementation A. The degraded ecosystem, its structure and function; ecological diagnoses; remedial eco-medical measures: ‘strict diet, medication, surgery’ B. Design of restoration method(s), calculations of cost (a) Measures to be taken in the catchment area (b) Measures to be taken in the lake/wetland ecosystem C. Information, argumentation, convincing; the public, politicians, administrators; the past and present situations compared with the future development of the ecosystem, with and without restoration measures; cooperation with decision-makers and contractors
74
S. Björk
D. Finalisation of financial commitment; decision-making E. Accomplishment of eco-technical measures (a) Ecological and technical cooperation (b) Ecological control programme F. Measures for food web management, if considered realistic, to tuneup the system 3. Post-project monitoring and project evaluation A. Post-project investigations/monitoring during several years for control of sustainable results (a) External and internal nutrient loading (b) Collection of chemical and physical data, investigations on organism conditions B. Ecological syntheses of analytical data; both foreseen and unforeseen positive and negative results should be reported; before-and-after comparisons utilised for training and education of students
5.3 Internal and External Nutrient Control Inland water ecosystems reflect the character – primarily the geology – of the catchment area, with surface and groundwater as carriers of solid and dissolved matter. Through rainwater and dry deposition, these systems also mirror conditions and activities transmitted from the air. The external loading of nutrients is decisive for the original productivity in water bodies – within the limits set by light, temperature, precipitation, water renewal, etc. for different areas at different latitudes. Properly-working lake ecosystems act as sinks for organic and minerogenic matter and this material in turn participates in adsorption reactions and chemical-complex formation. Nutrients are accumulated in the sediments and bound there. However, overloading by nutrients from external sources gives rise to an increased production of plant material in the lake and this in turn causes a spectrum of processes at the water-sediment interface. These processes may result in the leakage of nutrients accumulated in the sediment. This surplus supply of nutrients to the lake water constitutes the internal loading. When the internal loading is added to the external one, a lake ecosystem undergoes sudden and dramatic changes in the form of a rapid increase in primary productivity. This leads to a fast degradation of the whole system – severe oxygen deficiency, increased sediment growth rate, fish-kills, etc. Such changes in the character of an ecosystem hit by internal loading, were first described and explained by Ohle (1955, 1965, 1971) and Thomas (1955, 1963), who described the changes as being very sudden. As already stressed, only after the external loading has been reduced to, or close to, normal, should the restoration methods and techniques aimed at a reduction or
5 Principles, Planning and Accomplishment of Lake Restoration Projects
75
elimination of the internal loading in irreversibly damaged lakes be applied. Phosphorus precipitation, aeration, sediment treatment and sediment removal are all methods to counteract the internal loading and to transform the ecosystem from a nutrient source to a nutrient sink. If sustainable results are to be achieved, methods for the reduction and control of macrophytic vegetation have to be combined with methods for the adjustment of environmental conditions, as for example, by increasing water depth. Efforts to manage the food web might be a method to tune up the function of an ecosystem after the nutrient level has been brought under control and the necessary environmental conditions for all stages of the life cycle of esteemed species have been realised.
References Ohle W (1955) Die Ursachen der rasanten Seeneutrophierung (The causes of the very rapid eutrophication of lakes). Verh Int Verein Limnol 12:373–382, in German Ohle W (1965) Nahrstoffanreicherung der Gewässer durch Düngemittel und Meliorationen (Nutrient enrichment of water bodies by fertilisers and drainage). Münchner Beiträge 12:54–83, in German Ohle W (1971) Gewässereutrophierung (Water bodies and their surroundings as ecologicaI units in their importance for the eutrophication of waters). Gewässerschutz, Wasser, Abwasser, Aachen: 437–456 (in German) Thomas E (1955) Stoffhaushalt und Sedimentation im oligotrophen Aegerisee und im eutrophen Pfäffiker und Greifensee (Matter budgets and sedimentation in the oligotrophic Lake Aegeri and in eutrophic Lakes Pfäffiker und Greifen). Mem Ist Ital Idrobiol 8:357–465, in German Thomas E (1963) Experimentelle Untersuchungen über die Schlammbildung in unberührten und kulturbeeinflussten Seen der Schweiz (Experimental investigations on the formation of sediment in intact and human influenced lakes in Switzerland). Wasser und Abwasser. 1–21 (in German)
Chapter 6
Restoration of Eutrophic Lakes by Sediment Treatment Wilhelm Ripl
Abstract Sediment treatment with nitrate as an oxidising agent has been used successfully on Lake Lillesjön in southern Sweden, smaller ponds near Berlin or at the Alte Donau, a eutrophicated dead-arm of the River Danube in Vienna. Specially-designed equipment has been used to inject nitrate solution directly into the sediment so as to minimise the risk of increasing the nitrogen concentration in the water. Iron salts are usually added to enhance the binding of phosphorus in the sediment. Reactions involved when such treatment is implemented are processes that normally occur naturally in lakes in the top sediment layers. However, with an excessive input of nutrients to a lake from its catchment, the natural process of phosphorus binding becomes insufficient; due to the oxygen depletion at the bottom of the lake, phosphorus is released from the sediment to the water, enhancing primary production and thus disturbing the balance between production and decomposition of organic matter. To prevent further degradation of the lake ecosystem, sediment treatment can be used. Keywords Alte Donau • Lake eutrophication • Lake Lillesjön • Oxidation of organic matter • Phosphorus precipitation • Sediment treatment
6.1 Introduction Eutrophication problems in lakes are, generally, caused by an increased flow of nutrients and/or degradable organic matter to the lake basin. If such an excessive inflow has continued over a long period of time, the metabolism in the lake will change and the structure and function of the sedimenťs will alter. The excessive nutrient supply leads to increased primary production, which is often not adequately W. Ripl (*) Technical University of Berlin, Nuthestrasse 4A, D-14513 Teltow, Germany e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_6, © Springer Science+Business Media B.V. 2010
77
78
W. Ripl
compensated by decomposition processes due to an insufficient oxygen supply. This means that both nutrients and organic matter are accumulated at the sediment surface, leaving sediments poised at low redox conditions and in a state of high potential reactivity. Such conditions prevail even after the external loading has been reduced. The high potential reactivity of the sediment may lead to a very sudden change in lake metabolism. At the same time the lake is buffered against any effect from a reduction in external loading, a behaviour that could be referred to as ‘hysteresis’ or a delay in reaction after changed external conditions. If a quick improvement of a lake’s condition is required, sediment treatment in situ is a valuable option. However, it has to be stressed that any internal lake measures should always be preceded by measures that reduce the external loading from the catchment. Otherwise, any short periods of improvement obtained will be limited by the metabolism caused by external nutrient or energy (organic matter) loading.
6.2 Internal Restoration Measures and Processes Involved 6.2.1 Mineralisation of Organic Matter Usually, the first problem to tackle is to increase mineralisation processes in step with production by supplying oxygen to the top sediment layer. The supply of oxygen is necessarily accompanied by water movement thus increasing reactivity in a transport-limited sediment surface. The use of pure oxygen as well as the addition of very strong oxidising agents such as potassium permanganate and peroxides have both been tried. However, these substances proved to have a very low oxidation efficiency caused by the loss of gaseous oxygen and partly also by the sterilisation of the sediment surface. In contrast, aeration measures using compressed air have shown good efficiency in some cases, especially in deeper, drinking water reservoirs having small areas of sediment to be treated and having dissolution problems of iron and manganese. High concentrations of iron and manganese increase the treatment costs for the preparation of drinking water. In most eutrophicated lakes, however, the majority of sediments have accumulated over long time periods and much smaller layers have been deposited during periods of eutrophication. This is despite the fact that deposition rates during recent periods of eutrophication can be at least ten times or more that of earlier sedimentation rates. In these cases, a combined treatment, increasing phosphorus-binding on the one hand and oxidation of easily degradable organic matter on the other, can be a solution.
6.2.2 Combined Sediment Treatment with Nitrate and Iron The combined measures of iron treatment and oxidation by biochemical denitrification (with nitrate used as an oxidising agent) were applied, for the first time, in Lake
6 Restoration of Eutrophic Lakes by Sediment Treatment
79
Table 6.1 Lake Lillesjön (southern Sweden, treatment accomplished in 1975) Catchment area 1.0 km2 Lake area 4.2 ha Mean depth 2.0 m Maximum depth 4.2 m Treated sediment area 1.2 ha Chemicals used 13 t Fe Cl3; 5 t Ca (OH)2; 12 t Ca (NO3)2 Methods of application Direct injection by harrow Treatment period 3–4 weeks Further measures implemented −Stoppage of sewage inlet −Removal of vegetation
Lillesjön (southern Sweden) where calcium and iron salts were used. The experience has shown that a single treatment can change the lake metabolism instantly and provide water of bathing quality for a long period. Lake Lillesjön was treated in 1975 and its metabolism has not significantly changed since. Some data for the treatment of Lake Lillesjön and of Hambutten and Karutschen ponds (Berlin) are summarised in Tables 6.1 and 6.2. The oxidising agent used is nitrate and the reactions involved are processes that occur naturally in lakes in the top sediment layers. In lakes with relatively low nutrient loading, the phosphorus from the catchment is quickly precipitated with the iron or lime compounds that enter the lake with groundwater recharge. Such an immediate precipitation, during periods of high flow in winter and spring, leaves very little phosphorus available for maintaining biological production activities in the pelagic zone of the lake. The low planktonic production in such lakes is mineralised by a sufficient oxygen supply during the water circulation twice a year. The retention of phosphorus in the sediment is very high because of the greater than stoichiometric content of iron, or other phosphorus precipitating agents, in relation to the phosphorus inflow. An increased production in the lake, due to nutrient contributions or an input of organic matter from the catchment, leads to the accumulation of degradable organic matter at the sediment surface layer. This enrichment of the substratum is followed by increased bacterial activities and increased uptake of oxygen, causing extended periods of anoxic conditions at the sediment surface. This spontaneously reduces the oxidation potential, and other electron acceptors, rather than oxygen, are used for mineralisation (denitrification, desulphurication, methane production). In this reaction scheme, denitrification is the first process. This provides a terminal oxidation of organic matter by the production of carbon dioxide and the liberation of molecular nitrogen, described by the following chemical equation:
4 NO3 − + 5 CH 2 O → 5 CO2 + 2N 2 + 3H 2 O + 4 OH −
(6.1)
This process is still in a redox range where, iron is not reduced to bivalent ferrous ions, or where, hydrogen sulphide competes for ferrous ions with the phosphorus ions. This means that the phosphorus remains undissolved from the sediments and does not lead to enhanced primary production in the pelagic waters.
80
W. Ripl
Table 6.2 Hambuttenpfuhl and Karutschenpfuhl (small ponds within Berlin, treatment accomplished in 1991) Hambuttenpfuhl Karutschenpfuhl Catchment area 2.4 ha 23.3 ha Lake area 0.28 ha 0.29 ha Mean depth 0.8 m 0.9 m Maximum depth 1.9 m 3.7 m Treated sediment area Approx. total lake area Approx. total lake area Chemicals used 4 t Ca (NO3)2 4 t Ca (NO3)2 Method of application Entry into the lake water Entry into the lake water Treatment period Several days Several days −Iron treatment of the Further measures −Iron treatment of the sediment sediment implemented −Aeration by circulating plant −Aeration by circulating plant −Clearing of shore trees to −Clearing of shore trees to minimise foliage (leaf litter) minimise foliage (leaf litter)
Although a natural process, denitrification is not very significant in natural lakes because nitrogen in the oxidised state is seldom provided in sufficient amounts by tributaries. This is especially true during the periods of stagnation when oxidised nitrogen compounds quickly disappear by this process. After the disappearance of nitrate, desulphurication takes place in the sediments, as described in Eq. 6.2. The process of desulphurication oxidises organic matter and produces CO2, thereby reducing sulphate.
2 CH 2 O + SO 4 2 − + 2H + → H 2 S + 2 CO2 + 2H 2 O
(6.2)
Usually, sulphate is ubiquitously present in water at sufficient concentrations, maintaining desulphurication at the sediment surface. However, H2S is produced and this compound depletes any remaining traces of oxygen, the last traces of oxidised nitrous compounds, and reacts with iron to form ferrous sulphide, which leads to the liberation of phosphates (Eq. 6.4) from the sediments in just the moment when all ferrous hydroxides reacted to sulphides (Eq. 6.3):
2 FeO (OH )+ 3H 2 S → 2 FeS + S + 4H 2 O
(6.3)
2 FePO 4 + 3H 2 S → 2 FeS + 2 PO34− + S + 6H +
(6.4)
At this moment, free hydrogen sulphide poisons the sediments and prevents higher faunistic components in the benthic areas. Usually, when this process is stopped by the lack of sulphate in the deeper sediment layers, methane production takes place:
2 CH 2 O → CH 4 + CO2
(6.5)
6 Restoration of Eutrophic Lakes by Sediment Treatment
81
Since methane is only slightly soluble in water, sediment mixing occurs by methane bubbles and phosphates are liberated to the pelagic waters causing excessive primary production. Knowledge of these processes occurring in the sediment, along with their time and depth distribution, provides possibilities for a tailor-made sediment treatment with the aim of inactivating the sediment. Usually, the application of both nitrate and iron at the sediment surface is necessary, and ideally is conducted in the latter part of the spring circulation. Experience has shown that it is convenient to start with the iron treatment as it destabilises the hydrogen sulphide and other sulphide compounds in the sediment (Eq. 6.3). Thus these reduced compounds do not have to be oxidised in total by added calcium-nitrate solutions, which makes the nitrate treatment a little more efficient. Prior to the implementation of a sediment treatment in a given lake, laboratory experiments with the lake sediment should be conducted to analyse the iron, phosphorus, and energetic conditions (amount of easy degradable organic substances) in the sediment. Various additions of the agents have to be tested with the sediments in question. By conducting experiments with the iron and nitrate treatment, both the areal dosage and the concentrations of these agents are optimised. For nitrate, the penetration of the added calcium nitrate through the surface sediment layers has to be examined. With respect to iron, it is necessary to establish a dosage sufficient for excessive phosphorus binding, even if a desulphurication process in the sediments should occur. The case studies mentioned (Lake Lillesjön, Hambutten and Karutschen ponds, and Lake GrossGlienicker in Chapter 7) can only be regarded as examples. The dosages used in these cases should not be seen as general recommendations. In water bodies with low alkalinity, the introduction of iron compounds, in the form of iron chloride, could use up the buffer capacity, which would result in acid conditions in such lakes:
2 FeCl3 + 3H 2 O → Fe 2 O3 + 6H + + 6 Cl −
(6.6)
Acid conditions, however, retard the denitrification activity. Although the coenotic structure in the lake will change, few toxic effects to fishes and benthic organisms can be observed, as long as pH in the lake does not fall below about pH 6. In order to prevent acid conditions, it might be necessary to add calcium carbonate, in the form of fine particles with a high reactivity, together with the iron, to provide a sufficient amount of buffer in the water and at the sediment surface. In some cases, there is a high amount of iron accumulated in the sediments and this iron is sufficient for binding the phosphorus. In such lakes, the addition of iron can be omitted, and the application of nitrate would be sufficient. In other lakes, where iron has been depleted in the surface sediments by low redox conditions for long periods, only the addition of iron would be sufficient, especially when this measure is carried out in very shallow lakes, where water movement and oxygenation of the sediment surface is provided throughout the
82
W. Ripl
whole year. This is also true for lakes where artificial mixing or aeration are carried out at the same time, as was the case in Lake Gross-Glienicker (see Chapter 7). In our whole-lake studies, it was shown that for internal sediment-stabilising measures, each lake has to be treated individually according to the morphometric conditions on the one hand, and the water transport through the system and its pattern during the year on the other. This shows that standardised dosages of chemicals or standardised areas to be treated, in relation to the total lake area, cannot be given. The treatment can only be evaluated by thorough investigations over an adequate time period, with a sufficient amount of sediment sampling points in each lake. The sufficient number of sampling points required is determined by the homogeneity and isotropy (distribution) of the sedimentation process in each specific lake.
6.3 Application of the Chemicals to the Lake and Evaluation of Restoration Effects In the first whole-lake treatment experiments, the chemicals were injected into the sediment with a harrow-like device dragged over the sediment surface in order to cover the total area to be treated. For Lake Lillesjön the treatment device is shown schematically in Fig. 6.1. The concentrations of the treatment solutions were chosen in such a way that the distribution of the chemicals at the sediment surface was controlled by the gravity of the solutions. It was shown that such treatments give excellent results; however, the costs of the chemical distribution are extremely high. Recently, some experiments have been conducted with mixing the chemicals with the water of lake tributaries, or adding nitrate to the lake with dephosphorised and nitrified effluents from treatment plants. It was shown that even these kinds of treatment, when applied in a proper way, produce good and lasting results.
Chemicals FeCl3
Air
Ca(OH)2 Ca(NO3)2
Epilimnion
Hypolimnion Sediment
Fig. 6.1 Equipment used for the treatment of Lake Lillesjön, southern Sweden
6 Restoration of Eutrophic Lakes by Sediment Treatment
83
The application of dephosphorised and nitrified effluents from a sewage plant in Schleswig to the Schlei estuary (northern Germany) is shown in Fig. 6.2. Iron applications were conducted by premixing the agents on shore and distributing the iron over the lake with tubing mounted on a boat. The iron solution was introduced at about one metre below the water surface. Of course, the iron flocks colour the water temporarily. But within 1 or 2 days, the iron flocks settle to the bottom, leaving a dephosphorised water which is necessary to prevent excessive primary production in spring and avoid the loading from organic degradable matter by algal sedimentation. The phosphorus bond to iron is possible through two different processes: the phosphate can be bound in iron-phosphate or adsorbed to iron oxide-hydroxide (see Section 7.2.2). More recently, a treatment with iron chloride, combined with calcium carbonate for buffering and with an application of nitrate for oxidation of superficial sapropels was undertaken at Alte Donau – an eutrophicated dead-arm of the River Danube in Vienna that has been isolated from the river since 1875 as a result of regulation measures. Iron and nitrate treatment was carried out in 1995 and 1996 (Ripl and Wolter 1995; Ladinig 1995). Following the treatment, the content of phosphorus and seston in the lake water decreased (Donabaum et al. 1999) and after some initial planting, submersed macrophytes spread over the lake. The restoration of this 1.5 km2 Lake Alte Donau was a similar success to that of Lake Gross-Glienicker (Chapter 7) and a valuable reference for restoration of eutrophicated lakes by phosphorus binding.
Fig. 6.2 Proposal for the treatment of effluent of a sewage plant in Schleswig (Schlei estuary, northern Germany)
84
W. Ripl
References Donabaum K, Schagerl M, Teubner K, Dokulil MT (1999) Sanierung und Restaurierung der Alten Donau in Wien, Österreich. – Deutsche Gesellschaft für Limnologie (DGL) – Tagungsberichte 1998 (Klagenfurt): 264–268, Tutzing 1999 Ladinig G (1995) Gewässersanierung: Beispiel Alte Donau. Perspektiven 5(1995):46–51 Ripl W, Wolter K-D (1995) Sanierung Alte Donau (Wien). Begleituntersuchung zur kombinierten Eisen- und Nitratbehandlung. Textteil und Anhang: Graphiken. Im Auftrag der Stadt Wien, Magistratabteilung 45 – Wasserbau. Aquaterra-Consult-Gesellschaft m. b. H., Wien 77 pp. + annex
Additional Literature Rheinheimer G (1991) Mikrobiologie der Gewässer (Microbiology of water bodies). 5 Auf. – G. Fischer, Jena (in German) Ripl W (1976) Biochemical oxidation of polluted lake sediment with nitrate. A new restoration method. Ambio 5:312–135 Ripl W (1978) Oxidation of lake sediments with nitrate. A restoration method for former recipients. Coden Lunbds/(NBLl-1001)/1-151/(1989). ISSN 0348-0798 Ripl W (1983) Dümmersanierung. (Restoration of the Dümmer). TU-Berlin, FB 14, Fachgebiet Limnologie, Berlin (in German) Ripl W (1986) Restaurierung der Schlei. Bericht über ein Forschungsvorhaben. (Restoration of the Schlei estuary. Report on a research project). Hrsg.: Landesamt für Wasserhaushalt und Küsten Schleswig-Holstein, D 5. Kiel. (in German) Ripl W, Feibicke M (1992) Nitrogen metabolism in ecosystems. A new approach. Int Revue Ges Hydrobiol 77:5–27 Ripl W, Lindmark G (1978) Ecosystem control by nitrogen sediment metabolism. Vatten 34:135–144 Ripl, W. and Wolter, K.-D. (1993). Sanierungsmaßnahmen am Hambutten- und Karutschenpfuhl. (Restoration measures at Hambutten and Karutschen ponds). TU-Berlin, Gesellschaft für Gewässerbewirtrschaftung mbH, Selbstdruck, Berlin (in German) Ripl W, Leonardson L, Lindmark G, Andersson G, Cronber G (1979) Optimering av reningsverk/ recipient-system. Vatten 35:96–103 (in Swedish) Ripl W, Motter M, Wesseler E, Fischer W (1988) Regional-ökologische Studien zum Plankton und Benthos Berliner Gewässer (Regional ecological studies of plankton and benthos in water bodies of Berlin). TUB, Institut für Ökologie: Selbstdruck, Berlin (in German) Ripl W, Feibicke M, Heller S (1993) Nahrstoffelimination aus einem gering belasteten Fließgewässer mit Hilf eines bewirtschafteten Schilfpolders. (Nutrient elimination from a river with a low load by means of reedbeds). Jahresbericht 1992. TU-Berlin, FB 14, Fachgebiet Limnologie, GFGmbH, Berlin (in German) Stumm W, Morgan JJ (1981) Aquatic chemistry, 2nd edn. Wiley-Interscience, New York
Chapter 7
Restoration of Eutrophic Lakes by Phosphorus Precipitation, with a Case Study on Lake Gross-Glienicker Klaus-Dieter Wolter
Abstract After diversion of their external nutrient loading, eutrophic lakes with a high internal phosphorus release from their sediments may be restored by phosphorus precipitation. Iron or alum compounds are applied to the water to bind and subsequently sediment out phosphorus from the water column. As an example, the successful restoration of Lake Gross-Glienicker (Germany; Berlin/Brandenburg) is described. The injection of iron compounds into the lake (0.68 km2, mean depth 6.5 m) led to a reduction of phosphorus from 0.5 to about 0.03–0.04 mg l−1. The Secchi depth increased to the extent that the lake has become popular for bathing and has developed submersed vegetation. In general, the biocoenosis of the lake has improved. Boundary conditions for the application of iron (e.g. pH and alkalinity, sulphur compounds, combination with hypolimnetic aeration, hydrology of the lake and its near surroundings) and further measures to increase the quality of the lake are described. Keywords Iron treatment • Lake Gross-Glienicker • Lake restoration • Phosphorus precipitation
7.1 Introduction High concentrations of phosphorus in a lake’s water column will usually cause an increased primary production and undesirable algal blooms. To counteract this in some lakes elsewhere, an application of phosphorus-binding compounds to the water column has been used to inactivate the phosphorus, thus decreasing its availability for primary producers. It should, however, be stressed that this method
K.-D. Wolter (*) Kolberger Street 3, D-65191 Wiesbaden, Germany e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_7, © Springer Science+Business Media B.V. 2010
85
86
K.-D. Wolter
should only be applied after the input of phosphorus from the catchment has been reduced and any considerable loading of phosphorus to the lake has ended. Even after the diversion of nutrient loading from a lake, some lakes continue to show prolonged high nutrient concentrations – especially of phosphorus. If this is observed in a lake with a rapid water exchange (retention time considerably less than 1 year), it indicates either additional, but as yet unrecognised, sources of external loading or a high phosphorus release from the lake’s sediment – i.e. internal loading. This nutrient release from the sediment can be minimised by a suitable sediment treatment (see Chapter 6).
7.2 Phosphorus Precipitation from the Water Column In lakes with long retention times (considerably more than 1 year), lake recovery can be accelerated by the precipitation of phosphorus out of the water column. For this purpose, alum (aluminium sulphate, Al2[SO4]3) or iron compounds (iron-IIIchloride, FeCl3) can be used. They form relatively stable phosphorus-binding compounds which sediment out in the form of a gelatinous floc. As these preparations always have to be used in an over-stoichiometric amount (relative to the phosphorus in the water), an additional phosphorus-binding capacity is created at the sediment surface. For this reason, every in-lake preci pitation may be viewed as a sediment treatment too. The principle conditions formulated in Chapter 6 are also valid for phosphorus precipitation from the water column. Beside aluminium and iron, some authors also recommend the use of lime. In this case, over-saturated calcium oxide or hydroxide solutions are applied to the water. Calcium together with hydroxide and phosphate forms apatite (Eq. 7.1), a phosphorus-containing mineral, but no successful treatments of lakes with this preparation are known:
10 Ca 2 + + 2 OH − + 6 PO34− → Ca10 (PO 4 )6 (OH )2
(7.1)
Another alternative proposed by several authors has been the use of fly ash from the combustion of coal or oil. Fly ash may bind phosphorus by its high specific surface and the content of CaO, MgO, and Al (III). Nevertheless, the use of fly ash is not recommended because of the toxic effects caused by heavy metals or by the drastic increase or decrease in pH that has been observed after the treatment. The methods of phosphorus precipitation from the water column are best suited for smaller, shallow lakes. The application of the in-lake treatment should always be preceded by a significant reduction of nutrient loading from the catchment area in order to achieve effective and long-lasting results. Otherwise, the positive effects of treatment can be overwhelmed by the continued external loading in just a few months.
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
87
7.2.1 The Use of Aluminium Sulphate (Alum) In the case of aluminium, inorganic phosphate is bound directly or by adsorption to aluminium hydroxide (Eqs. 7.2 and 7.3):
Al (OH )3 + PO 4 − → AlPO 4 + 3 OH −
(7.2)
Al (OH )3 + PO 4 3 − → Al (OH )3 ~ PO 4 3 − (aq )
(7.3)
The best phosphorus fraction to bind to aluminium is phosphate-phosphorus. Organic fractions of phosphorus are bound less. Dissolved organic phosphorus is less effectively removed than particulate phosphorus. Total phosphorus in the form of particulate phosphorus (algae, detritus, bacteria) cannot be bound to aluminium hydroxide directly, but the particles may be enclosed into the forming aluminium flocs and subsequently settle to the sediment. For this reason, precipitation may also be carried out during the summer vegetation period. In this case, bathing should be prohibited during the treatment. However, the trapping of particles into the aluminium flocs only occurs when high concentrations of alum (above 5 g Al3+ m−3) are used, which may be restricted by the cost of the chemicals or the lack of buffering capacity (alkalinity) in the lake water (see also Section 7.2.3). The surplus aluminium settles to the sediment and increases its phosphorusbinding capacity. With time and with further sedimentation on top, this static layer of aluminium is buried deeper in the sediment. Although phosphorus binding to this layer can be observed in the laboratory for several years, in practice the burying of the aluminium leads to a quick cessation of the phosphorus binding to the aluminium particles. For this reason, long-lasting phosphorus trapping probably cannot be achieved. However, compared to iron the use of alum is cheaper.
7.2.1.1 Toxicity of Alum The use of aluminium sulphate has often been reported to have toxic effects on organisms. Therefore, treatment with alum has to be carried out with care. Concentrations below 50 mg Al l−1 in the lake water during treatment will probably have no harmful effects on organisms. In low-buffered waters, the treatment with alum has led to the enrichment of aluminium in fishes. There are indications of bioaccumulation in rainbow trout (Oncorhynchus mykis) and chronic toxicity in Chironomidae (Cooke et al. 1986). In some cases aluminium is taken up by aquatic plants, thus reducing their physiological ability of phosphorus-uptake by the roots. Laboratory experiments with aluminium and also with iron have often shown increased mortality of certain species. In lakes, these effects are usually lower because the unhomogeneous distribution of iron and aluminium during treatment allows for escape reactions by the organisms.
88
K.-D. Wolter
7.2.2 Precipitation with Iron Phosphorus precipitation with iron has been used in many cases. Similarly as with aluminium, the phosphate is bound in iron-phosphate minerals or it is adsorbed to iron oxide-hydroxide (Eqs. 7.4 and 7.5):
FeO (OH )+ H 3 PO 4 → FePO 4 + 2H 2 O
(7.4)
FeO (OH )+ PO 4 3 − → FeO (OH ) ~ PO 4 3 − (aq )
(7.5)
From the phosphorus fractions, inorganic phosphate is best bound by iron. No significant bond of organic phosphorus has been observed. In contrast to aluminium, it is not possible to precipitate organic phosphorus-containing particles by their inclusion into the forming iron flocs. For iron, chronic toxic effects on organisms in the lake are unknown. After sedimentation, iron continues to bind phosphorus at the sediment surface. In contrast to aluminium, the bond of phosphorus to iron is redox sensitive. When the sediments become anoxic and hydrogen sulphide is formed, iron-bound phosphorus is released from the sediment and may be transported to the water column. To prevent this event, in many cases the treatment of sediment (see Chapter 6) has to be combined with phosphorus precipitation in the water column. However, the redox-sensitive behaviour of iron has also some positive aspects. As lakes and their sediments are dynamic structures being altered by processes of production, sedimentation and respiration, the dissolution of iron may be viewed as a process enhancing the iron buffering capacity. Iron dissolved in deeper sediment layers migrates along its concentration gradient to the sediment surface. Generally, an enrichment of iron occurs at the reductive/oxidative boundary layer near the sediment surface where the exchange of phosphorus with the water and especially the binding of phosphorus from the water is a quantitatively significant process. Therefore, iron can be seen as a dynamic phosphorus trap, also functioning if the lake system shows some spatially- or temporally-restricted increase in phosphorus concentration.
7.2.3 Effect of Aluminium and Iron Compounds on pH of Lake Water Aluminium sulphate and iron chloride have to be used with care because of their effect on the pH of the water. When applied to the water, the preparations undergo hydrolysis which liberates protons as described in the reactions (Eqs. 7.6 and 7.7) below:
Al 2 (S0 4 )3 + 6H 2 O → 2 Al (OH )3 + 3 SO 4 2 − + 6H +
(7.6)
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
2 FeCl3 + 3H 2 O → Fe 2 O3 + 6H + + 6 Cl −
89
(7.7)
In water bodies with low alkalinity, the introduction of iron chloride or a luminium sulphate could use up buffer capacity, which would result in acid conditions in the lake. Although the coenotic structure in the lake will change, toxic effects on fishes and benthic organisms are usually not observed, as long as pH does not fall below about 6. In order to prevent acid conditions, it might be necessary to add calcium carbonate (in the form of fine particles having a high reactivity) together with the iron or aluminium preparations, to provide a sufficient amount of buffer in the water and at the sediment surface.
7.3 Treatment Schedule Prior to the in-lake treatment, it is very important to make an analysis of processes involved in phosphorus metabolism in the lake. The measurement of concentrations of individual phosphorus fractions during one or more years would allow the identification of the most suitable period for treatment. The best time for phosphorus precipitation is when the phosphate fraction is at its relative maximum – usually from late autumn to early spring. The treatment of the lake has to be finished before intensive plankton growth starts in spring. The use of solid preparations is recommended only when these can be distributed over an ice area with high weight-bearing capacity. Fine solid particles of the acidforming preparations have to be handled with care. Distribution from aeroplanes may allow fine particles to drift and cause damage to the lake surroundings. After applying the preparations on the ice, its structure changes, the ice weakens, and should not be walked on. Most often, the preparations are applied in dissolved form. Figure 7.1 shows a schematic illustration of the application of iron oxide-hydroxide and iron chloride to the lake. The solutions are transported to the lake by flexible, high-pressure polythene tubing and applied by a harrow from the boat to the lake. In the case that high amounts of dissolved organic substances in the water prevents flocculation of the iron or aluminium compounds, fine mineral particles (clay, bentonite) may be added to reduce the activity of these organic substances. During the lake treatment, chemical analyses of the water should be made, including: • • • •
Iron or aluminium concentrations Concentrations of individual phosphorus fractions pH Alkalinity
After the treatment, these parameters along with parameters indicating the biological production of the lake should be monitored.
90
K.-D. Wolter
Fig. 7.1 Schematic illustration of the application of iron oxide-hydroxide and iron chloride to the lake water
7.4 Case Study of Lake Gross-Glienicker 7.4.1 Introduction Lake Gross-Glienicker (in German, Gross-Glienicker See), located right on the former border between West-Berlin and former East Germany, used to be one of the best lakes in Berlin, in terms of water quality, during the 1950s and 1960s. However, in the 1970s and 1980s, the lake showed increasing concentrations of phosphorus. After the unification of Germany in 1989, it was decided to restore the lake.
7.4.2 Description of the Lake and Its Degradation Lake Gross-Glienicker developed from a subglacial channel during the last Weichselian glaciation. It is oriented north–south, with its channel-shaped basin being 1,800 m long and about 400 m wide. The lake has an area of 0.68 km2, its volume being 4.2 Mm3 with a maximum depth of 11 m and mean depth of 6.5 m. The lake catchment area of about 16 km2, determined by the topography of the surroundings, is mostly occupied by urban settlements, agriculture and forestry (Fig. 7.2).
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
91
Fig. 7.2 Land use in the catchment area of Lake Gross-Glienicker before 1990
In the 1920s and 1930s, any urbanisation had the character of a holiday s ettlement. However, since the 1950s and 1960s, more and more houses were built – but with no central sewage treatment being installed. Sewage was disposed into septic tanks and it is likely that some of them were not water tight and acted like soakaways, thereby loading the groundwater with nutrients. In 1972, military barracks were built to the north of the lake; their sewage treatment plant was virtually ineffective and, after passing two little ponds, ‘untreated’ sewage was entering Lake Gross-Glienicker. About 380 m3 of sewage reached the lake per day and this was most likely one of the most significant sources of nutrients. Since the 1920s or 1930s, a waterworks had been located on the east side of the lake. As a consequence, the water table had been lowered, preventing some of the nutrientloaded groundwater from flowing into the lake. However, pumping of the groundwater also brought negative consequences: the soil and aquifer were degraded by the leaching of nutrient-binding agents like iron, small silt and clay particles. It was probably for this reason that, within a few years after the closure of the waterworks in 1977, the concentration of phosphorus in the lake remarkably increased.
7.4.3 Processes in the Lake During the Degradation Period The summer water transparency (as measured during 1989–1992) decreased to values lower than 1 m (see Fig. 7.6) and blue-green algae became an important
92
K.-D. Wolter
phytoplankton group. In the hypolimnion, oxygen was depleted usually by May (soon after the establishment of thermal stratification; the thermal stratification of Lake Gross-Glienicker being of a rather high stability with a summer thermocline between 5 and 6 m) and as a consequence, hydrogen sulphide was formed. The littoral zone of the lake was poorly developed; although some reeds occurred, submersed macrophytes were hardly evident. The positive effect of such structures increasing surface areas and thus improving nutrient cycling was therefore missing. Communities in the lake were dominated by phytoplankton. Due to the production of hydrogen sulphide, the benthos was poorly developed. The retention of phosphorus in sediments is controlled by processes of sulphate and nitrate reduction. In Lake Gross-Glienicker, the large amount of easily-degradable organic matter produced by phytoplankton in summer settled to the bottom of the lake. As oxygen and oxidised nitrous compounds (nitrate) were lacking in summer, the organic substances in the sediments were mainly degraded by sulphate reduction: in this process hydrogen sulphide is formed. The hydrogen sulphide can reduce iron and form iron sulphide – thus releasing the iron-bound phosphorus. Sediments with high S/Fe-ratios (above 1 or 1.5) do not normally show a good phosphorus-binding capacity (see also Chapter 6). The status of a lake’s sediments can be evaluated with a sediment analysis of iron, sulphur and phosphorus. Before iron treatment in Lake Gross-Glienicker, the molar ratio of sulphur to iron (S/Fe) was 1.3, indicating that a relatively high amount of iron was present in the form of iron sulphide. Hence, relatively small concentrations of phosphorus were found in the sediments. Pre-investigations showed the upper sediment layer to contain: about 33% dry matter of organic substances, most of them easily degradable; about 22% dry matter of lime; and up to 41% dry matter of acid-insoluble residues (AIR). Besides these main components, the sediment content of iron, sulphur, and phosphorus is particularly important for the evaluation of phosphorus exchange between water and sediment. In Lake Gross-Glienicker, sulphur reached 1.5% dry matter and phosphorus about 0.2% dry matter, while the iron content of 2% dry matter was very low in comparison to other Berlin lakes. Related to the geological structure of their catchment areas, some lakes in the Berlin region contain much more iron in their sediments. Under oxidised conditions, this additional iron is able to bind phosphorus. Knowledge of these processes helps anticipate the process determining the phosphorus concentration in the water. In Lake Gross-Glienicker, analyses of lake water had started more than 3 years prior to its restoration was undertaken. As the lake restoration proceeded, activities such as phosphorus inactivation by the addition of iron, and additionally mixing the water by aeration, led to significant changes in the concentrations of all chemical compounds involved. An adequate assessment of how these processes altered during restoration can only be made with knowledge of the processes prior to restoration. In temperate lakes, a major process causing phosphorus-related problems is the net release of phosphorus from the sediment during the vegetation period. In Lake
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
93
tot-P mg/l
Tegeler, in Berlin, for example, net phosphorus release can be demonstrated by examining the phosphorus-balance – this is obtained by measuring phosphorus input, output, and its content in the lake (Ripl et al. 1993). In Lake Gross-Glienicker, no measurement or estimation of inflows and outflows were made, and therefore a balance of nutrients could not be calculated. However, as it was assumed that Lake Gross-Glienicker had no significant exchange of water (inflow and outflow) during summer, some conclusions about phosphorus processes were made – just by knowing the concentration pattern of phosphorus in the whole water column, in epilimnion, and hypolimnion. The volume weighted concentration of phosphorus in the hypolimnion increased between March and August in the years 1989–1992 (Fig. 7.3, filled triangles). Due to the overcompensation of the phosphorus concentration increase in the hypolimnion by its decrease in the more voluminous epilimnion (Fig. 7.3, open triangles), the water column as a whole lost phosphorus (Fig. 7.3, middle line). This means that, under the circumstances prevailing at that time, sedimentation was the main phosphorus process during summer. Compared to other lakes, which show internal loading of phosphorus during summer, this pattern of Lake Gross-Glienicker was untypical – caused by the lack of iron in its sediment. The increase of total phosphorus during the cold season, from about September to March, was probably caused by mineralisation of the organic sediment and the inflow of phosphorus-rich water via surface or ground waters.
1,2 1,0 0,8 0,6 0,4 0,2
iron treatment
0,10 0,05 0,00
PO4-P mg/l
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 1,0 0,8 0,6 0,4 0,2
iron treatment
0,04 0,02 0,00 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 volume weighted mean
total volume
hypolimnion
epilimnion
Fig. 7.3 Volume-weighted concentrations of total phosphorus and phosphate phosphorus in Lake Gross-Glienicker during the years 1989–2000
94
K.-D. Wolter
7.4.4 Restoration of the Lake The first step was to divert the nutrient loading away from the lake; however, the lake showed no response to the removal of this external loading. Given these circumstances, restoration measures had to be undertaken in the lake itself. The most suitable measure seemed to be the precipitation of phosphorus from the water – by treatment with iron. In lakes like Lake Gross-Glienicker, with high amounts of easily-degradable organic substances in the sediment, a release of iron-bound phosphorus from the bottom can occur. To increase the mineralisation of such a reduced sediment as this – and to avoid the reduction of the iron compounds in the sediment – oxidising agents, in the form of nitrate or of oxygen by hypolimnetic aeration, need to be delivered to the sediment surface. 7.4.4.1 Hypolimnetic Aeration During the pollution period while the lake was heavily loaded with nutrients, large amounts of algae would be produced – which died and settled on the bottom of the lake. During the summers, the hypolimnion would be anaerobic and no oxidative degradation of organic matter would take place; thus much easily-degradable organic substances accumulated in the sediment. Even after restoration, in the mid-1990s, this sediment exhibited a very high capacity for oxygen consumption, which also influenced the water body. This oxygen consumption could be controlled by hypolimnetic aerators. The first step in the restoration of the lake itself was the installation of four hypolimnetic aerators. The main function of the aerators was to increase the concentration of oxygen in the hypolimnion. In addition, the aerators act as equipment controlling the water flow at the sediment–water interface (Verner 1994; Ripl and Wolter 2005). Thus oxidising agents are brought to the sediment surface; these agents can be mainly oxygen or naturally-occurring nitrate. Aerators are therefore controlling and enhancing the supplies of these oxidising agents and thus increasing the binding of phosphorus, especially to trivalent iron. 7.4.4.2 Iron Treatment As a second step, iron in the form of solid iron hydroxy-oxide and dissolved iron chloride was added to the lake water, in order to enhance the phosphorus-binding capacity even further (see Section 7.2.2). Planned to be carried out during the autumn circulation, the iron treatment, due to several delays, was carried out from December 1992 until February 1993. In both cases (iron hydroxy-oxide and iron chloride), the area to be treated by adding iron was confined to the deeper parts of the lake; iron deposited in the lake’s shallow areas (erosion zone) would be transported to deeper parts during phases of turbulent circulation. In Lake Gross-Glienicker, sedimentation is limited to depths
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
95
Lake Gross-Glienicker main wind direction transects of iron treatment
water depth (land) 0-1.5 m 1.5-2 m 2-3 m 3-4 m 4-5 m 5-6 m 6-7 m 7-8 m 8-9 m 9-10 m 10-10.5 m >10.5 m 0
300 m
Fig. 7.4 Map of Lake Gross-Glienicker depth. Arrow in the lake area shows transects during iron treatment
more than 3 or 4 m; thus, only areas more than 4 m deep were treated. Both preparations were distributed more or less equally over this area. The transects taken by the boat on the lake are shown in Fig. 7.4. Application of Iron Hydroxy-oxide Iron is common in groundwater. In waterworks that use groundwater for their drinking-water supply, iron hydroxy-oxide is formed during water conditioning. This material can be added to lakes in order to increase their phosphorus-binding capacity. However, it has to be tested for phosphorus content prior to its application as it may contain precipitated phosphorus. The molar quotient Fe/P of the iron hydroxy-oxide should be 20, but if necessary no lower than 10. The iron hydroxyoxide is a wet, fine-grained material which can be suspended in water and pumped
96
K.-D. Wolter
to the lake. It forms an iron buffer inside the sediment. In Lake Gross-Glienicker, 250 g Fe m−2 of lake area were used.
Application of Iron Chloride Iron chloride is a very acid solution and has to be handled with care. In Lake GrossGlienicker, a further 250 g Fe m−2 of lake area were applied in the form of iron chloride. The pH of the 40% iron chloride solution is below one. After dilution with water, the Fe3+ ions are precipitated as oxides and hydroxides; this process releases additional protons by the hydrolysis of water, and these protons have to be buffered by alkalinity. In Lake Gross-Glienicker, the addition of 112 mg l−1 FeCl3 led to a drop of alkalinity from 1.96 to 0.36 mM (Fig. 7.5). The pH of the water during the treatment with iron chloride reached a minimum of 6.26 (Fig. 7.5).
Effect of the Littoral Zone In lakes with a long retention time, the littoral zone may have an effect on the nutrient flow in the lake. Nutrients leaking from the untight septic tanks could enter the lake through the littoral zone. Phosphorus might also be released from the littoral 10
pH
9 8 7 6
alkalinity mM
3.0
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 iron treatment
2.5 2.0 1.5 1.0 0.5 0.0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 volume weighted mean
total
hypolimnion
epilimnion
Fig. 7.5 Volume-weighted means of pH and alkalinity in Lake Gross-Glienicker during the years 1989–2000
Secchi depth m
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
97
0 1 2 3 4 5 6 7
Chl-a L mg/m3
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 iron treatment
300 250 200 150 100 50 20 10 0
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Fig. 7.6 Water transparency and chlorophyll a concentrations in Lake Gross-Glienicker during the years 1989–2000
zone when people were bathing and stirring the sand with their feet. In one sample, a phosphorus concentration of 0.5 mg P l−1 was found in the interstitial water of the sand only 35 cm below the surface. To prevent the nutrient input from the stirring of littoral sands by bathers, some measures, schematically drawn in Fig. 7.7, were recommended: • Artificial ‘islands’ should be built to offer bathing facilities without disturbing the sand. • A trail should be built near the shore line to allow the observation of littoral biotopes without damaging the sensitive macrophytes.
7.4.5 Project Evaluation Ten years after restoration, the treatment of Lake Gross-Glienicker can be regarded as being successful. After the iron-treatment, the hydrological (water level, retention time), chemical and biological parameters were monitored. Initially, the water level decreased temporarily due to reduced inflow and low precipitation. The theoretical water retention time, as estimated by the successive decrease of chloride concentrations (that had been enriched by the FeCl3-treatment), increased to about 20 years. During the summer vegetation periods, low concentrations of phosphorus (about 30–40 µg l−1 P-tot) and chlorophyll a (that had decreased from a mean value
98 Fig. 7.7 Proposal for artificial islands and ‘observation’ trail near the shore in Lake GrossGlienicker
K.-D. Wolter artificial bathing isles possible trail
0
300 m
of 49 mg m−3 to a mean value of 6.5 mg m−3) were maintained (Figs. 7.3 and 7.6). The oligotrophication of the lake, together with the continued hypolimnetic aeration, has led to continuous aerobic conditions in deep water and at the sedimentwater boundary layer (Wolter 2007). In the littoral, the successive development of submerged macrophytes could be observed. First, isolated occurrences were found in 1996; then, between 1996 and 2000, these isolated stands of macrophytes spread to 9% of the whole lake area or 30% of the littoral zone with less than 5 m water depth. Above all, Myriophyllum spicatum, Elodea canadensis, Potamogeton berchtoldii and Potamogeton pectinatus were found (Körner 2000). In the plankton community, there occurred a general decrease in phytobiovolumes and a shift of composition after iron treatment. The decrease of phosphorus led to a better coupling of phytoplankton production to growth of zooplankton. Compared to the previous situation, the ratio of zooplankton to phytoplankton increased, indicating a higher efficiency in the food chain; and thereby a reduction in the losses to sedimentation. There was also an improvement in the fish community. The tendency towards the deformation of underfed whiting decreased; and in relation to predatory fishes, whiting diminished in number. In the 10 years since the iron treatment, enormous phosphorus releases from the sediment (internal loading with noticeable increase of surface phosphorus concentration) have not been observed. Nevertheless, since 2001 local minor increases of deep water phosphorus have occurred. In deeper sediments (5–10 cm), part of the iron (about 50–75%) was used up by desulphurication and the subsequent formation of iron-sulphide. On the whole, the increase of iron-content at the sediment surface has led to a tight phosphorus binding. The aim of the restoration of Lake Gross-Glienicker, situated in the denselypopulated area of Berlin, was to use the lake for recreation, bathing and fishing.
7 Restoration of Eutrophic Lakes by Phosphorus Precipitation
99
The basic prerequisite for the redevelopment of the lake was the diversion of sewage. The sewage from the military barracks polluting the lake was diverted prior to restoration measures being undertaken in the lake itself. In subsequent years, leaking septic tanks from settlements in the lake surroundings have been replaced by sewage pipes connected to a centralised sewage plant. In the case of nutrient leaching from soils, the seepage of groundwater into the lake should be kept to a minimum. This can best be achieved by decreasing the rate of groundwater recharge by increasing evapotranspiration with the help of an increased vegetation cover. Raising the water level in the lake as high as possible would also ensure that seepage of water from the lake to the soil increases, thereby minimising the nutrient loading from the soil to the lake. To summarise, two principles had to be respected in order to restore the lake and its catchment area: • Keep water in the soil – thereby achieving maximum evapotranspiration and maintaining short-circuited water cycle (see Chapter 1; and Ripl and Wolter 2002). • Keep minerals and nutrients within particulate or solid structures, i.e. in the soil, terrestrial vegetation, emersed and submersed macrophytes and sediments.
References Cooke GD, Welch EB, Peterson SA, Newroth PR (1986) Lake and reservoir restoration. Butterworths, Boston, 392 pp Körner S (2000) Aktuelle Besiedlung des Groß-Glienicker Sees mit submersen Makrophyten. Auftraggeber: Senatsverwaltung für Stadtentwicklung. Institut für Gewässerökologie und Binnenfischerei (IGB), Berlin, Manuscript, 12 pp. + annex Ripl W, Wolter K-D (2002) Ecosystem function and degradation. In: Williams PJ le B, Thomas DR, Reynolds CS (eds) Phytoplankton productivity. Carbon assimilation in marine and freshwater ecosystems. Blackwell, Oxford, pp 291–317 Ripl W, Wolter K-D (2005) The assault on the quality and value of lakes. In: O’Sullivan PE, Reynolds CS (eds) The lakes handbook, vol 2, Part I – general issues, chapter 2. Blackwell, Oxford, pp 25–61 Ripl W, Heller S, Koppelmeyer B, Markwitz M, Wolter K-D (1993) Limnologische Begleitstudie zur Entlastung des Tegeler Sees. Endbericht. Im Auftrag der Senatsverwaltung für Stadtentwicklung und Umweltschutz. Technische Universität Berlin und Gesellschaft für Gewässerbewirtschaftung, Berlin, 51 pp. + Anhang, Manuskript Verner B (1994) Aeration of the hypolimnion as a tool for restoring eutrophic lakes. In: Eiseltová M (ed) Restoration of lake ecosystems – a holistic approach, vol 32. International Waterfowl and Wetlands Research Bureau, IWRB Publication, Slimbridge, pp 119–129 Wolter K-D (2007) Nachhaltige Entwicklung – Prozessbetrachtung und Konsequenzen für die Ökosystemsteuerung am Beispiel der Seerestaurierung. Habilitationsschrift. Technische Universität Berlin, Fakultät VI – Planen Bauen Umwelt. Manuskript, 286 pp
Chapter 8
Restoration of Lakes Through Sediment Removal, with Case Studies from Lakes Trummen, Sweden and Vajgar, Czech Republic Sven Björk, Jan Pokorný, and Václav Hauser
Abstract For removal of nutrient-rich sediment from polluted and irreversibly damaged lakes, suction dredgers have been constructed to meet the demands formulated by limnologists. This limnological-technological cooperation has aimed at the precision dredging of sediment strata responsible for the internal nutrient loading of lake ecosystems. Suction dredging is mainly restricted to small lakes of high environmental value. Prior to in-lake measures, the external loading of an ecosystem has to be brought under control. Whenever possible pumped sediment ought to be used as fertiliser. The heavily polluted Lake Trummen, dredged in 1970–1971, was turned into an environmental asset. The cyanobacterial blooms (summer transparency 15–20 cm) disappeared and were replaced by a plankton community rich in species. The ecosystem as a whole recovered to a functional unit, characterised by a balance between production and decomposition. In the urbanised area of Växjö, Lake Trummen became available for a variety of recreation activities. The ecosystem conditions are continuously monitored. As in Lake Trummen, Lake Vajgar was in a degraded condition characterised by heavy blooms of cyanobacteria caused by intensive internal phosphorus loading. After unsatisfactory experiments with a suction dredger available on the market, an automatically-controlled, precision dredger was designed, constructed and used to remove the defined top sediment layer in 1991–1992. The Vajgar ecosystem immediately changed from being a source to functioning as a trap for phosphorus. However, because the external loading from the catchment area remained high, Lake Vajgar again gradually developed conditions favourable for cyanobacteria.
S. Björk (*) Department of Ecology, Limnology, University of Lund, SE-223 62 Lund, Sweden e-mail:
[email protected] J. Pokorný ENKI p.b.c., Dukelská 145, 379 01 Třeboň, Czech Republic V. Hauser ENVI Ltd., Dukelská 145, 379 01 Třeboň, Czech Republic M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_8, © Springer Science+Business Media B.V. 2010
101
102
S. Björk et al.
Keywords Biodiversity • External and internal nutrient loading • Limnological control programme • Sediment • Suction dredging
8.1 Introduction The removal of nutrient-rich sediment deposited in lakes must be considered the most radical and definite method to restore eutrophicated lakes. For a number of reasons this method should be practised only in shallow water bodies suffering from high sediment deposition. In waters deep enough to preserve their lake character with open water, other methods to reduce phosphorus release from the sediment should be applied (e.g. in-lake treatment of sediment, aeration). Compared with sediment removal such methods are technically less complicated, as well as being fast and cheaper. However, in shallow lakes in which the ageing processes are taking place at a high rate, sediment has to be removed in order to re-create sustainable, balanced systems and regain a sufficient water depth to prevent shallow lakes with high deposition rates from terrestrialisation.
8.2 Removal of Sediment by Suction Dredging 8.2.1 Technical Requirements The removal of defined layers of top sediment, free of the roots of growing macrophytic vegetation, requires suction-dredgers to be used. As most water bodies selected for restoration through suction dredging are small, dredgers should be small, lightweight and easily transportable on a lorry from one lake to another. During dredging operations, there should be no turbidity caused by nozzles, dredgers and pumping – a sign of stirring up and moving sediment, indicating undesirable fertilisation of the lake by nutrient-rich interstitial water. In some of the early projects, commercially-available suction-dredgers which were not constructed for this specific purpose were used. Lake Trummen was the first demonstration project on lake restoration through sediment removal, for which a special nozzle was developed, whereby it was possible to avoid turbidity. Experience gained during the 1960s and 1970s from sediment suction-dredging projects made it clear to limnologists that the dredging technique must meet the following three demands: • No turbidity should be created. • Water admixture to the sediment should be no more than that required to allow it to be pumped. As a rule, the nutrient-rich top sediment can be pumped without any addition of water. • The proportions of sediment and water should be constant when homogenous sediment is pumped.
8 Restoration of Lakes Through Sediment Removal
103
8.2.2 Horizontal Control of Dredgers So far, two methods for controlling the horizontal movements of dredgers have been applied, by means of: • Two hydraulically operated spuds (stabilisers) at the rear of the float • Wire arrangements In the first case, either the nozzle moves along the circumference of a semicircle as the float is fixed to the bottom by alternating starboard and port spud or both spuds are fixed to the bottom and the nozzle is mounted on a moveable arm at the stem of the float. In the second case, the moving pattern is dependent on the arrangement of the wires from the float to the points of attachment on the shore or to anchors. In both cases, rows of buoys have been needed to mark treated areas.
8.2.3 Pumping and Deposition of Sediments The conditions for the deposition of removed sediment on land vary from case to case depending on dredging technology and the space available at the specific water body selected for restoration. If it is necessary to deposit the pumped sediment in settling ponds, the run-off water, being a mixture of lake and nutrient-rich interstitial
Fig. 8.1 Schematic illustration of lake restoration by means of suction dredging (From Björk 1972)
104
S. Björk et al.
water, has to be treated as wastewater in order to avoid pollution downstream of the settling ponds (Fig. 8.1). As degraded lakes that need dredging are often located in urbanised regions, it is sometimes necessary to pump the sediment long distances to suitable deposition sites. This is made possible by means of the installation of booster pumps to increase the pressure in the pipeline, in order to overcome the distances and increases in elevation. To protect the booster pumps from mechanical damage, a grid for the separation of stones ought to be installed between the nozzle pump and the first booster pump. One of the aims for a new, suction dredging technology designed to meet the three limnological demands described above, is to make it possible to deposit the sediment directly onto fields. Thus the need to construct settling ponds and treat the run-off water is avoided. The new equipment is characterised by being automatically controlled. This is based on a continuous analysis of the pumped medium. Detailed maps showing the stratigraphy and horizontal distribution of the type of sediment that should be removed are essential for the operation.
8.2.4 Automatically Controlled Suction Dredging Technology In principle, an automatically controlled suction-dredger consists of three parts: the float, the nozzle, and the automatic control system. 1. The float is designed in the simplest way. Its size is determined by local conditions (size of the lake, waves, transport conditions, etc.). It is imperative that the size of the dredger corresponds to the size of the actual water body in order to optimise transport and installation costs. 2. The nozzles developed for soft organic sediment include pump and grid for protection against stones. (Primarily, the sediment [gyttja] to be removed does not include stones. However, stones, ammunition, bicycles, etc., have often been thrown into lakes, especially in urbanised regions). For other types of sediment, a special design and function of the nozzle is needed to fulfil the limnological demands. 3. The control system consists of measuring devices from which signals are continuously transmitted to the unit directing the position of the nozzle that pumps a predetermined concentration of sediment, the speed of movement of the dredger, its location on the lake, etc. The measuring devices of the automaticallycontrolled sediment pumping system include reading of water depth, density-, flow- and other meters for recording physical and chemical parameters of pumped sediment and sediment/water mixtures. The type of precision dredging most often desired for lake restoration is the removal of sludge, layer after layer, by pumping sediment according to a programmed, constant, dry matter concentration. A serious drawback concerning old types of suction-dredgers was the pumping of large volumes of ‘unnecessary’ water. This made the construction of big settling ponds inevitable. Through automatically controlled
8 Restoration of Lakes Through Sediment Removal
105
pumping, ‘unnecessary’ water is avoided as the dredger can be programmed to pump sediment with the smallest possible admixture of water. This is accomplished through the continuous control of the nozzle’s position and movement (horizontally and vertically). In summary, the highly flexible, computerised, automatic control system makes it easier to find deposition sites for the sediment. The reasons for this are: • The amount of water can be minimised or optimised with respect to transport costs, sediment mixtures for special purposes, deposition conditions, etc. • In order to compose a soil mixture appropriate for agricultural field crops, horticultural mixtures, etc., fertilising elements, organic matter, and mineral particles of chosen fraction can be continuously added to the pumped sediment before its deposition. The pH can be adjusted and substances for the binding of metals, for example, can also be added. Provided the pumped sediment has proved to be suitable as a medium for fertilising and soil conditioning, direct deposition on arable land is a cheap and ecologically sound method of returning matter from eutrophic water bodies back to the catchment area. However, removal of contaminated sediment needs special caution and methods should be designed specifically for each project. When contractors offer dredging operations for the restoration of lakes, the high pumping capacity of conventional dredgers is often emphasised as a competitive argument. However, high capacity is not at all the decisive qualification. On the contrary, it is often a disadvantage for this type of suction dredging. Instead, precision in the removal of an exactly defined sediment layer is decisive. In this connection, high capacity dredgers are problematical because of the lack of correspondence they show between pumping capacity and the bottom area that has to be cleaned per unit time, in order to supply the dredger with the volume protection of sediment for which it has been designed. Because the dredger cannot possibly move at the speed required in order to collect enough sediment from a layer 20–50 cm thick, the action results in the pumping of either excess water or sediment which should remain in the lake. The limnological aim of a restoration project by sediment removal from a eutrophicated lake is to carefully clean a defined area within which the polluted sediment causes the troublesome internal nutrient loading. It is the area of removed sediment that should be checked after the project has been finished. The lake-bottom area that has been dredged is much more important to asses the quality of a completed job than the volume of removed sediment. It is undesirable that the underlying sediment harmless for the lake ecosystem or, through its adsorptive capability, even beneficial, would have been removed, thereby unnecessarily increasing the total volume of pumped sediment. The high precision required to clean the bottom areas according to the limnological restoration plan, demands positioning equipment with a vertical accuracy of the order of centimetres. The required precision for the lateral (horizontal) movements of dredger and nozzle is obtained by means of laser technology. Cleaned areas should be automatically recorded in detail during the dredging operation.
106
S. Björk et al.
8.3 Removal of Sediment Overgrown by Macrophytes Eutrophicated shallow lakes with heavy water blooms are typically devoid of submersed plants but often characterised by luxuriant stands of floating-leaved and emergent littoral vegetation. The presence of living roots in the sediment causes great problems and often makes suction-dredging impossible. Before dredging a polluted lake overgrown with floating-leaved plants (Nuphar, Nymphaea, Potamogeton, etc.), their petioles and stems should be cut short above the lake bottom. After that, their rhizomes and roots have to be cut by a rotovator mounted on pontoons (see Fig. 9.5). In soft bottoms, it is possible to loosen the rhizomes from the bottom by the use of cutting bars (one horizontal and one vertical), also mounted on pontoons. The loosened material, drifting at the water surface, is transported to wind-exposed shore sections. It is sometimes necessary to carry out this type of treatment in sub-areas surrounded by floating booms. After finishing the work within the sub-area, all loosened material is dragged, with the boom encircling it, to the shore where the material is removed from the lake using a conveyor belt. The pre-treatment of the lake bottom, to get rid of rhizomes and roots, in order to make suction-dredging possible, ought to be carried out at least 1 year before the dredging starts. Root material remaining in the bottom has then decayed to such an extent that it is no longer a problem and the negative ecological effects (turbidity), caused by the treatment of the bottom, are remedied by the subsequent dredging. Plaur formations developed in the lakeward vegetation zone should best be cut in portions suitable for being towed to the shore and removed by an excavator. Draglines, amphibious excavators, etc., are used for cutting the plaur into manageable pieces. Firmly rooted vegetation along shores should, if necessary, be removed by draglines and excavators. All such work should also be carried out in due time, for ecological reasons, before the final suction-dredging takes place. Excavated bottom material can also be loaded into a pulper/macerater, transferring it to hydraulic plunger pumps having the capacity to transport it (at a pressure of up to 100 bar) over several kilometres, even with differences in height of tens of metres.
8.4 Case Studies 8.4.1 Lake Trummen, Sweden 8.4.1.1 Background In Sweden, thanks to the intensive construction of sewage-treatment plants with phosphorus precipitation, the conditions in hundreds of eutrophicated lakes have improved. Nevertheless, some lakes had been so severely affected that they did not recover.
8 Restoration of Lakes Through Sediment Removal
107
10 15 Skirviken 20 −21 10
21
13
10 13 20 21 15 10
Typha Carex latifolia rostrata Equisetum fluviatile Mire vegetation Nuphar luteum Phragmites Shrub vegetation communis Schoenoplectus lacustris
0
100
200 300m
Fig. 8.2 Lake Trummen. Map of vegetation and water depth (in m) (From Persson 1969 and Digerfeldt 1972)
8.4.1.2 General Description and Type of Disturbance of Lake Trummen Lake Trummen (56º52¢N, 14º50¢E) at the town of Växjö in south-central Sweden was a good example of the large group of polluted lakes which had deteriorated within and around urbanised areas. The originally oligotrophic, brown-water Lake Trummen (drainage area 12 km2, lake area 1 km2 and depth, until 1970, 2 m – Fig. 8.2) had been used for swimming and for water supply, at least to some degree, until the 1920s. Up to 1957–1958, it was exploited as a recipient of sewage from the town of Växjö and, from 1941 to 1957, of wastewater from a flax factory. An extensive survey concerning the human impact on the lake during the nineteenth and twentieth centuries has been compiled by Lettevall (1969, 1977). The pollution became more and more severe and in the 1940s the hypertrophic ecosystem collapsed. After that there were regular fish-kills due to total oxygen depletion in winter and the summer transparency was depressed to 15–20 cm by cyanobacteria (‘blue-green algae’), especially Microcystis. Despite that the sewage and industrial wastewater discharge to the lake was stopped in 1957–1958, the lake did not recover – a very interesting but not surprising fact to limnologists and a severe problem for the town authorities. About 10 years after the diversion of wastewater from the lake, the summer transparency was still only 15–20 cm. At this point, the town authorities definitely had to face the Trummen problem when planning the urbanisation of the surroundings. Of course, the simple, narrow-minded technological solution of filling the lake was considered, but the authorities decided to follow a proposal to restore the lake. Later, the restoration project developed a very successful cooperation between limnologists, palaeolimnologists and other ecologists, politicians, technologists and administrators. The scientific pre-project investigations carried out during 1968–1970 were designed as a typical ecosystem-oriented study.
108
S. Björk et al.
Fig. 8.3 Lake Trummen. Section through a Typha-Equisetum plaur area (From Björk & Digerfeldt 1965)
8.4.1.3 Pre-project Investigations Palaeolimnological investigations have revealed that in the intact, oligotrophic Lake Trummen, the sediment growth rate was between 0.2 and 0.4 mm per year, but due to the very high nutrient supply during the recipient period it increased to 8 mm (Digerfeldt 1972). In the 1960s the sediment of Lake Trummen showed a very characteristic stratification – a brown, well-consolidated gyttja overlaid by the 20–50 cm of black, loose deposits from the recipient period (Björk and Digerfeldt 1965). The irreversibility of the damage to Lake Trummen was due to the presence of that layer from which nutrients were being released (Bengtsson and Fleischer 1971; Bengtsson et al. 1975) and to which the plankton crop was being deposited after every vegetation period. Broad reedbeds, developed as a floating plaur zone, surrounded the lake (Fig. 8.3). 8.4.1.4 Restoration In 1970, the topmost half metre of sediment was removed by suction dredging, and in 1971, another half metre was dredged. Altogether about 400,000 m3 of sediment (gyttja) were removed during nine working summer months. Furthermore, at least 200,000 m3 of lake water were mixed with the pumped sediment. The limnologists requested a nozzle for dredging which would make it possible to suck in the sediment without making the lake water turbid, and with a minimum intake of lake water. The sediment was pumped to simple, settling ponds constructed in an abandoned farming area from which the topsoil had first been removed. The run-off water from the settling ponds – a mixture of lake and interstitial water – was treated with aluminium sulphate in a simple plant for precipitation of phosphate and suspended matter. Before restoration, the total phosphorus content of the
8 Restoration of Lakes Through Sediment Removal
109
Fig. 8.4 Lake Trummen. Drying sediment in the settling pond (Photo: Sven Björk 1970)
lake water was about 600 µg l−1. The phosphorus concentration of the water from the settling ponds was of the order of milligrams per litre before the treatment. After the precipitation, the total phosphorus content of the run-off water was about 30 µg l−1. The pre-project investigations also comprised the growth test with the sediment to prove its suitability for agricultural and horticultural purposes. The pumped sediment was, after drying, sold by the town to be used in parks, gardens, roadsides, etc. (Fig. 8.4). The income financed the preparation of green areas around the lake. A separate overgrown bay was left as a waterfowl reserve (an observation tower was constructed on the shore), and an artificial island was built for the birds. 8.4.1.5 Overall Project Evaluation As was foreseen, the changes in the lake itself were dramatic (Cronberg 1982). The phosphorus and nitrogen concentrations decreased and the transparency increased (Fig. 8.5). The cyanobacterial blooms disappeared, and a plankton community with a higher diversity replaced the monocultures of Microcystis. The population of the freshwater mussel Anodonta that was wiped out in the hypertrophic lake, recolonised the bottom immediately after dredging. The reappearance of the mussels means that the ecosystem became enriched with their function of an efficient seston filtrator. The restoration caused such a change in the structure of the ecosystem that the lake recovered to a functioning unit, characterised by a balance between its production and decomposition. The lake became suitable for sport fishery, swimming, windsurfing and other forms of recreation within the urbanised area. Only nine summer months were
110
S. Björk et al.
Fig. 8.5 Lake Trummen. Concentrations of phosphorus and nitrogen, total biomass of phytoplankton and total biomass of blue-green alga, 1968–1978 (From Cronberg 1982)
needed to transform the lake from an environmental problem to an environmental asset. The total cost was about US$500,000 (in 1971). The Lake Trummen restoration project was used for the training of professional limnologists and other ecologists (about 20 have been involved). It also served as a demonstration project for converting politicians and administrators to believers in ecology and nature conservation offering sustainable, high quality environmental conditions. In Lake Trummen, the nutrient-rich sediments had been deposited over more or less the whole lake area (with the exception of the bay preserved as a waterfowl reserve). Therefore, it was necessary to dredge the extensive part of the lake. However, in other lakes, sediments of a sewage-sludge character could be concentrated to a restricted area, and the restoration effect could be reached after dredging only part of the lake bottom.
8 Restoration of Lakes Through Sediment Removal
111
8.4.1.6 Management and Monitoring Programmes Following the highly significant nutrient reduction through the sediment removal, intensive fishing was carried out (during the period of 1976–1979) to reduce the populations of bream (Abramis brama) and roach (Rutilus rutilus) (16 t, i.e. 190 kg ha−1 of fish were removed), while predators, such as pike (Esox lucius) and perch (Perca fluviatilis), were left in the lake (Andersson et al. 1978; Andersson 1985). There is hardly any doubt that the synchronous decrease in total phosphorus and nitrogen concentrations as well as in phytoplankton biomass was caused by this intensive selective fishing. It has, however, not been possible to obtain stability in these respects (Fig. 8.6). Reduction of planktivorous fish took place also in 1994, 1996–1997 and 2000. In the case of Lake Trummen there is a considerable exchange of fish between the lake and its tributaries and outflow including the lakes located upstream and downstream. The occasional development of a food resource, such as large planktonic animals, is utilised by both invading and in-lake reproduced fish. In a eutrophicated lake, after reduction of, for example, bream and roach, the eventual appearance of a valuable food resource (cladocerans, etc.) will not be left untouched and unexploited. It again opens up the possibilities for the prosperous development of consumers (fish) that will reproduce successfully until their biomass becomes the food resource of a higher consumer or is reduced through laborious fishing. The oscillations might be great in disturbed systems, but the base of the trophic pyramid will always remain broad and the top narrow. If a snapshot of biomass distribution in the open water should happen to show upside-down relations, these are of limited duration only. It should be stressed, that Lake Trummen was restored by suction dredging after sewage and industrial wastewater had been diverted from the lake. The external loading could, therefore, be considered as fairly normalised with respect to regional limnological conditions. However, since the restoration, the surroundings have been urbanised, i.e. the character of the catchment area has changed considerably. In the limnological management plan for the restored Lake Trummen, it was suggested to make arrangements for continuous and careful control of the water quality in the tributaries. Among other things, construction of basins for the reduction of nutrients, and the protection against oil spill from industrialised sections of the catchment were recommended. The suggested arrangements have in course of time largely been realised. Changes in the lake ecosystem following the restoration were studied by a research programme covering the period 1968–1980. Since then, the lake development is followed as part of the routine environmental control programme executed by the local authorities. Ten to 15 years after the restoration of Lake Trummen, there appeared indications of increased plankton turbidity caused by tiny cyanobacteria (among these Cyanodictyon imperfectum) decreasing the transparency. In order to revert to the restored stage, it is necessary to reduce the external loading from the recently urbanised part of the catchment area, including nutrient trapping at the mouths of the tributaries. The diagram in Fig. 8.7 illustrates the striking differences in phytoplankton biomass before and after restoration as well as between-year oscillations during the period 1968–2003.
S. Björk et al.
Total nitrogen, mg N. 1−1
Total phosphorus, mg P. 1−1
112
0.8 0.6 0.4 0.2 0 8 6 4 2 0 80
Total
40 20 0 80
Blue-green algae
Phytoplankton biomass, mg. 1−1
60
60 40 20 0 1970
1975
1980
1985
Fig. 8.6 Lake Trummen. Results of nutrient reduction and intensive removal of fish on phosphorus and nitrogen concentrations and phytoplankton development (From Björk 1988)
8 Restoration of Lakes Through Sediment Removal
113
120
Biomass [mg 1−1]
100
80
60
40
20
0
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
95 96 97 98 99 00 01 02 03
Fig. 8.7 Lake Trummen. Monthly values of phytoplankton biomass. The turn of the months June–July indicated. The restoration was performed in 1970–1971 (Data from Gertrud Cronberg)
8.4.2 Vajgar Lake, Czech Republic 8.4.2.1 Background The Vajgar Lake is an artificial water reservoir built with the aim of fish production, therefore locally called a fish pond. The fish ponds in central and eastern Europe, built mainly in the Middle Ages, are shallow water bodies ranging in size from less than a hectare to several hundred hectares. Situated mostly on sites of former marsh, swamp, bog or fen, or in original floodplains of small watercourses, the fish ponds and their littoral zones have become a specific type of wetlands (Dykyjová and Květ 1978) and may look like natural lakes. In a country with very few natural lakes, these fish ponds are an important part of the hydrological system, serve as water purification systems and sediment traps, provide habitats for many aquatic and wetland plant and animal species, serve for recreation, but above all, are dedicated to fish production. The original oligo- or mesotrophic character of the fish ponds has changed to eutrophic or even hypertrophic conditions during the second half of the twentieth century. Radical intensification of fish production, based on direct fertilisation and liming of fish ponds along with higher fish-stock, has increased the production of fish nearly tenfold since the 1930s. The recent intensive management of fish ponds, combined with the high external loading of nutrients from agricultural fields or wastewater and storm water discharge from urban areas, has resulted in the massive development of phytoplankton and cyanobacterial water blooms, along with great fluctuations in oxygen and pH.
114
S. Björk et al.
8.4.2.2 Description of Vajgar Lake/Fish Pond and Its Catchment Vajgar fish pond (c. 40 ha) is situated in the district town of Jindřichův Hradec (23,000 inhabitants, 49°09¢N, 15°01¢, altitude 476 m asl). The fish pond formed part of the town’s medieval fortification, the first written documentation concerning Vajgar fish pond coming from the thirteenth century. The fish pond originally provided water and fish, served as a recreational area, and, from 1795, it also became a recipient of the town’s wastewater (Teplý, 1927). Up until 1970, wastewater (biological oxygen demand BOD 200–250 mg l−1, total phosphorus Ptot c. 5 mg l−1, total nitrogen Ntot 50 mg l−1) was still discharged into Vajgar fish pond (about 15 l s−1). In 1970, a wastewater treatment plant was built, treating about 10,900 m3 per day. In 1990s, 12 septic tanks situated under the level of the wastewater pipe system, were finally connected to the system by means of pumps. Unfortunately, four overflow storm water drains still discharge into the fish pond. Up to 1965, the fish pond was regularly emptied and the fish harvested. Then, due to some damage to the main outlet mechanism, and, later on, also due to a small dam constructed between the so-called ‘Small’ and ‘Large’ Vajgar, the pond could no longer be drained. Vajgar fish pond itself serves as a sedimentation basin of Hamerský stream (on which several fish ponds have been constructed) and its catchment of 220 km2. 8.4.2.3 Pre-project Investigations In the summer of 1985, a mass development of Stephanodiscus hantzschi, an indicator of eutrophic waters, was observed in Vajgar fish pond as well as in several other fish ponds on the Hamerský stream. Dense cyanobacterial blooms occurred regularly in summers of late 1980s and caused environmental problems – the pond could not be used for recreation anymore and the smell of degrading blue-greens was annoying the town inhabitants. Cyanobacteria of the genus Anabaena and Aphanizomenon flosaquae prevailed during summers in late 1980s, whilst dense cyanobacterial blooms dominated by Microcystis sp. occurred mostly in the summers of 1990 and 1991. The chlorophyll a concentration reached almost 350 µg l−1 in May 1991. The seasonal course of total phosphorus budget estimated from total phosphorus concentrations of Vajgar fish pond (Fig. 8.8) in the inflow and outflow water, has shown that the fish pond served as a phosphorus source in late spring. At that time, the sediment begins to warm up, organic matter starts to decompose removing oxygen from the bottom and, thus, allowing the release of phosphorus from the sediment. This observation confirmed that in-lake restoration measures have to be taken should the nutrient loading be reduced. Sediment cores were taken in order to measure the thickness of black (nutrientrich) sediment which should be removed from the pond. Map of the black sediment thickness were made for the whole pond using data obtained on transects 100 m apart and at 50 m distances along each transect (the thickness of nutrient-rich sediment was upto 60 cm). In the centre of the pond, the deepest possible core was
8 Restoration of Lakes Through Sediment Removal
115 inflow outflow
P tot [mg I-1]
400 300 200 100 0
Jan
Apr
May
Aug month
Oct
Dec
Fig. 8.8 Seasonal course of phosphorus concentrations in inflow and outflow of Vajgar fish pond in 1991
taken in order to try to reach the original bottom of the valley basin. This approximately 2.4 m deep core was divided into 42 five-centimetre segments and the following analysis made for each segment: dry matter, organic matter, Al, As, B, Bi, Ca, Cd, Co, Cu, Cr, Fe, Hg, Mg, Mn, Mo, Na, Ni, P, Pb, S, Se Sr, Ti, V, W, Zn and pollen. Dry matter content of the sediment throughout the vertical profile (Fig. 8.9) shows a range of 10% at the surface, 25% at about 1 m, and almost 80% at 2 m of sediment depth. Organic matter content of sediment (Fig. 8.10) is highest at the surface where it reaches about 20% and then decreases in deeper layers. Vertical profiles of selected heavy metals and nutrient concentrations in fresh and dry weight of sediment are shown in Fig. 8.11. The concentrations of heavy metals do not exceed the limits for composts (as stipulated by the Czech Ministry of Environment). The release of phosphorus from the upper black layer of sediment and the lower grey layer were tested in laboratory under aerobic and anaerobic conditions (Fig. 8.12). The highest release of phosphorus is clearly from black, upper layer, of sediment under anaerobic conditions. The dating of the whole profile was made by radiocarbon 14C and the dating of upper layers by caesium isotopes 134Cs and 137Cs. The dating by 14C showed that the original bottom was flooded about 1,000 years ago. The age of the layers at 170 and 100 cm depth was 1292 ± 60 and 1322 ± 60, respectively. The rate of oxygen consumed by the different layers was also studied and could be correlated to organic matter content. The pH, alkalinity and phosphate in the interstitial water were measured and microbiological tests were undertaken. The analyses showed that the sediment could be safely applied to fields for agricultural use. 8.4.2.4 Restoration and Project Evaluation As the sediment accumulated in the pond served as an uncontrolled source of nutrients, it was decided to remove the top, nutrient-releasing, layer of the
116
S. Björk et al. depth [cm] 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
0
10
20
30
40
50
60
70
80
90
100
dry matter [%]
Fig. 8.9 The dry matter content in the sediment profile of Vajgar fish pond in October 1990
sediment from the pond. In 1987, an old type of suction-dredger (Czechoslovakmade suction dredger SB 20) was tested. The dredger was able to pump sediment onto the bank but the sediment was very diluted (less than 4% of dry matter), the dredger made holes in the bottom which could cause problems during fish-harvesting, the maximum length of transport and capacity of the dredger was low, and so systematic dredging of the bottom surface was very difficult. Hence, there was the need for a suction dredger which would respect the demands of limnologists (as stressed earlier). The prototype of such a dredger, designed in Sweden, was constructed for the sediment removal in Vajgar. The sediment-pumping started in August 1991. By the end of 1992, about 330,000 m3 of sediment had been pumped out from the pond and transported in pipes for about 2.5 km to seven settling lagoons, each about 1 ha in area and 3–5 m deep. The sediment transported from the fish pond had a dry mass content of 10–15%. The town of Jindřichův Hradec, supported by the district office, financed the restoration works; the sediment removal and scientific work being also supported by the Ministry of Environment. The whole cost was approximately US$850,000.
8 Restoration of Lakes Through Sediment Removal
117
depth [cm] 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
0
5
10
15
20
25
30
organic matter [%]
Fig. 8.10 The organic matter content in the sediment profile of Vajgar fish pond in October 1990
Plans to use the pumped and dried sediment to fertilise agricultural land were, unfortunately, not fulfilled. During the transitional period of the early 1990s, a period when state and cooperative farms were changing into privately owned farms, the farmers were not able to pay for the preparation of manure and the state no longer paid subsidies for composting and fertilisers as it did formerly. Most of the sediment, thus, has remained deposited in the settling ponds and become overgrown by bulrush, willows and alder, despite that the valuable fertilising potential of the sediment was proved through the experimental cultivation of various vegetables on mixtures of the sediment and nutrient-poor soil.
8.4.2.5 Situation After Restoration In April 1993, Vajgar fish pond was partly drained in order to remove the ‘temporary’ dam placed between the Large and Small Vajgar and to allow the complete drainage of the Vajgar fish pond again. After the fish pond was refilled with water, it was stocked with heavier carp (Cyprinus carpio), 1–2 kg, about 250 kg ha−1.
118
S. Björk et al. 0 25
0
50
50
75
75
100
depth [cm]
depth [cm]]
25
100 125 150
150 175
175
200
200
225
225 250
125
0
1
2
3
4
5
250
6
0
20 40 60 80 100 120 140 160 180
Ca [g/kg]
Pb [mg/kg]
0
0
25
25
50
50 75
depth [cm]
depth [cm]
75 100 125 150
100 125 150
175
175
200
200
225
225
250
0
800
1600
2400
3200
250
4000
0
1
2
3
0
0
25
25
50
50
75
75
100
100
125 150
175 200
225
225 4,00
6,00
8,00 10,00 12,00
Hg [mg/kg]
6
7
8
150
200
2,00
5
125
175
250 0,00
4
S [g/kg]
depth [cm]
depth [cm]
P [mg/kg]
250
0
10
20
30
40
50
60
Cr [mg/kg]
Fig. 8.11 Vertical profile of concentrations of selected elements in fresh (+) and dry (*) weight sediment of Vajgar fish pond
8 Restoration of Lakes Through Sediment Removal
119
Ptot [µg l−1]
1000 800 600 400 200 0
0
2
4
6
8
10
12 14 time [h]
16
18
20
22
24
26
Fig. 8.12 The release of total phosphorus from the upper (black) layer (, ) and lower (grey) layer (●, +) of sediment from the Vajgar fish pond into distilled water bubbled by air, i.e. aerobic conditions (, +) and N2, i.e. anaerobic conditions (, ●)
500 inflow P tot [mg l-1]
400
outflow
300 200 100 0
Feb Mar Apr May Jun
Jul
Aug Sep Oct Nov Dec
month
Fig. 8.13 Seasonal course of total phosphorus concentrations in inflow and outflow of the Vajgar fish pond in 1993
These carp were added to the original fish stock in order to disturb and mix the rest of the black sediment. The seasonal course of phosphorus concentrations before sediment removal (in 1991, see Fig. 8.8) and after its removal (1993–1995, see Figs. 8.13–8.15) show the positive effect of the organic-rich sediment removal on phosphorus binding. Phosphorus budget was calculated, for the year 1991, from phosphorus concentrations and water discharge. For April and, in particular, for May 1991, it shows higher phosphorus concentrations as well as higher absolute amounts of phosphorus measured in the outflow than in the inflow. In May 1991, about 200 kg more total P was leaving the fish pond than was entering it. After the sediment removal (1993), less total phosphorus was
120
S. Björk et al. 500 inflow P tot [mg I-1]
400
outflow
300 200 100 0 Jan Feb Mar Apr May Jun
Jul Aug Sep Oct Nov Dec
month
Fig. 8.14 Seasonal course of total phosphorus concentrations in inflow and outflow of the Vajgar fish pond in 1994
500 inflow
P tot [mg l-1]
400
outflow
300
200 100
0 Jan Feb Mar Apr May Jun Jul month
Aug Sep Oct Nov Dec
Fig. 8.15 Seasonal course of total phosphorus concentrations in inflow and outflow of the Vajgar fish pond in 1995
leaving the Vajgar fish pond than was entering it. Nevertheless, the capacity of remaining sediment to bind phosphorus gradually decreased between 1993 and 1995. As for phytoplankton, the dynamics and structure of the community showed signs of improvement during the first year after sediment removal (Poulíčková et al. 1998). The summer peak of biomass was lowered and the species diversity increased. In 1995, however, the phytoplankton structure and dynamics returned to a similar state as that before the sediment removal. Absent after the sediment removal in 1993 and 1994, an extensive population of Microcystis was found
8 Restoration of Lakes Through Sediment Removal
121
again in Vajgar fish pond in 1996. The temporary absence of Microcystis might be attributed to the removal of the upper layer of sediment that contained the inoculum of Microcystis. Colonies of Microcystis sink to the bottom at the end of each growing season, they winter in the top layer of sediment protected by a layer of mucus (Maršálek et al. 1996) and the colonies that successfully overwinter serve as an inoculum for the coming growing season. It is suspected that a new inoculum was introduced to Vajgar from upstream fishponds (Poulíčková et al. 1998). Despite that the internal loading of Vajgar fish pond was reduced by the sediment removal, the external loading remained high. Thus, due to the regular high supply of nutrients from the catchment, the Vajgar fish pond maintains the characteristics of eutrophic waters with conditions favourable for the mass development of cyanobacteria. Much more attention should have been, therefore, paid to the reduction of external sources of nutrients in the Vajgar catchment before the in-lake restoration project was attempted.
References Andersson G (1985) The influence of fish on eutrophic lake ecosystems. In: Proceedings of the international congress on lake pollution and recovery. European Water Pollution Control Association, Rome, pp 112–115 Andersson G, Berggren H, Cronberg G, Gelin C (1978) Effects of planktivorous and benthivorous fish on organisms and water chemistry in eutrophic lakes. Hydrobiologia 59:9–15 Bengtsson L, Fleischer S (1971) Sedimentundersökningar i sjöarna Trummen och Hinnasjön 1968–1970. (Sediment investigations in the lakes Trummen and Hinnasjön 1968–1970). Vatten 1:73–94, In Swedish with English summary Bengtsson L, Fleischer S, Lindmark G, Ripl W (1975) Lake Trummen restoration project. I. Water and sediment chemistry. Verh Int Verein Limnol 19:1080–1087 Björk S (1988) Redevelopment of lake ecosystems – a case study approach. Ambio 17:90–98 Björk S, Digerfeldt G (1965) Notes on the limnology and post-glacial development of Lake Trummen. Bot Notiser 118:305–325 Björk S (1972) Swedish lake restoration program gets results. Ambio 1:153–165 Cronberg G (1982) Phytoplankton changes in Lake Trummen induced by restoration. Longterm whole-lake studies and food-web experiments. Folia Limnologica Scandinavica 18:119 pp Digerfeldt G (1972) The post-glacial development of Lake Trummen. Regional vegetation history, water level changes and palaeolimnology. Folia Limnologica Scandinavica 16:104 pp Dykyjová D, Květ J (eds) (1978) Pond littoral ecosystems. Structure and functioning. Ecological studies, vol 28. Springer Verlag, Berlin, 490 pp Lettevall U (1969) Den kulturpåverkade sjön Trummen. Historik och utvecklingstendenser. (The human-influenced Lake Trummen. History and trends of development). Medd. fr. Forskargruppen för sjörestaurering vid Lunds univ., 24. Stencil, Inst. of Limnology, Lund, 18 pp (in Swedish) Lettevall U (1977) Sjön Trummen i Växjö. Förstörd–restaurerad–pånyttfödd (Lake Trummen, Växjö. Ruined–restored–recovered). Kronoberg County Government, Växjö, 32 pp (in Swedish with English summary) Maršálek B, Keršner V, Marvan P (1996) Vodní květy sinic. (Cyanobacterial water blooms). Nadatio flos-aquae, Brno, 142 pp (in Czech)
122
S. Björk et al.
Persson F (1969) Makrofytvegetation och litoraltopografi i sjön Trummen. (Macrophyte vegetation and littoral topography of Lake Trummen). Medd. fr. Forskargruppen för sjörestaurering vid Lunds univ., 32. Stencil, Institute of Limnology, Lund, 12 pp (in Swedish) Poulíčková A, Pechar L, Kűmmel M (1998) Influence of sediment removal on fish pond phytoplankton. Algol Stud 89:107–120 Teplý F (1927) Dějiny města Jindřichova Hradce (History of the town of Jindřichův Hradec). A. Landfras and syn. Jindřichův Hradec (in Czech)
Chapter 9
Treatment of Overgrown Shallow Lakes Through Macrophyte Control: The Case Study of Lake Hornborga, Sweden Sven Björk
Abstract A great number of lowered shallow lakes have become overgrown by perennial macrophytes. Because their ornithological and other environmental values have become lost, efforts are made to restore them. The most serious damage on the ecosystems is caused by the development of a rhizome and root felt and the accumulation of huge masses of coarse detritus. Removal of macrophyte detritus and biomass (including the root-felt) necessitates access to a variety of technical equipment in order to clean the bottom before the water level is raised in a lowered, overgrown lake. One such example is Lake Hornborga (30 km2), before degradation following water-level lowering and drainage, once ranked as one of the most valuable waterfowl lakes in northwestern Europe. After drainage the lake had become almost completely overgrown by reed and sedge. After the Swedish Government had decided to investigate the possibilities to restore the lake, limnological, ornithological and other field studies started in 1967. Ecotechnical restoration methods and prognoses for a restored lake were presented in 1972. The project goal was to transform the reed areas to open water (about 11 km2) and to keep the sedge-covered part (about 18 km2) for emergent vegetation. In 1982, the Parliament unanimously accepted the restoration plan including a raise of the water level by 1.4 m. The implementation of the plan was, however, delayed by drawn-out discussions about dikes and water depths most suitable for birds. Not until 1995 was the water level finally raised by 0.8 m. The long-term development of the ecosystem is being documented. Keywords Accumulation of coarse detritus • Perennial macrophytes • Raising of the water level • Root-felt • Water level lowering
S. Björk (*) Department of Ecology, Limnology, University of Lund, SE-223 62, Lund, Sweden e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_9, © Springer Science+Business Media B.V. 2010
123
124
S. Björk
9.1 Introduction In a great number of wetlands and shallow lakes, water levels have been lowered in order to get access to arable land. But because the aeration of the peaty and organic sediment soils causes increased mineralisation, in such cases, the newly-reclaimed agricultural land usually suffered rapid subsidence. The subsidence often amounts to 1–2 cm per year in temperate latitudes and to about 4 cm per year in tropical countries. Thus, after only some decades, the lowering of water levels in shallow lakes has nearly always created problems. In many cases the remaining lakes have become overgrown by macrophytes and the obtained arable land impossible to cultivate, being too wet due to the soil’s subsidence. Some of the lowered shallow lakes used to be outstanding waterfowl biotopes, and efforts are now being made to restore them to their former ornithological value, reclaim fishing-waters and recreate the qualities of the landscape in general.
9.2 Mass Development of Macrophytes in Shallow Lakes 9.2.1 Competition Between Phytoplankton and Aquatic Plants Shallow lakes, suffering from eutrophication, are often characterised by turbid water and heavy plankton blooms during the summer. As the light conditions in the water are bad, such lakes are typically devoid of submerged vegetation. However, floating-leaved and emergent macrophytic vegetation is often luxuriant with stands expanding. The lakeward margin of the reedbeds can develop into plaur formations (see Section 9.4.5). Alternatively, in shallow lakes which have become productive because of a moderate external supply of nutrients or by a lowering of the water level, submersed macrophytes, such as Ceratophyllum, Elodea, Potamogeton and Myriophyllum, may reach such a mass development that efforts to reduce them are needed in order to attain the goal of the management programme. Even charophytes, like Chara tomentosa, can appear in such masses that a reduction is required. When dealing with lakes of this type, the reduction of submersed vegetation has to be made with care. This is because the ecosystem might oscillate between two different types of structure and function: 1. One characterised by richly-developed submersed vegetation and clear water 2. The other characterised by richly-developed phytoplankton, turbid water, and no or sparse submersed vegetation A switch from submersed vegetation in clear water to phytoplankton and turbid water can also follow the introduction of grass carp (Ctenopharyngodon idella) as
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
125
a means to diminish macrophytes. In general, the available high concentrations of nutrients are utilised by primary producers: either the submersed macrophytes or the phytoplankton.
9.2.2 Notes on Basic Macrophyte Ecology In perennial macrophytes with well-developed rhizome and root systems, essential nutrients are assimilated from the bottom. In spring, these nutrients are translocated to the growing shoots but then, at least in part, returned, together with soluble carbohydrates, to the rhizomes (e.g. K) before the end of the vegetation period (Fig. 9.1). In contrast to those elements that are more or less preserved within the plant, other elements (especially Mg, Ca, Fe) are not transported back to the rhizomes in autumn. Thus, for this latter group of elements, there exists a one-way transport from the bottom, via the plant, straight to the water and the top sediment layer (Fig. 9.1). The fact that essential nutrients can be preserved and repeatedly utilised in the development of shoots, as in a perennial plant like Phragmites australis, has important ecological consequences. The occasional addition of nutrients (e.g. a temporary outlet of sewage) to the superficial soil layer can have a prolonged effect on the productivity of reed, as tested in field experiments with sewage
Fig. 9.1 Schematic illustration of the role of Phragmites and other rooted, perennial, aquatic macrophytes in the transportation of elements within the limnic ecosystem (From Björk 1967)
126
S. Björk
(Björk 1968). In lakes where eutrophication has resulted in the luxuriant growth of perennial macrophytes, the diversion of sewage, for example, does not have any immediate effect in the form of reduced productivity because the essential nutrients are re-circulated. The plants assimilate as much nutrients as they need for their normal physiological function. Above this level, a ‘luxuriant consumption’ can appear. Nutrients preserved within the perennial plant system may also have been taken up from the soil within the rhizosphere over a long time period. The deeply-rooted macrophytes in the minerogenic littoral zone frequently seem to be furnished with nutrients from the subsoil water – from which enrichment can take place. Thereby the standing crop can successively increase. The nutrient concentration of the water penetrating the ground and seeping through the littoral zone is therefore important for the development of the deeply-rooted aquatic vegetation growing in minerogenic soil landwards from the organic sediments. This is often vividly illustrated along shores where the groundwater is enriched by infiltrated sewage from summerhouses. The reproduction strategies of different aquatic plant species (cf. Sculthorpe 1985) is of decisive importance for their dispersal and recolonisation within treated wetland areas. The quantitatively-dominating species are characterised by rapid vegetative reproduction. Although the distribution of Phragmites australis is dependent on both generative and vegetative reproduction, the colonisation within a wetland where this species is already most successfully established takes place by means of rhizomes and drifting rhizome pieces. The seeds of Phragmites have very specific environmental demands for the germination and development of seedlings. These demands are realised only under conditions corresponding to those found in the moist portion of the littoral zone or appearing over large areas during low-water periods. New stands of the common reed never develop from submersed seeds, in contrast to Schoenoplectus lacustris which successfully colonises new areas in open water by means of both plant fragments and submersed seedlings (cf. Ekstam et al. 1992).
9.3 Macrophyte Control 9.3.1 Pre-project Investigations and Planning When planning the restoration of shallow lakes overgrown by macrophytes, it is important to consider all interest groups focused on the wetland area. The most common combinations of interests are recreation (like swimming, canoeing and windsurfing), ornithology and angling. In the latter case, the ecological demands of actual fish species over the whole life cycle deserve attention. The overall aim ought to be to create varied environmental conditions to enhance a high diversity of flora and fauna. A mosaic of macrophyte stands and open water is most often optimal, although some bird species need large reedbeds for nesting
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
127
and other species need large open water areas. Other aspects such as the environmental requirements of birds during migration, nesting and moulting periods, as well as the food requirements of young waterfowl, also need to be given attention. Preceding all activities for the control of macrophytic vegetation, there should be some pre-studies undertaken, including: a compilation of the deterioration history of the wetland, mapping of vegetation, water depth during the year, bottom conditions (organogenic, minerogenic, soft or hard, location of boulders, etc.), nesting sites of bird species, etc. The information collected during these pre-studies is of great importance for the correct evaluation of results, as well as for the correction and improvement of methods.
9.3.2 Restoration Methods 9.3.2.1 Technical Equipment Among the three different types of methods generally mentioned in connection with the reduction of undesirable vegetation, i.e. biological, chemical and mechanical, only mechanical methods are treated in this handbook because they are the only ones possible to have under continuous and complete control. The available technical equipment covers a wide spectrum from simple scythes, small boats with cutting bars, to expensive aquatic harvesters. In addition to the amphibious and pontoon machines (see Section 8.3), a pontoon harvester for removal of submerged vegetation has been constructed for Scandinavian wetlands. The cut material is automatically loaded and brought to land by the harvester. Contrary to other machines with about the same function, this type can be easily transported from one site to another. Pontoon machines can also be equipped with rakes for collection and transport of cut material or masses of freefloating plants to the shore. The most suitable propulsion of pontoon machines is by paddle wheels. 9.3.2.2 The Need to Remove the Cut Plant Material Plant material cut by mowing-machines or loosened by rotovators should be transferred to land. If, in large projects, plant material is disintegrated in a macerator and pumped to land, precautions should be made to prevent pollution from the sap. Easily degradable fresh material and plant sap consume oxygen and act as a fertiliser. Slowly-degrading matter covering the bottom, like coarse Phragmites detritus, constitutes an unsuitable substrate for submersed plants and for bottom animals. Such detritus is also easily moved along the bottom by water currents and can accumulate in stands of, for example, Schoenoplectus, originating from submersed
128
S. Björk
seedlings in areas of open water. Contrary to the stems of Phragmites, those of bulrush (Schoenoplectus) wither in autumn. However, because the basal portions remain, the bulrush stands growing in open water act as very efficient traps for drifting detritus, both minerogenic and organogenic. Most intensively, the accumulation among stubbles and young shoots takes place during the windy, spring high-water periods. In this way ‘verlandungs’-islands are created. Because of the tough rhizome system of Schoenoplectus (acting like reinforcing netting) such vegetative islands are often difficult to get rid of. 9.3.2.3 Time Schedule In principle, the most suitable time for cutting perennial emergent plant species is when they have their highest concentration of nutrients in the shoots. The removal of the standing crop thus means a maximal depletion of nutrients from the plant. In the case of Phragmites, this stage is reached just before the appearance of the panicles. Repeated cutting can result in the extermination of a species within a treated area. However, provided the environmental conditions are not altered, the same or similar species will successively recolonise the cleared bottom. Although the removal of green emergent plant material from a wetland also means a certain reduction in the total nutrient capital of the ecosystem, the real importance of this activity for the nutrient economy is rather negligible. However, when dealing with the huge biomass of submersed plants like Ceratophyllum and Myriophyllum, it is imperative to remove it from the wetland before any decomposition of the plant material has started, i.e. before the release of nutrients from the easily-degradable matter. Another aspect to observe is that the cutting should be done before seeds have become fully developed and germinative, and thus get easily lost from the cut material of, for example, Potamogeton. Both when cutting standing crop and rotovating the root systems of hydrophytes, especially submerged plants, late during the vegetation period, the cleaned area can also be seeded with hibernacula of various kinds, i.e. buds adapted for overwintering and propagation. 9.3.2.4 Use of Fire In wetlands overgrown by emergent vegetation and with the bottom covered by a layer of coarse detritus (see Fig. 9.4), it is sometimes possible to make use of fire to retard the ageing process, or as a means to prepare for treatment of the root-felt. The best conditions for burning are mostly in late summer and early autumn (appropriate training for firemen). Because the rhizomes and roots are not damaged by the fire the re-growth of Phragmites shoots occur. The burning of reed may thus result in an increase in reed biomass in the following years due to better light and temperature conditions after the removal of the detritus. If the aim of the
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
129
treatment is to get rid of the emergent vegetation, rotovation of the root-felt is necessary after burning the stems and coarse detritus layer.
9.3.2.5 Management of the Littoral Zone As long as the littoral zones of lakes were grazed by cattle, this zone was kept open up to the depth which cattle reached when grazing aquatic vegetation (Fig. 9.2). When grazing stopped, the littoral became rapidly overgrown, and nesting as well as foraging sites for waterfowl became lost. As training of military vehicles takes place in this zone (under the agreement with the Swedish Environmental Agency), the root systems of the emergent perennial plants get destroyed and the littoral becomes productive with respect to a diverse fauna (insects, molluscs etc.) and flora (Lemna, Ceratophyllum, Hydrocharis, etc.). Like grazing, this form of artificial treatment has to be regularly undertaken along shore reaches that are not exposed to ice and strong water movements, should it be decided that the littoral zone be prevented from overgrowing.
9.4 Case Study: Lake Hornborga, Sweden 9.4.1 Background Among the numerous lakes degraded through water-level lowering in Sweden, Lake Hornborga (in Swedish, Hornborgasjön, location 58°29'N, 13°34'E, Fig. 9.3) in the province of Västergötland, is the best known. This is partly because the lowering was a legal scandal and a big economic failure, and partly because extensive basic investigations were made there, concerning the possibilities to restore this type of shallow, drained lakes. The eco-technical restoration methods elaborated at Lake Hornborga were applied there already in the 1960s (Björk 1972).
9.4.2 Causes of Degradation Until man interfered with the well-organised complex of components that functioned within the Lake Hornborga ecosystem, it maintained a rather high degree of productivity without suffering from rapid ageing. This is quite remarkable for a lake of this size (30 km2), shallowness and high trophic level. The lake used to have a maximum depth of 3 m, but most of the lake was much shallower. A constellation of factors such as rapid water renewal, well-situated inlets and outlets (Fig. 9.3), strong water and strong ice movements, provided a good system for transporting
130
S. Björk
Fig. 9.2 The effect of grazing on the littoral vegetation (From Luther and Munsterhjelm 1983)
matter out of the lake and prevented it from becoming overgrown. The emergent vegetation was mainly restricted to wind-protected shore areas. The organic matter of the richly-developed submerged vegetation easily decayed and was transported out of the lake, while the precipitated calcium carbonate particles covering these
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
131
plants settled to the bottom. Before its degradation, following the lowering of its water level and finally, in 1933, its complete drainage, Lake Hornborga ranked as one of the most valuable lakes for waterfowl in northwestern Europe. Since 1802, the lake had been lowered five times in attempts to obtain arable land. The last big failure, in 1932–1933, resulted in a bottom drained in the summer and which, in a considerable part of the lake, consisted of lake marl. The hilly land area of about 616 km2 that drains into Lake Hornborga needs the lake as a reservoir to catch the rain and snow meltwater that rushes down the hills to the plain below. In 1954, the water level was slightly raised in the northern part (12 km2) of the lake that was diked-in. This could be looked upon as a large-scale field experiment. As described below, it revealed, amongst other things, the necessity of removing the macrophytic vegetation and of restoring a clean sediment bottom in areas intended to be re-created and preserved as open water with submerged vegetation. During the 1930s until the middle of the 1960s, the Hornborga area went through the typical development for a lowered, shallow lake and its drained surroundings. The lake area became overgrown by emergent vegetation. In fact, the slight raising of the water level in 1954, to a maximum depth of c. 80 cm in the diked-in portion, considerably improved the environmental conditions for common reed which had colonised the lake bottom. Huge masses of coarse detritus accumulated and filled in the lake basin (Fig. 9.4). At the same time, the surrounding organic soil areas subsided. Altogether this resulted in fairly severe flooding and it became impossible to cultivate the successively wetter organic soils. Lake Hornborga lost its value as an outstanding waterfowl habitat because the lake ecosystem’s structure and function were completely destroyed. Monocultures of common reed (Phragmites australis, Fig. 9.9 – upper left photo) and slendertufted sedge (Carex acuta) nearly covered the whole lake area in 1967. Mixed plant communities including living and dead willow bushes (Salix spp.) made large areas almost impenetrable.
9.4.3 Basic Research, Arguments and Directives for Restoration The legal course of events, including the long series of illegal ingredients, leading to the destruction of the great natural asset of Lake Hornborga, was revealed by Swanberg (1959a, b, 1968, 1971). He elucidated the very serious economic and ecological consequences of the drainage project. Thanks to Swanberg’s basic research and persevering, convincing argumentation, it was successively realised by all parties concerned that the drainage project was a multidimensional failure that had severely hit both nature and culture. The Swedish Government finally decided to investigate the possibilities to restore Lake Hornborga. According to governmental directives, investigations to determine whether or not the lake could still be restored and whether restoration would be sufficiently permanent were to be carried out (Vilborg 1973). In the case that restoration would be theoretically possible, methods to attain this goal were to be elaborated.
132
S. Björk
Fig. 9.3 Lake Hornborga, Sweden. Location; topographic map from 1967 to 1969; and recent overgrowth by emergent vegetation (1905, 1917, 1936, 1968)
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
133
Fig. 9.4 Lake Hornborga. Distribution of living and dead material in stands of Phragmites. At the start of the restoration of the bottom areas overgrown by Phragmites, the thickness of the B and C layers amounted to ca. 75 cm (From Björk 1972)
9.4.4 Restoration Goals The ecological goal for most restoration projects in lowered and overgrown lakes is to create an open water area and a mosaic of open water and emergent/submerged vegetation. The restored lake/wetland should be brought to a state of permanence where no future extensive management programme is needed. For responsible, long-term planning of wetland restoration, management and protection, this is of the utmost importance. Shallow lake ecosystems are not at all static but highly dynamic systems and can suffer from very rapid ageing processes. After water level lowering, as well as after an insufficient increase in water levels as a means of ‘restoration’, such lakes usually pass through a transient period of flourishing birdlife. Sometimes, as in the Lake Hornborga case, it is still difficult to convince the general public that this period is of just short duration and that ecologically realistic measures to counteract the rapid ageing must be taken in order to give the ecosystem a sustainable character. When planning for the restoration of such systems to sustainable units – according to
134
S. Björk
human’s time-scale – it is necessary to apply a holistic approach not only in space, i.e. to comprise the whole ecosystem including its catchment area, but also in time, i.e. to allow for the factors which have an influence on the speed of the ageing processes.
9.4.5 Field Investigations The field investigations started in 1967 and were carried out by a team of ornithologists (under the leadership of Dr. P.O. Swanberg), limnologists, technologists, hydrologists, economists and agriculturalists. In 1968, the lake and its subsiding surrounding areas were mapped using aerial photographs to obtain contour maps with a 25 cm interval. Following the governmental directives, the necessary limnological restorative measures and developmental prognoses were elaborated, based on the specific conditions in the degraded Lake Hornborga itself. As the ecological properties in every lake/wetland ecosystem are unique, a tailor-made restoration plan has to be designed for each individual project. One year of limnological studies made it quite clear – theoretically – that Lake Hornborga could be restored (Björk 1972). The most serious problem following the lowering of water levels in shallow lakes is the development of a root-felt in the upper sediment layer and the accumulation of coarse plant material (detritus, see Fig. 9.4) produced by a highly-productive macrophytic vegetation. With such clear evidence, the results of the limnological investigations demonstrated that before the decisive step for the restoration of Lake Hornborga – i.e. the raising of the water level – was to be taken, the bottom had to be treated in order to get rid of the coarse plant material and the root-felt. In 1968, large-scale field experiments were begun in order to develop the technical methods necessary to counteract the otherwise irreversible damage, i.e. the huge masses of accumulated detritus as well as the root-felt of the emergent macrophytes had to be removed. If the water level is to be raised over areas overgrown by sedge (Carex), reed (Phragmites), bulrush (Schoenoplectus) etc., the gas (mainly methane) produced in the bottom accumulates in and beneath the root-felt. After a period of time the rootfelt would, therefore, float to the surface and would soon be overgrown by new vegetation. This process leads to plaur formation (floating stands of emergent macrophytes).
9.4.6 Project Design The basic limnological investigations carried out in all parts of Lake Hornborga included both the inflowing streams and the outflow, and the well-documented changes in the ecosystem following the drainage. Bottom levels, sediment, peat, water and flora were studied and the results were synthesised with findings from
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
135
the groups responsible for ornithology, hydrology, etc. The ecological restoration plan for Lake Hornborga was presented in 1973 (Plan/73, cf. Björk 1972; Swanberg 1972; Vilborg 1973). In Lake Hornborga, as elsewhere, the sedge root-felt posed a problem, as it was thick, resistant and for economic reasons impossible to remove from the very large areas covered by it (ca. 18 km2). In small water bodies, the floating root-felt can be removed by means of a dragline, or cut up, towed to the shore and removed. The reed root-felt can easily be cut by amphibious rotovators that were constructed for the Lake Hornborga project (Fig. 9.5). Following the governmental directives, the limnological project goal for Lake Hornborga was to transform the reed (Phragmites) areas to open water (about 11 km2) and to keep emergent vegetation in the area covered by sedge (Carex) (about 18 km2). When the water level was raised, the sedge root-felt would float to the water surface and be recolonised by reed (Phragmites) and bulrush (Schoenoplectus). Before raising the water level, the tough sedge root-felt could be removed from small areas by means of amphibious excavators. After raising the water level these parts would be preserved as open waters surrounded by plaur vegetation and be attractive for birds. The procedure to replace emergent plant cover by submerged vegetation is illustrated in Fig. 9.6. During the field experiment in the years 1968–1972, convincing results for a successful restoration using the above series of methods were obtained. For cutting macrophytic vegetation and destroying the root-felt, prototypes of
Fig. 9.5 Lake Hornborga. Prototype of rotovator for removal of the stubble mat and rhizome layer of Phragmites. The slowly-moving knives cut the rhizomes in long, easily-floating pieces (Photo: Sven Björk 1972)
136
Fig. 9.6 The Lake Hornborga restoration project (From Björk 1972)
S. Björk
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
137
amphibious and pontoon machines were developed. The use of the constructed equipment is schematically shown in Fig. 9.6, according to the dry and flooded conditions characteristic for the drained Lake Hornborga in the 1960s and 1970s. During dry periods in late summer and early autumn, it was sometimes possible to quickly clean several square kilometres from accumulated reed detritus by means of fire, especially in the sections of the lake where willow bushes prevented the immediate use of machines. However, the fire did not destroy the root systems. Therefore, the productivity of the reed subsequently became higher since the fire had improved the environmental conditions by the removal of the layers of settled detritus and dry stems which had been decreasing both temperature and light at the bottom. It is usually most practical to start the removal of the macrophytic vegetation along the wind- and wave-exposed shores. During periods of high water level, the cut and loosened plant material is transported to the exposed shore. During low water periods it is possible to burn it on the shore or to collect and compost it. In small-scale projects, it is practical to carry out the cutting and root destruction in sub-areas surrounded by floating booms. All the loosened material is dragged within the boom to the shore and removed from the lake by means of a conveyer belt. With the procedure as described in Fig. 9.6, enormous masses of coarse detritus deposited in the stubble mat were loosened and together with the rhizomes and roots transported by the spring high water to the shores along the wind-exposed portions of the lake (Fig. 9.7), where they were subsequently burned in the summer. When the detritus was removed, the consolidated original gyttja (mainly lake marl) again became the bottom. If covered by sufficiently deep water, the reed
Fig. 9.7 Lake Hornborga. Phragmites rhizome pieces after cutting with rotovators (Photo: Sven Björk 1969)
138
S. Björk
monoculture would be replaced by submerged vegetation, and a rich bottom fauna, microbenthos and periphyton develop. The biotope changes, already documented during the experimental period, resulted in a very obvious improvement in the waterfowl fauna (Fig. 9.8). The restoration of bottom areas primarily overgrown by common reed (Phragmites) was an easy procedure by means of the technical equipment developed for the Lake Hornborga project. Even if the root-felt of bulrush (Schoenoplectus) and sedge (Carex) could have been cut by rotovators, the main part of the loosened material would have remained on the bottom, stationary or partly drifting. In the case of common reed, the internodes of big pieces of stems and rhizomes are filled with gas, providing them with excellent buoyancy (Fig. 9.7); it is imperative, therefore, that the rotovators cut the rhizomes in big pieces. Machines making slurry of both stems and rhizomes are destructive because the buoyancy of the material gets lost and the bottom remains uncleaned and becomes covered by a layer of coarse, partly drifting particles. Methane gas develops in the degradation process of the organic matter when prepared in this way. After treatment of 11 km2 of the bottom as described, and manual removal of trees and especially bushes along the shores and in the bays of the former lake, the restoration plan presented in 1973 (Plan/73) recommended the raising of the water level in two steps, first by ca. 1 m and then, after about 5–10 years by a further 0.5 m. This procedure was recommended in order to get the wind-exposed shores
Fig. 9.8 Lake Hornborga before and after the demonstration experiments for directing the primary production from emergent to submersed vegetation. Water level not yet raised. Comparison between conditions in 1965 and 1971 (Data from H. Berggren and P.O. Swanberg; from Björk 1972)
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
139
worked up, cleaned and washed by the waves in a natural way. The primary raising of the water level should be sudden and big enough to prevent both successive recolonisation of emergent species, mainly Phragmites, from the littoral zone and the survival of initial stands in open shallow water areas. Before the lake degraded completely, Schoenoplectus was the characteristic species. After the completion of a restoration according to Plan/73, this species would again develop outside Phragmites stands along the shores and within plaur areas. With the distribution of the emergent vegetation under control by means of water depth, wave and especially ice action, the sediment bottom cleaned from coarse detritus could be largely covered by underwater vegetation (Fig. 9.9 – middle row left).
Fig. 9.9 Lake Hornborga. Demonstration of restoration methods and recolonisation of treated areas (Photos: Sven Björk)
140
S. Björk
According to Plan/73, lowland areas adjacent to Lake Hornborga should not be cut off from the restored lake as shallow shores and vast wetlands also belonged to the most characteristic features of the former lake (Vilborg 1973). Plan/73 recommended the construction of: (1) a compulsory dike at the outlet for making it possible to raise and regulate the water level; and, in addition to that, (2) short dikes for the protection of still arable land against high-water floods, part of these to be landwards from large plaur areas (Fig. 9.10).
9.4.7 Project Implementation In 1977, the limnological restoration plan of 1973 was unanimously accepted by the Riksdag (Parliament) and in 1982, the Water Court and the Government granted permission to raise the water level, in two steps, by 1.4 m. The Swedish Environmental
a
Tranum
b
Tranum Härlingstorp
Härlingstorp
a b
a b Stenum
Fl
ia
n
Stenum
FlStenumsmaden ian
Stenumsmaden
Fågeludden
Fågeludden
Trestena
Trestena
Ytterberg
Ytterberg
Ore nabb
Skara
Ore nabb
Skara AlmeöHornborgamaden Hor nbo rga ån
Almeö Hornborgamaden Hor nbo rga ån
Dagsnäs
Dagsnäs
Bjurum Bredegården n så m ru
u Bj Stora Bjurum 0
Falköping 1
2 km
Sätunamaden
Bjurum
Bredegården
n så
Dike at the outlet (compulsory) Dikes considered to be economically realistic and ecologically safe. Dikes not considered to be realistic
um ur Bj Stora Bjurum Falköping 0
1
2 km
Sätunamaden
Dike at the outlet Suggested dikes. The Hertzman dike across the Ytterberg bay and other dikes should have been constructed on several metres of organogenic deposits. The Environmental Protection Agency later cancelled their plans for extensive diking.
Fig. 9.10 Lake Hornborga. Distribution of open water (a) with submerged and emergent vegetation and areas predetermined as plaur (b) after raising the water level according to restoration PIan/73. Map A: location of dikes according to Plan/73. Map B: dikes according to the Environmental Protection Agency, Stockholm
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
141
Agency was responsible for the implementation of the plan. However, under their management, dikes, finally reaching a total length of 25 km, were projected around the lake and along tributaries. In addition to these dikes, efforts were also made to get permission for construction of the Hertzman’s dike for reclamation of one of the lake’s broad bays (Fig. 9.10). The latter was, however, rejected by the Government. Because geological investigations had not been made at the suggested location of the dikes, the risks for leakage had not been assessed. Pleading the argument that depths greater than 0.4 m is a waste of water with regard to the environmental demands of many duck species, together with that of the leakage risks, the Environmental Agency changed the plan for the restoration of Lake Hornborga. The projected dikes (with the exception of the one at the outlet) were abolished and the water was to be raised by only ca. 0.8 m, with the first water rise by 0.2 m in 1992. An overview of the treatment of Lake Hornborga, after shrinking the Plan/73, was presented by Hertzman and Larsson (1999). They argue that ‘the technical planning and the presentation of the restoration measures were formulated before the ecological implications of the suggested steps were fully analysed’ and that ‘the prevailing idea’ was ‘that if the limnological aspects were adequately addressed, the necessary ecological conditions for the birds would more or less automatically be created’ (Hertzman and Larsson 1999, p. 68). Their plan, however, was not equipped with any environmental impact assessment (EIA). In accordance with the limnological prognoses given in the Plan/73, plaur formation (the base provided by root-felt of Carex, etc., being lifted up by methane gas produced at the bottom and coming to the surface) soon started after the rise in water level and ice movements occurred. The bird life also improved considerably. Incorrectly-treated bottom areas were characterised by an intense production of methane. An unforeseen result has been the strong increase in the number of greylag geese (Anser anser) efficiently consuming reed-bed plants. The striking reduction of Phragmites has forced more than 75% of the marsh harriers (Circus aeruginosus) to breed in Salix bushes instead of in reed-beds. In 2001, a total of 36 breeding pairs were recorded at the lake (Pettersson 2002). The future development of the ecosystem will serve as an illustrative example of the importance in restorative planning of a holistic view in both space and time. The further the ageing processes (filling in with sediment, peat and coarse detritus) have advanced in a shallow lake basin, the more difficult it is to restore it by raising the water level; ultimately, it becomes impossible. If Lake Hornborga should be restored in a sustainable way, then that part not covered by revegetated plaur and bottom-rooted emergent plants must be large and deep enough to give the ecosystem some form of functional longevity, as stressed in the original governmental directives for the restoration of the lake, and not just give it the temporary appearance as an attractive waterfowl habitat. The open water should act as the ‘heart and kidney’ of the whole area, constituting an integrated part of the complex system, functionally powerful enough to preserve its vitality after restoration. The persistence of the presently-achieved results would, otherwise, be a matter of decades, and not centuries. Every rise in the water level would, of course, cause a flourishing period for waterfowl, but the lower the rise the shorter the period of a functioning lake.
142
S. Björk
It must not be forgotten that even a limnologically correctly-restored Lake Hornborga would be severely damaged by the former water-level lowering – and that the restored system must be given the structural and functional capacity to withstand the increased nutrient loading. It should, amongst other things, have the capacity to metabolise nutrient-rich water from the agricultural catchment area (the 1954 diked-in lake was supplied with such water for only a short spring period), as well as the anoxic water pressed out from the plaur sections during periods of decreasing water level. In this connection, chain reactions, including periodic oxygen deficiency and nutrient release, would appear and would affect animal survival and the fertilisation of downstream waters. Within the open water area, the bottom should consist of sediment cleaned from root-felt and coarse detritus. Furthermore, the open water should be large enough to enable the formation of blue ice capable of pressing and pushing to keep exposed shores open as the ice did in the lake before its degradation. The evolution of shallow lakes/wetlands is, in several respects, dependent on their conditions in the extreme years. Most important are the effects of dry periods with low water levels, especially if these effects in regulated lakes cannot be compensated by the effects of extreme high water periods. In dry years, the artificiallyshallow lake, overgrown by emergent vegetation, will suffer from extra water losses due to the intensive transpiration from the plant cover. Within the dry littoral zone, plants will transpire the groundwater which would have otherwise supplied the lake. In addition to this, agricultural irrigation may reduce the supply of water during dry seasons. The summarising effect of the low water levels in extreme years is the lakeward expansion of emergent macrophytic vegetation, resulting in an increased production of accumulating coarse detritus. The Plan/73 for the restoration of Lake Hornborga also included measures to secure an even bottom in order to prevent sites from accumulating detritus and developing stands of emergent plants. Dikes along the drainage canals had, therefore, to be levelled as part of the efforts to restore, as much as possible, the former wind-induced regime for in-lake water and ice movements.
9.4.8 Methods to Increase Water Depth When the water depth of a lake becomes too shallow, there are in general three possibilities to increase it: to raise the water level; to lower the bottom; or to combine both these measures. Under all circumstances the bottom has to be treated in order to get rid of the accumulated coarse plant material and the root-felt, a procedure which also results in a lowering of the bottom level. In Lake Hornborga, the removal of the root-felt and the dense stubble mat according to the Plan/73 would have lowered the bottom by ca. 40 cm. Lake Hornborga is situated in an area where the bedrock, in some parts, consists of limestone. The characteristic sediment deposit in the northern section of the lake
9 Treatment of Overgrown Shallow Lakes Through Macrophyte Control
143
is lake marl, i.e. calcium carbonate precipitated as a result of the photosynthetic activity of the rich submersed vegetation. The marl is of high quality and excellent for the liming of acidified areas instead of spreading finely-ground limestone/chalk that is obtained from quarries and crushed and ground to fine fractions. In Sweden, where inland waters in oligotrophic areas are suffering from acidification, the liming of acidified lakes, brooks and catchment areas is state-financed (ca. US$11 million per year in the middle of the 1980s). In lowered and completely-drained lakes with lake marl as the characteristic sediment, modern automatic extraction methods have opened up possibilities for efficient sediment-mining aimed at the restoration of the lowered lakes, and the treatment of acidified land and waters. For Lake Hornborga, situated on the geological borderline between limestone and gneiss, lake marl is a characteristic sediment (Fig. 9.11), and immediately southwest of the lake is the area of Sweden where the soils and waters are most severely affected by acidification. At the start of the Hornborga project, in the mid-1960s, acidification was not yet identified as an environmental problem and large-scale liming as a remedial measure was, of course, not introduced until later. As pointed out at the beginning of this case study (see Section 9.4.2), the ageing of shallow, productive lowland lakes is caused by the deposition of sediment and peat, produced in the ecosystem. In Lake Hornborga, the sediment in that part of the lake where it is possible to restore to open water consists of thick, homogeneous deposits of high-quality lake marl. The removal of the top layer of marl would not have changed the character of the bottom. The increase in water depth, would, on the contrary, have had a real rejuvenating effect on the system. Thus, after liming had been given such a high economic priority in environmental protection financed by the state, there would have been an excellent opportunity to investigate this innovative alternative for reaching the goal of a sustainable Lake Hornborga
Fig. 9.11 Lake Hornborga. Stratigraphical section from north (N) to south (S). Figures on both sides of the diagram denote level above the sea (m). The upper dashed line connects the upper levels for consolidated gyttja and the lower line the upper levels for minerogenic soils
144
S. Björk
through the combination of lowering the bottom and raising the water level, although this possibility did not appear until after half-time in the protracted project. Within the lake and wetland restoration-research sector there is an urgent need for the collection of experience of ecologically-realistic and economically-sound methods applied in demonstration projects.
References Björk S (1967) Ecological investigations of Phragmites communis. Studies in theoretic and applied limnology. Folia Limnol Scand 14:248 Björk S (1968) Makrofytproblem i kulturpåverkade vatten. (Macrophyte problems in waters affected by humans). Limnologisymposion, Helsinki, 8 pp (in Swedish) Björk S (1972) Statens naturvårdsverks utredning beträffande Hornborgasjöns framtid. Den limnologiska delutredningen. Förutsättningar, metoder och kostnader for Hornborgasjöns restaurering (The national environmental protection agency’s investigation concerning the future of Lake Hornborga. The limnological investigation. Conditions, methods and costs). Statens naturvårdsverk, Stockholm, PM 280, 2nd ed 1978, 55 pp (in Swedish) Björk S (1976) The restoration of degraded wetlands. In: Smart M (ed) Proceedings of the international conference on conservation of wetlands and waterfowl. Heiligenhafen 1974, pp 349–354 Ekstam B, Granéli W, Weisner S (1992) Establishment of reedbeds. In: Ward D (ed) Reedbeds for wildlife. Proceedings of the conference creating and managing reedbeds with value to wildlife. Histon, Cambridgeshire 1991, pp 3–19 Hertzman T, Larsson T (1999) Lake Hornborga, Sweden – the return of a bird lake. Wetlands International Publication, Wageningen, The Netherlands, 82 pp Luther H, Munsterhjelm R (1983) Inverkan av strandbetets upphörande på hydrolitoralens flora i Pojoviken (Influence of the ceased shore grazing on the hydrolittoral flora of the Pojoviken inlet, S. Finland). Memoranda Soc Fauna Flora Fennica 59:9–19 (in Swedish) Pettersson B (2002) Kärrhökars boplatsval i Hornborgasjön före och efter restaureringen (Harriers’ choice of nesting sites in Lake Hornborga before and after restoration). Vår fågelvärld 2:24–25 Sculthorpe CD (1985) The biology of aquatic vascular plants. Edward Arnold, London, 610 pp Swanberg PO (1959a) Hornborgasjön som fågelsjö. (Lake Hornborga as a waterfowl lake). Från Falbygd till Vänerkust, Lidköping, in Swedish Swanberg PO (1959b) Hornborgasjöns sänkningar. (The lowerings of Lake Hornborga). Från Falbygd till Vänerkust, Lidköping, pp 132–151, in Swedish Swanberg PO (1968) Hornborgasjön och människan (Lake Hornborga and man), vol 22. Falbygden, Falköping, pp 175–206, in Swedish Swanberg PO (1971) Hornborgasjön. (Lake Hornborga). In: Johansson H (ed) Boken om Gudhem. Falköping, pp 274–303 (in Swedish) Swanberg PO (1972) Metodik i den ornitologiska inventeringen av Hornborgasjön 1969–1971 (Methods used in the ornithological census of Lake Hornborgasjön in 1969–1971). Vår Fågelväld 39:369–376 (in Swedish with English summary) Swanberg PO (1973) Hornborgasjön som fågelsjö. Omitologisk undersökning i statens naturvårdsverks utredning om sjöns framtid (Lake Hornborga as a waterfowl lake. The ornithological study of the national environmental protection agency’s investigation concerning the future of Lake Hornborga). Statens naturvårdsverk, Stockholm, PM 280, 2nd edn, 1978, 100 pp (in Swedish) Vilborg L (1973) Hornborgasjöutredningen (The Lake Hornborga investigation). Statens naturvårdsverk, Stockholm, PM 280, 34 pp (in Swedish)
Chapter 10
The Stream and Beyond: Reinstating Natural Functions in Streams and Their Floodplains Bent Lauge Madsen
Abstract Stream restoration has moved from in-stream habitat restoration to reinstating an intact stream-floodplain system. This approach restores not only the dual stream-riparian habitats, but influences the whole stream system – water quality improves, and discharge fluctuations are dampened. Water quality in downstream water bodies also benefits – a waterlogged meadow can eliminate nitrate and lock iron in solid state, and phosphorus can be retained in a flooded meadow. While traditional stream management tended to minimise the retention capacity, ecologically-reasoned management should maintain a good retention capacity for particulate and dissolved organic matter – the main food source for invertebrates. Stream quality can be described in five dimensions: water quality, discharge, in-stream physical variation, longitudinal continuity, and mutual stream-floodplain hydrological contact. Prudent weed cutting has proved a most suitable and cost-efficient tool in the re-establishment of lost salmonid and invertebrate habitats. The inherent disturbance capacity of running waters should be acknowledged: prudent stream management should fully utilise the running waters self-regenerative capacities. Keywords Flooding • Floodplain • Habitat • Stream quality • Stream restoration
10.1 Introduction More than a quarter of a century ago, the father of stream ecology, Noel B. Hynes (1975), stressed: We must, in fact, not divorce the stream from its valley in our thoughts at any time. If we do, we lose touch with reality.
B.L. Madsen (*) National Agency of Forest and Nature (retired), Watercastle Old School Research Station, 38 Kirkensgaardvej, DK-7620, Lemvig, Denmark e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_10, © Springer Science+Business Media B.V. 2010
145
146
B.L. Madsen
This wise statement – a harbinger of the new millennium’s agenda of stream-valley management – was saying: the valley rules the stream. During the quarter of a century since, the intimate relationship between a stream and its catchment – a more modern but less poetic word for valley – has become a focal point in stream science as well as in stream management. Most emphasis has been focused on the neighbourhood nearest the stream, the riparian zone or the floodplain, i.e. the parts of the valley that are regularly flooded. Water, sediment and nutrients enter the stream from the catchment (synonym for drainage basin, or watershed in American usage), but the relationship between a stream and its catchment is not a one-way affair. Sediments and nutrients that flow in the stream can be returned to the floodplain through regular flooding events. This makes the floodplain a very attractive place to live – and not only for plants and animals. Since ancient times, mankind too has benefited from the regular flooding which made sustainable agriculture possible. Floodings were a blessing but also a curse. Excessive floodings put the riparians, i.e. the people living near the streams, at risk. The increasing habitation of the floodplains fostered a desire to rule the stream and to tame the flood in order to protect life and property. The engineering answer to flooding was to get rid of the water by transporting it as fast as possible to the sea. Streams were confined to artificial channels, constrained by dykes and managed as efficient gutters. The intimate, mutual relationship between stream and valley became fragile and, eventually, ended in divorce. The stream water, along with its nutrients, was prevented from moving back to the floodplain. A steady deterioration of water quality appeared, caused by sewage discharged from urban areas into the streams and nutrients leached from intensive agriculture. While farmers and factories prospered, fishermen suffered. And the rich nature in the streams and floodplains vanished, not only because of pollution but also due to habitat destruction. Increasing public and political awareness has initiated measures to restore nature’s lost values, starting with a gargantuan effort to clean the water and followed by restoration work to reinstate habitats in streams. However, many years elapsed between Hynes’s words in 1975 and the time when first approaches to reunite the stream with its floodplain were considered. Violent floods in recent years have opened our eyes to the wisdom in Hynes’s words. Finally, we are realising that the valley rules the fluvial behaviour of the stream, and that nature’s own devices may be the key to natural and less expensive flood control. This insight has paved the road to new and exciting challenges to guide stream restoration out of its infancy. We started by re-instating some lost in-stream structures, such as meanders and riffles, in order to make streams a better place for plants and animals. Now we are moving the restoration out into the floodplains, realising that it is the functioning floodplains that are crucial for flood control as well as for reinstating natural biological communities in the stream and in the floodplain. The increasing evidence that riverine wetlands are a mediator of a cleaner environment in downstream water bodies, including the open sea, and that they can ease
10 The Stream and Beyond: Reinstating Natural Functions in Streams
147
violent floods, greatly widens the perspectives of stream restoration. Hidden in our separated stream-valley system is a natural functioning system, waiting to be revealed when the two are reunited. As Francis Bacon (English philosopher 1561– 1626) said: Nature is often hidden, sometimes overcome, but seldom extinguished.
This ancient statement sets the theme for this chapter: described in the first part are some crucial fluvial and biological properties of naturally functioning streamcatchment systems which are impaired in our channelised streams. In the second part, ways how some of these lost properties can be re-instated to approach a natural stream floodplain function are discussed.
10.2 More Than a Stream Channel Streams are often regarded as unidirectional, linear systems. They look that way – they are waterways, flowing downhill. The current, quite literally, orchestrates the fluvial and biological processes from the stream’s source to its mouth. It also maintains the mosaic of habitats; it has been a driving force in the evolution of stream flora and fauna. The current is what makes streams alike, and what makes them different. But streams are much more than waterways. Not the stream itself but its whole catchment is the fundamental unit in stream functioning and stream-related management. The water that becomes visible in the stream channel has perfused the entire catchment, and sets its mark on it, and though in a more subtle way than in the streams, not less significantly. Water and water-borne material from the catchment end up in streams – this means that all changes that take place in catchments are mirrored in streams. As Noel Hynes (Hynes 1984) put it: The stream is a sort of pulse for the landscape.
The river integrates and reflects the activities occurring upstream in the catchment. Streams connect various structural elements and are conduits for the short and long-term free movement and spreading of terrestrial and aquatic organisms in the landscape. The importance of placing the streams in a landscape context has been emphasised by Décamps (1984). He sees streams as corridors in the landscape connecting the mosaic of habitats, i.e. riparian strips or zones with trees and other vegetation that can cover the whole floodplain, and can even cover parts of the upland. The riparian strips or zones are inherent features of streams that have evolved by disturbances and other geomorphological processes at the stream borders. The riparian zone is the larder of the stream: it delivers an often substantial part of the food for the stream’s invertebrates and, in the end, the fish life, as Hynes (1975) emphasised in his famous paper on the stream and its valley.
148
B.L. Madsen
10.2.1 The Stream in Four Dimensions The interactions between streams and their floodplains are often described in four dimensions (e. g. Giller and Malmquist 1998). They have a longitudinal dimension, and then a lateral and a vertical dimension extending far beyond their channels’ width and depth. And they also have a temporal dimension: they are ever changing. 10.2.1.1 A Longitudinal View Many properties, other than the sheer volume of discharge, change when we follow a stream in a humid region from its sources to the mouth, and many of the changes seem to follow certain patterns that are repeated from one stream to the next. Among the first to identify functional patterns that change in a consistent and predictable way from source to mouth, were fluvial scientists, and their understanding of the fluvial behaviour has dissipated in a fruitful way to stream managers today, most notably through the textbooks of Leopold (1964, 1994), Schumm (1977) and Newbury and Gaboury (1993). That the slope decreases in a downstream direction, while water volume, channel depth and width increases, is obvious. Less obvious, perhaps, is the fact that channel width increases relatively faster than channel depth, and that this can explain part of the behaviour of the different stretches of a stream. Another general but, somewhat surprising, pattern is the increase in mean velocity of rivers downstream in alluvial deposits, despite the more gentle slope. This is explained by the decreasing flow resistance in the downstream direction where a greater volume of water is experiencing less frictional contact with the bed. In some streams, e.g. glacial, with many stones in the lower reaches, the current may be slowed. A high velocity in smaller, high gradient streams may be real, especially at high flows, but the high velocity can be an illusion – caused by the erratic flow pattern over the rough bed material. The progressive downstream increase in a stream’s cross-profile is a function of one specific, powerful discharge event. This channel-forming discharge often has a mean occurrence interval of about twice in 3 years (Leopold 1994). The width dimension responds more to increases in discharge than does the depth. A stream will respond to a consistent change in discharge, e.g. due to change in climate, abstraction of water, etc., by adjusting its shape to the new consistent bankfull event. More frequent flooding leads to more deposition. Therefore there is a tendency towards equilibrium between discharge and floodplain level. While some longitudinal changes are gradual, at least along stretches with no branches, more or less distinct zonal patterns are recognisable. With the identification of predictable relationships between the stream and its climatic-geological and physical settings, various functional classification systems are emerging, based on a stream’s response to disturbance, and on sediment transport processes. A now widely-agreed division is: the upper eroding headwaters; the middle transfer-zones; and the lower depositional zone – the terminology referring to the
10 The Stream and Beyond: Reinstating Natural Functions in Streams
149
most dominant fluvial processes. Schumm (1977) documented that in alluvial channels, i.e. channels that run through the same material as is transported in them, sediment transport modes (suspended or bed load) and width-depth ratios determine bed and bank stability. He recognised three different channel forms: straight, meandering and braided, ranked in decreasing stability. The headwaters, typical small streams in mountains and hilly areas, have the steepest gradients and run in confined, V-shaped valleys (see Fig. 10.1), where they are intimately coupled to hillslope erosion, which can initiate landslides and fallen logs. They respond rapidly to heavy precipitation because of their small storage capacity. Their individual catchment areas are small, but their accumulated impact on the downstream part of the stream system can be immense. Their share of the total catchment area, at least in pristine areas, can make up to 70–80% of the total drainage basin (Gomi et al. 2002). The headwaters merge at lower elevations into streams flowing on more gentle slopes. In this transfer zone sediment transport is the dominating process, and it is here the braiding and meandering begins, and the floodplain is built (see Fig. 10.2). The meandering continues in the depositional zone, where the deposition of sediment is the dominant process. Eventually the meandering changes into several, individual or anastomosing channels over the terminal part with the smallest slope. The recognition that different reaches of a stream respond in their own way to the various fluvial processes, has great value in stream management, and several attempts have been made to develop an operational stream classification system. A promising one is the Rosgen stream classification system (Rosgen 1985; Rosgen and Silvey 1998). It describes not only the physical patterns, but also the behaviour to discharge, not of a catchment, nor a whole stream, but for a restricted stretch of stream. He has recognised seven relevant morphological major stream types, occurring
Fig. 10.1 A Danish headwater stream (Gjessø Bæk) with a confined V-shaped valley
150
B.L. Madsen
Fig. 10.2 A Danish meandering stream (Spørring Å, Å is a Danish word for stream) which has evaded the widespread straightening during the past century
in a wide variety of landscapes from bedrock slopes to low gradient alluvial plains. Rosgen has circumvented the exponential growth of stream types with ever increasing numbers of parameters; by choosing a few, but significant, parameters, and then identifying the most frequent combinations of the parameters and subclasses of bed material, he has restricted the types to the most relevant ones, ending up with some 40 distinguishable stream types, about half of which are rare. To meet local conditions, the fluvial principles in Rosgen’s system can be modified and supplied with other relevant parameters. Mernild and Hasholt (2003) have proposed a classification system as a practical administrative tool for the management and planning of lowland streams and their catchments in a glaciated lowland landscape, where Rosgen’s system alone results in too few types. A key parameter in their system is the hydrology in combination with geomorphology and soil types in the basin. All is expressed in a logical numbering system. We have not met, and probably will never find, a final answer to stream classification, but Rosgen’s system has proved viable in practical river work. It is important to keep in mind that his system meets a major quality in stream classification systems: it has great communicative properties because of its logic, its simplicity and Rosgen’s excellent graphical descriptions. It helps its practitioners to view the stream in a landscape context, to interpret the importance of the channel processes, and to communicate it to stakeholders and decision-makers. But the system has its limitations, especially when it comes to designing the dimensions of stable, alluvial stream channels. This should rely on well-documented equations, based on the known physics of the alluvial stream: with discharge, gradient and bed material grain size as the primary parameters (Leopold 1964).
10 The Stream and Beyond: Reinstating Natural Functions in Streams
151
10.2.1.2 A Lateral View Streams have a lateral impact that – in the end – shapes the landscape. The lateral interactions are linked to the continuous effect of the current’s erosive forces, to the sedimentation and formation of point bars, and to the periodic deposition occurring during flooding of the floodplain. In the headwaters the lateral impact is a one-way affair, with water and material only moving into the stream. However, in floodplain alluvial streams it is a two-way affair, with water and material also periodically moving from the stream channel into the floodplain as well. Many plants and animals exhibit lateral movements too, sharing both habitats; for example, insects with an aquatic larval stage and a terrestrial adult stage (see Section 10.2.2.3). The stream engineers its own floodplain when it erodes the outer (concave) meander bend and adds to the point bars at the inner (convex) bend, which in the end builds the floodplain, supplemented by the depositions during flooding. During their lateral migration, the meanders occasionally erode the valley, thus widening the floodplain and bringing a new supply of coarse sediment into the channel. In the floodplain, oxbow lakes are traces left from abandoned meanders, some of them invisible as subsurface palaeochannels. Both add to the diversity of habitats on the floodplain. 10.2.1.3 A Vertical View The most visible vertical dimensions are the deposits within the floodplains, and the dynamic evolution of pool-riffle structures. But there are more elusive vertical dimensions too. While since ancient times it has been common knowledge that salmon eggs are buried deep in the stream-bottom, and their alevins live their own life there, the existence of a rich habitat extending under the streambed has largely evaded recognition. Often a substantial part of a stream’s water is flowing in this ‘hyporheic’ zone under the visible bottom, into and away from the visible stream and it can extend far beyond the stream channel into the floodplain. It is a subterranean, slow-flowing stream, finding its way through conductive layers, ‘palaeochannels’, e.g. past meanders (Stanford and Ward. 1993). More than water is exchanged between this zone and the visible stream. Stonefly-larvae (Plecoptera) and other invertebrates spend part of their life in the hyporheic zone. Plenty of them have been sampled deep in the floodplain 2 km from the visible stream border. Dissolved and particulate organic matter is transferred out from the valley into the stream via the hyporheic zone, contributing to the food web (see Sections 10.2.3.1 and 10.2.3.2), and during spates the tiny inhabitants in the stream can seek refuge in the interstitials. 10.2.1.4 A Temporal View The recognition of the temporal stream dimension dates deep into the past as can be demonstrated by the saying of Heraclitus as follows: ‘You can never step in the same river twice.’
152
B.L. Madsen
Streams and their floodplains are ever changing. It can be seen at many scales, from seconds to centuries and beyond. The saltating grains on the bed are changing from second to second. Discharge and flooding from season to season, shape the channel within years, the floodplain from century to century. A trained eye can read the history in the floodplain and valley. Past fluctuations in precipitation have set their marks here as 100- or 500-year flooding events. Traces of more persistent climatic fluctuations are left as terraces that are remains of ancient floodplains. They are mute storytellers of past times, when discharges were higher. It must be remembered that the present floodplain is the one constructed by the river in the present climate and which is inundated during present, moderate flow events (Leopold 1964).
10.2.2 A Place to Live 10.2.2.1 Habitats A habitat is a spatial unit: defined as the place where a species successfully survives, grows and reproduces. Although basically defined from the viewpoint of a species, it is often used in a broader sense, describing a restricted area, defined by some conspicuous, characteristic properties and often inhabited by several species. Running water, simply because it is running, is a richer habitat than still water. This statement by Hynes (1970) summarises the uniqueness of streams as habitats for animals and plants. The current not only orchestrates the fluvial processes, but the biological ones too. It creates and maintains a multitude of substrata, offering a stone for a trout to hide under and insects to graze, gravel for salmon to spawn in and insect predators to hunt among, and at the lower end of the current velocity scale it offers mud for worms to browse. The current enhances the uptake of food, nutrients and oxygen, and it has been a driving force for the evolution of ecological adaptations to a life in the water flow. There is an ecological consensus that a great diversity in habitats is a prerequisite for a high biological diversity and species abundance. The principle was formulated early in the last century by Thienemann (1918): Je variabler die Lebensbedingungen einer Lebenstädte, um so grösser die Artenzahl der zugehörigen Lebensgemeinschaft. (With a larger diversity in habitat structure, the more species compose the community).
The principle was finally summarised in his treatise on European freshwater fauna (Thienemann 1950). As Hynes (1970) remarks, this principle has been rediscovered and put forward and presented as a new idea repeatedly – an apparently continuing process. Natural streams and their floodplains have a large gamut of habitats, or a high degree of ‘patchiness’, and these characters are crucial in understanding stream and river ecology (Hildrew 1996). They are, as are many other of the properties of streams, scale-dependent: a stream, a stream segment, a stream stretch with or
10 The Stream and Beyond: Reinstating Natural Functions in Streams
153
without its meadow, and a stream reach, are all habitats. All can be subdivided: a debris dam, a big stone and a riffle are habitats in their own rights. Even a gravel particle, a sand grain, or a space within the biofilm on a stone can be viewed as habitats for some tiny creatures (see Fig. 10.3). The scale chosen depends on the object of study: beavers and otters need habitats that are 2–3 orders of magnitude larger than that of trout, whose habitats are 2–3 orders of magnitude larger than those of many insects. Habitats have a temporal dimension too. Habitats, both in the stream and in the floodplain, are continually changing in response to discharge and flooding events. Also the habitat requirements of their inhabitants change during their various developmental stages. Salmonids need small gravel streams for spawning, successively larger streams for various size classes of young, and the open sea for the adult sea trout and salmon. Stream insect larvae dwell in aquatic habitats, while their adults usually are terrestrial. 10.2.2.2 Stream Habitats A key determinant of the quality of stream habitats is the current – in concert with the material. The current creates and maintains habitats by sorting the sediment. The habitats change in response to the fluvial processes that rule the stream. As fluvial processes change along a stream, habitats will change too. In headwaters, the current erodes, sorts and deposits the material in patches of different substrata in a mosaic structure. Adding to the patchwork are the leaf packets and the debris dams. There is a broad spectrum of current velocities, repeatedly changing from subcritical to supercritical, offering opportunities for a variety of feeding specialists. Steep gradient headwaters seldom have riffles, while pools can develop; they add to the habitat diversity and provide shelter, where migrating fish can rest after passing rapids. Stationary fish can seek shelter here during droughts and during floods, and they find cover against herons and other predators. In floodplain streams, the current cuts the meanders (Fig. 10.4) and shapes the riffle-pool sequences (Fig. 10.5). Meanders, pools and riffles offer a variety of current velocities, substrata and physical shapes. And their patterns repeat over and over again down the stream channel with an interval of about seven times the bankfull width of the stream. The mosaic physical structure promotes a mosaic vegetation of macrophytes in unshaded floodplain streams. Their different growth forms offer a variety of habitats for fish and invertebrates. Further downstream, where the deposition processes take over, the variation in bottom substrata decreases. While in the main channel a fine-grained bottom without macrophytes dominates, the decrease in habitats diversity in the main channel may be more than balanced by new habitats in extensive bank zones and backwaters. This can result in an often very rich biodiversity as the genuine stream organisms are enriched by those favouring stagnant water inhabiting backwaters. A variety of fish other than salmonids inhabit these downstream reaches, which have a rich bird
154
B.L. Madsen
Fig. 10.3 Stream systems can be viewed in many different scales, from the channel segments in the entire catchment into the smallest detail, each with a multitude of habitats and flow conditions (reproduced with permission from Newbury and Gaboury 1993)
10 The Stream and Beyond: Reinstating Natural Functions in Streams
155
Fig. 10.4 Idealised full-wave meander showing the regular sequence of riffles and pools, sites of sedimentation (point bars) and sites of erosion (steep, undercut banks), all adding to the multitude of habitats. The ‘wavelength’ is about 10–14 times the bankfull width (From Madsen 1995)
Fig. 10.5 Even in a straightened channel reach riffle and pool can develop and be recognised mimicking the natural pattern (From Madsen 1995)
life too. A survey of Danish lowland streams (Wiberg et al. 2000) has shown that the number of caddis fly species increases considerably from the headwaters to the lowermost floodplain streams. This gradual downstream increase is probably related to the greater availability of habitats rather than to the size of the streams.
156
B.L. Madsen
10.2.2.3 Riparian Habitats The habitats in the stream cannot be viewed without their riparian surroundings, from the nearest edge to the wide floodplain. Riparian trees along small streams provide food, and their exposed roots are shelter for fish and substrate for invertebrates. Trees also shade the water so it remains cool in summertime. Riparian vegetation is the habitat for most adult stream insects and this terrestrial habitat is as important for the adults as the aquatic habitat is for the larval stages. Here they go through metamorphosis, mate and feed, and some use the vegetation for oviposition. Sufficient qualities in both habitats are vital for the successful life cycle of species (Fig. 10.6). During summer, the shade from overhanging riparian vegetation provides cover for fish. Fish also receive a substantial amount of their diet from terrestrial insects that drop into the water from the vegetation. In smaller, unshaded streams with a gradual transition from the stream channel to meadow, many plants are amphibious, sometimes with two different growth-forms for their aquatic and terrestrial habitat. By sharing these two habitats they can benefit from having the best of both worlds, e.g. the often high content of carbon dioxide in water seeping out from the meadows, the mild winter temperatures in the water, and the unrestricted access to light. The floodplain, growing larger and larger downstream, has a variety of habitats, all created by the fluvial events. The flood-pulses fertilise vegetation, and they fill
Fig. 10.6 Stream insects with a terrestrial stage need suitable habitats in the stream and in the terrestrial surroundings as well (From Madsen 1995)
10 The Stream and Beyond: Reinstating Natural Functions in Streams
157
the depressions with water, so amphibians can breed and waders find food. The recently abandoned meanders continue as oxbow lakes or swamps, adding to the habitat heterogeneity. 10.2.2.4 Biological Engineers Many stream organisms are habitat modifiers – the most spectacular example being the beaver, which radically changes streams and their floodplains with debris dams to an extent exceeded only by man and major natural events. On a smaller, but still substantial scale, salmon and trout engineer their habitats, when they excavate spawning pits in riffles. Fine gravel particles, sand and silt are washed away, leaving coarser particles, so the eggs are buried under clean gravel, pebbles and cobbles with unclogged interstices, facilitating the oxygen supply needed for their eggs (Fig. 10.7). The salmon activity on the bed maintains the porous, hyporheic invertebrate habitats, which are the larder of food for the newly-hatched alevins. Another important implication of salmon modification of stream bottom is that the coarser surface stabilises the reed, because it becomes less susceptible to streambed scouring. In a Pacific gravel-bed stream the critical shear stress on a spawned reed increased by 73% as compared with a non-spawned (Montgomery et al. 1996). If salmon or trout spawning is absent for a few seasons, the riffle particles can cake together to an impenetrable pavement (Madsen 1995). An artificial loosening of hardened riffles, not succeeded by spawning, reverted to the hardened state in 2–5 years. On an even smaller scale, insect larvae can function as biological engineers too. Minute, tubebuilding chironomid midge larvae can increase and modify habitats for microorganisms. In a nutrient-poor Michigan stream, substrata densely covered with midge tubes, supported 7–12 times more diatom biomass than substrata without midges (Pringle 1985), and they influence the sediment transport by increased suspension (Vogel 1994). In shallow, unshaded streams which meander through floodplains, submerged macrophytes are extremely important in creating and modifying habitats (SandJensen 1997). The macrophyte habitats are very dynamic over short time scales
Fig. 10.7 Trout engineers the spawning bed in concert with the current. The shape facilitates a pressure gradient so fresh water can flow through the egg pockets (From Madsen 1995)
158
B.L. Madsen
such as months. High flows can uproot them, and in winter many species vanish. The macrophyte habitats range from a mosaic pattern, with patches of different growth forms alternating with exposed parts of the bottom, to a dense, monotonous carpet of one or a few dominating species such as Canadian waterweed (Elodea canadensis) and submerged burr reed (Sparganium emersum). Such a monoculture is often the terminal effect of a management practice that has changed the natural, fluvial processes. Macrophyte stands offer a diversity of habitats for many microorganisms and invertebrates. The grazers feed on their biofilm (see Section 10.2.3), the filtrators enjoy a place in the free water-flow elevated from the bottom, and salmonid fry find shelter and territory among the amphibious plants. As ecological engineers, plants modify other habitats by changing the flow pattern, attenuating the current velocities – thus facilitating sedimentation and entrapment of detritus inside and in the leeward edges of patches. Some macrophytes with a dense growth form, such as starwort (Callitriche spp.), can deflect and amplify the bypassing current, exposing coarse substrates and eroding the banks, thus shaping the stream course. A marginal band of the amphibious water cress (Nasturtium spp.) and burr reed (Sparganium emersum) can narrow the stream profile, when their roots stabilise the sediment accumulated in the patches. These characteristics firmly place macrophytes as key organisms in the prudent management of many floodplain streams (see Section 10.3.2). 10.2.2.5 Life in the Current The life in streams can be risky, but the benefits pay off. The currents transport matter downstream, sometimes very violently. The challenge of plants and animals is to stay put, and many adaptive mechanisms have evolved to cope with the risks. Macrophytes in the current are flexible, bending and changing form in response to the current, offering less resistance, so they can withstand the strong forces. Black flies anchor themselves by hooks, fastened to a silk pad they have glued onto the substrate. Caddis flies fasten their retreats to the substratum with silk. Many small invertebrates take advantage of the fact that current velocities close to the substratum are much reduced, so they can move around in almost still water (see Fig. 10.3); others have forms that deflect the current (Vogel 1994). But most find shelter against the current; they utilise the wide spectrum of reduced currents that exist in the heterogenous substrata and still benefit from the current. They stay leeward of stones and macrophytes, live in leaf-litter packets, crawl into crevices, or bury themselves in gravel. Trout and other fish hide behind stones, boulders and macrophyte patches. When salmon are negotiating strong rapids and riffles, they rest in the pools. If invertebrates are accidentally moved by the current, a phenomenon called drift, many exhibit a compensatory behaviour. Some move up against the current. A few spin a lifeline. In some insects, prone to drift, the adult females move upstream before the eggs are deposited (Madsen et al. 1973; Macneal et al. 2005). The existence of shelter, or refugia, is crucial for survival during floods and spates with increased sheer forces – and shelters are intimately associated to the
10 The Stream and Beyond: Reinstating Natural Functions in Streams
159
heterogeneity of the stream. Invertebrates can seek shelter in crevices and bury deep into the bottom. Lancaster and Hildrew (1993) have documented that when discharge increased, the relative density of stonefly nymphs (Plecoptera) increased in the parts of habitats with the least flow impact. Fish can survive in backwaters and in the inundated floodplain. Stream sections that have been denuded by a severe spate can be recolonised from the survivors drifting in from more sheltered, upstream parts and emerging from their refuges. Recolonisation can be surprisingly fast (Thorup 1970).
10.2.2.6 Access to Oxygen Oxygen dissolved in the water is a critical factor for the distribution of animal life in fresh water. The oxygen concentration in fresh water (in equilibrium with the atmosphere), is only ca. 14 mg l−1 at 0°C or approximately 20 times less than in the atmosphere. The oxygen concentration is only one side of the harsh conditions in water. The availability of oxygen is much lower than in the air, because the diffusion of oxygen is 10,000 times slower in water than in air. But animals in streams (see Fig. 10.8) have one advantage over those in still waters: the current continuously brings a fresh supply of oxygen to the respiratory organs, creating steep diffusion
Fig. 10.8 The influence of water movement on the lethal oxygen concentration level for a stonefly (Brachyptera risi) exposed to the current on the top of stones. It can survive a lower oxygen concentration in moving water (left) than in stagnant water (right) (From Madsen 1968)
160
B.L. Madsen
gradients close to organisms’ surfaces. This was eloquently formulated by another of the founding fathers of limnology, F. Ruttner (1926): Flowing water is physiologically richer in oxygen than still water.
Usually the groundwater feeding streams is low in oxygen, but oxygen enters the stream water by two pathways. One is by photosynthesis, when macrophytes and algae produce oxygen during daytime. In slow or moderately flowing water this can cause considerable, diurnal oxygen-fluctuations: saturation or supersaturation during daytimes, and values below, sometimes far below saturation level during night-time due to the respiration of all organisms. But the vertical stratification well known from lakes and ponds is a rare exception in streams, where even a slow current mixes the water, and enhances the exchange of oxygen and other gases with the atmosphere. A very important exchange mechanism is the entrapment of air pockets in the transient ‘hydraulic jump’ zone where a supercritical velocity is decelerated to subcritical velocity (Newbury and Gaboury 1993; see Fig. 10.3). They are regular structures in headwaters and steep streams; their characteristic noise is caused by the bubbles released from the air pockets. This fast exchange mechanism ensures that in such streams oxygen concentrations seldom deviate from the air equilibrium. The oxygen concentration at air equilibrium dissolved in water depends on the water temperature: solubility decreases with increasing temperature. And what makes it more critical is that the oxygen consumption of stream organisms increases with temperature, at high temperatures often around 10% per °C. This is to a certain degree counteracted by the increasing diffusion of oxygen at higher temperatures. But as a general rule, oxygen availability is regarded a major constraint for the downstream distribution of salmonid fish and other animals with high oxygen needs as exemplified in the classical division of central European streams with the upper trout zone and the lower barbell and bream zones (Hynes 1970). Temperature changes predictably in the downstream direction. Groundwater entering streams is cool, often approaching the annual mean temperature of the region. Thus the springs and headwaters are “warm” in wintertime and cool in summer relative to air temperature. Solar insulation heats the water along its downstream course, but the warming is dependent on shading, depth and retention time. The heating and cooling is highest in the open floodplain stretches, but with increasing water volume downstream the temperature fluctuations, especially between day and night, are dampened. This predictable temperature pattern – together with slower water-atmosphere exchange, higher photosynthesis and higher decomposition rate – sets the large differences between living conditions in the upper and lower parts of stream system. The combination of cool water during summer, oxygen saturation and fair current velocity makes the upper parts of a stream system a preferred habitat for salmonids and insects that need a high availability of oxygen. Contrary to common belief, oxygen availability however decreases with increasing altitude (Jacobsen 2000), despite the decrease in temperature with increasing altitude. The reasons are that the diffusion of oxygen decreases with lower temperature, and there is less oxygen dissolved in the water
10 The Stream and Beyond: Reinstating Natural Functions in Streams
161
due to the decrease in air pressure with increasing altitude. The distribution of organisms in downstream reaches is restricted by higher summer-temperatures and lower oxygen availability. Species more tolerant of high temperatures and low oxygen availability dominate here; salmonids are succeeded by barbel and by bream. Changes in oxygen availability in streams are not only a longitudinal phenomenon – it is also a consequence of local physical heterogeneity. Good oxygen availability in patches with stones and gravel can occur side by side with lower availability in mud and detritus deposits in patches with slow current and no turbulence (Madsen 1968).
10.2.3 The Larder of the Stream Only a fraction, some 10–30%, of the living macrophytes in terrestrial ecosystems is consumed directly by herbivores while much larger part is consumed in the form of detritus, i.e. leaf-litter and other dead plant remains. The most conspicuous primary producers in floodplain streams are macrophytes, but they have not been considered a significant food source for invertebrates, which is different to some, but not all terrestrial ecosystems. Recent work (Jacobsen and Sand-Jensen 1992), however, has shown that several species of pondweeds (Potamogeton spp.) are consumed by invertebrates to an extent similar to that seen in terrestrial plants, while other submerged species such as waterweed (Elodea spp.) and starworts (Callitriche spp.) are grazed very little. The main importance of the macrophytes in the food web is indirect; their large surfaces are covered with a biofilm of algae and other microorganisms (periphyton, ‘Aufwuchs’) (see Section 10.2.2.4) which are consumed by a multitude of mayfly larvae and other grazers. Forest-bordered headwater streams can also hold primary producers that contribute to the food web, despite the heavy shading during summer. From early spring, before the deciduous forests set leaves, stones and gravel, even sand, can develop a dense, nutritious layer of algae, mostly diatoms, ‘the grass of streams’. They are consumed by invertebrates that ingest the sand, or graze the surfaces. A very significant spring development of a dense cover of algae can be observed in floodplain streams, such as the Danish river Suså prior to the macrophyte shading of the bottom, which had a tenfold increase in algae biomass in May (Sand-Jensen 1997). In some headwater streams from biomes devoid of riparian trees and bush vegetation, diatoms and other minute primary producers are the main or sole source of organic matter. Invertebrates can graze the benthic algae so heavily that they are in control of the algal biomass; when the grazers are removed, there is a spectacular increase in the algae biomass (Kjeldsen 1996). The development of phytoplankton in a free water body is mainly a lake, not a stream, phenomenon, with the exception of some large, slow-flowing rivers and their backwaters. Phytoplankton from lakes, swept into an outlet stream, can also be a very significant food source and support a very rich community of filter feeders in downstream stretches.
162
B.L. Madsen
A biofilm community of microorganisms embedded in a slimy organic layer of mostly polysaccharides, produced by the microorganisms (Lock et al. 1984), covers all surfaces in the water. The community contains a wide variety of microorganisms: fungi, bacteria, algae and ciliates, together with free enzymes. Here is a minute ecosystem with its own nutrient cycles and food chains, in close contact with the water passing by, utilising inorganic nutrients and dissolved and fine particulate organic matter. The biofilm is an important food source for invertebrates that graze the layer. 10.2.3.1 Particulate Organic Matter As in other ecosystems the great majority of primary production enters the detrital food web. But the major part of the detritus input is often allochthonous, i.e. produced by autotrophs outside the stream. This is particularly the case in streams with a forested riparian zone. In macrophyte-rich floodplain streams, the detritus is mostly autochthonous, i.e. produced from within the stream. The importance of an allochtonous food source was emphasised by Hynes in his famous river-valley lecture (Hynes 1975). He had found it increasingly evident (Kaushik and Hynes 1971) that autumn-shed leaves and other plant debris imported from the terrestrial environment played a greater role than the organic matter produced in the stream. Since Hynes groundbreaking experiments, numerous studies have added to this theme. While recent work has produced vast amounts of valuable quantitative details, excellently summarised by Allan and Castillo (2007), the basic processes are largely undisputed. Autumn-shed leaves that fall and blow into streams are conditioned by fungi and bacteria that colonise the leaves. This colonisation of the leaf-litter enriches the leaf-microbe complex with protein, thus providing a valuable food basis for stream invertebrates; the larvae of caddis flies and crane flies eat the leaf material (Fig. 10.9). During this process the invertebrates contribute significantly to its decomposition, because leaves, stems, even wood, are fragmented into smaller particles, which re-enter the water as faeces or particles lost in the shredding process. Other invertebrates, such as the freshwater shrimp (Gammarus), and some stoneflies (Nemoura) graze the nutrition-rich layer of microorganisms on the surface of the leaf-litter. These are called ‘scrapers’ or ‘grazers’. The small particles (FPOM) are re-colonised by microorganisms that are the food for other invertebrates, sharing the name ‘collectors’. These include ‘filtrators’ such as midge larvae (e.g. simuliids and Rheotanytarsus) and caddis larvae (e.g. Hydropsyche) that collect particles suspended in the water, and ‘gatherers’ that browse on the sediment. For the various invertebrate species involved see Fig. 10.10. How important the allochthonous leaf-litter is for the ecology of a stream was demonstrated in a large-scale experiment in a North Carolina stream, bordered by deciduous forest. A 180 m long stretch of a stream was fenced by a net, so no leaf litter entered the stretch from the riparian zone. During the 3-year study, most of the dominant detritivore invertebrates and their predators suffered major reductions in number and biomass (Wallace et al. 1997).
10 The Stream and Beyond: Reinstating Natural Functions in Streams
163
Fig. 10.9 Alder leaf (Alnus glutinosa) before and after being shredded in the stream
Fig. 10.10 Invertebrates from the stream; from left to right: the freshwater shrimp (Gammarus pulex) is a very versatile inhabitant of streams – it is a shredder, a grazer and a predator; the limpet (Ancylus) is a grazer; the caddis larvae (Hydropsyche) is a filtrator; and the alder fly larvae (Sialis) is a predator
A special link between the stream and its riparian zone is the import of terrestrial insects and even small mammals. Terrestrial food is an important food source for grayling (Dahl 1962) and in four Japanese headwater streams it contributed about 50% of the annual salmonid consumption (Kawaguchi and Nakano 2001). 10.2.3.2 Dissolved Organic Matter Dissolved organic matter (DOM) is usually the largest pool of organic carbon in the stream water (Likens 1984). It enters the stream along a variety of pathways, one
164
B.L. Madsen
being rainwater intercepted by the canopy. Another important source of DOM is leaf-litter, which begins leaching as soon as it ends up in the water, undergoing considerable weight loss within a few days (Kaushik and Hynes 1971). But there is evidence that a major part of the DOM seeps into the streams from the soil in the valley, where it is released from microorganisms, from the roots of living plants, and from the decomposition of detritus. Sugars and other labile fractions easily enter the food web, usually via microorganisms, but a substantial part of the DOM is more or less inaccessible to food webs. Dissolved organic matter is also produced by macrophytes and their periphyton within streams. Not much of their quantitative biological impact is known from streams, but it is well documented from lakes, where they can fuel a planktonic, microbial food chain (Søndergaard 1997). No doubt a substantial part of food webs in macrophyte-rich floodplain streams are fuelled by these exudates too. A byway for the DOM into the food web is its transformation to fine organic particles, which seems to be enhanced by turbulent water over riffles and in rapids (Petersen 1986). The particles can be utilised by filtrators and collectors. 10.2.3.3 Patterns in Biological Communities and the Food Web Biological communities in the stream reflect food availability. Gradual change in food availability along the course of a stream from its source to the mouth has given rise to the River Continuum Concept (Vannote et al. 1980). One of the central tenets in the concept is that the feeding habits of the invertebrates reflect the food sources as they change with stream order. The River Continuum Concept applies to an ideal stream system in deciduous forest biomes, beginning with headwater streams, shaded by a dense forest canopy. The system then gradually changes as the stream runs through more open floodplain country, until it ends as a deep, wide river. In the upper part of the stream system the food is allochthonous, derived from the riparian vegetation, and the autotrophic production within the stream is negligible. Shredders and collectors are the dominating invertebrates. As the stream widens and runs into an open floodplain, more light and suitable substrates promote the growth of macrophytes and periphyton, and the food becomes predominantly autochthonous. The proportion of grazers increases relative to shredders. The collectors’ share remains unchanged, because fine, organic particles are available in this part of the stream. As the stream grows larger and deeper further downstream, the grazers are reduced in number or even disappear as the decreased light availability reduces the macrophytes and periphyton on which they feed. The collectors, especially the fine-particle filter feeders, become the dominant part of the community. The allochthonous-derived fine particles transferred from upstream are supplemented by fine particles seeping in from riparian soils (Meybeck 1982), and in some large slow-flowing rivers autochthonous phytoplankton can add to the food web. This proposed distinct, gradual change in food availability along an ‘ideal’ river, causing a gradual change in the composition of the invertebrate community, is in reality much more complex (‘noisy’), because there are other progressive changes than just changes in species and food availability along the stream’s course.
10 The Stream and Beyond: Reinstating Natural Functions in Streams
165
Temperature regimes change (see Section 10.2.2.6) and increasingly more habitats become available in the downstream direction. Snails can graze the periphyton on reed and floating macrophytes, shredders find decaying plant remains in larger streams. Adding to the noise in the concept is the fact that invertebrates’ feeding modes and preferred food sources are not always as clear-cut as assumed in the model. Shredders can feed as scrapers, even predators. Not all headwaters have riparian forests, but macrophytes in the middle floodplain streams can perform a similar entrapment function as debris dams, and the dead macrophytes can deliver detritus, which substitutes the leaf-litter. In its original version the River Continuum Concept neglected the import of allochthonous material in the lower reaches from tributaries, from flooding, and from riparian margins. A study in some small Danish streams has shown that there is no difference in the composition of invertebrate fauna in forest-covered streams compared with streams of similar size in open landscapes (Jacobsen and Friberg 1997). In some regions the River Continuum Concept does not apply because of a special hydrology and geology. In New Zealand, for example, spates can flush the streams of detritus (Winterbourn et al. 1981). Statzner and Higler (1985) suggest that the critical determinants in pattern in invertebrate communities are due to the zonal pattern of hydraulic characteristics, with transition zones distributed along streams from source to mouth. Despite justified criticism, the River Continuum Concept is a valuable and useful tool for the formulation and testing of new ideas in stream ecology. A virtue of the theory is that it can place the biological community in a given part of a stream into the context with the catchment.
10.2.4 More Than a Gutter One contrast between the nutrient cycles in lakes and in streams derives from the fact that autotrophs in a lake compete for the limiting inorganic nutrients in the water, while in streams, at least European lowland streams, such nutrients seldom are limiting for the autotrophs (Madsen 1995). The streams autotrophs take a share of these nutrients while they are transported downhill towards the sea. A stream’s ability to delay the flux of inorganic and organic nutrients is part of a well-functioning stream-floodplain system, so they are not merely flushed through but are utilised, transformed, and deposited transiently or permanently (Fig. 10.11). This capability to delay such fluxes is named the retention capacity, and it is influenced by the biological and physical structures in the stream and its floodplain. 10.2.4.1 The Water Cycle It shall be stressed that the catchment dictates the water inputs to the stream. This was demonstrated in a large-scale project in the Hubbard Brook experiment in New Hampshire (Likens 1984). When a catchment was cleared of trees and its re-vegetation suppressed, the annual runoff from the catchment in the two following years
166
B.L. Madsen
Fig. 10.11 Straightened streams have a reduced retention- and transformation capacity. They just transfer matter downstream. In natural streams matter is not only transferred downstream, but also transformed and utilised by the organisms, or deposited during flooding. Natural streams are more than a gutter
increased by 39% and 28% over the values expected had the catchment not been cleared. During summertime, the increase was respectively 414% and 380% greater than expected (Likens et al. 1970). This clearly demonstrates the impact of evapotranspiration from intact vegetation in the catchment. The catchment retention capacity for water is highly dependent on the vegetation cover and the soil type. In an intact forest floor with a well-developed litter and humic layer, the infiltration capacity is rarely exceeded. Even during a storm, overland flow is uncommon. The water percolates through the loose soil layers, worm- and root holes, idling as groundwater towards the stream. In the case of a storm, part of the water may flow through the upper layer as shallow subsurface flow, which eventually can reappear as overland flow, but on its route to the stream the water has been delayed. In such areas, soil erosion is negligible. The canopy eases the impact of the raindrops, and there is no substantial overland flow which is the prerequisite for erosion (Likens 1984). 10.2.4.2 Nutrient Retention The flux of organic matter and of nutrients in stream channels is not a continuous and uninterrupted process. While the water steadily flows downhill, the nutrients make a stepwise journey with many stop-overs. The nutrients are trapped and transiently stored in organisms, in sediments, in debris dams, in the hyporheal, and some end their journey in the floodplains. Many nutrients have been transformed into different forms perhaps several times. The downstream transport and transient retention of a nutrient atom can be viewed as an imaginary spiral or helix, and is termed ‘nutrient spiralling.’ It combines
10 The Stream and Beyond: Reinstating Natural Functions in Streams
167
both the storage and the transport of nutrients. The spiralling length of an atom is a loop between one point of release to the water phase and the point where it returns after having been stored in the biota. Newbold et al. (1984), applying radio-assay in a small woodland stream in Tennessee, determined a mean spiralling length for a phosphorus atom to be 190 m. In one loop the atom travelled 165 m in the water phase and 25 m incorporated in microorganisms in leaf litter, biofilm, etc., while it travelled only 0.05 m in consumers. This small contribution to the spiralling length reflects the fact that a very small amount of the phosphorus ends up in the consumers but that they have a very good retention capacity. The spiralling length is much influenced by physical structures in the stream channel, and by the consumers, which retain the matter for a shorter or longer period before it re-enters the free-flowing water. Deadwater areas, pools and a rich macrophyte flora, especially if covered by filtrators such as blackfly larvae (Simuliidea), shorten the spiralling length. 10.2.4.3 Physical Entrapment The retention of organic particles and transient storage occur at many spatial scales, from debris dams and macrophytes in a resolution of metres, over caddis nets in resolutions of centimetres, and in the resolution of micrometres within biofilms. The retention occurs in a variable temporal scale as well. Particulate matter stored in sediment and debris dams can be released by irregular fluvial events, or regular seasonal events such as the autumnal senescence of macrophytes. Important retention mechanisms are the organic debris dams made by logs and twigs that trap leaf-litter and other coarse particulate matter – frequently found in headwaters where even small logs and twigs can lock and collect material. How crucial these debris dams are for the functioning of a stream has been documented by experiments: where debris-dams were removed in a 175 m section of a brook in a forested catchment (Likens 1984). The annual export of particulate matter out of this section increased by 530%, and dissolved matter by 6%. In the section with intact debris dams, the export of dissolved matter exceeded the export of particulate matter by a factor of 5, while this was reversed in the section without debris dams to 0.5–1. The soluble organic matter and nutrients, released by living and dead organic material, do not evade retention. They are trapped in the microorganisms in the biofilms on stones, macrophytes and detritus, and they can enter the water again as particulate organic matter when released as faecal particles from invertebrates. The retention capacity of macrophytes in open floodplain streams may be comparable to debris dams, but it is not permanent, because their biomass is reduced in winter. Macrophytes slow down the water flow and promote the sedimentation of particles, and they offer large surfaces for filtrating invertebrates and biofilms. In the macrophyte-rich Danish stream Suså-river approximately 80% of the organic particle transport was retained in the sediment when the vegetation was dense (Sand-Jensen 1997). The accumulated material washed away during the high
168
B.L. Madsen
winter discharge but this does not make the retention in the macrophytes insignificant. The retention during summer prevents the sedimented particles, with their content of phosphorus and nitrogen, from entering downstream lakes and estuaries, where they would have facilitated excess oxygen consumption and algal bloom during that season. That macrophytes and their attached periphyton can retain dissolved nutrients has long been known (Keup 1968), and even in headwater streams with no other macrophytes available than mosses, rapid sorption has been reported (Likens 1984). Intensive Danish studies in a small stream in an agricultural catchment (Kronvang et al. 1999) have documented a significant temporary storage of organic nitrogen and phosphorus in stream sediments during low flow situations from spring to late summer. Net retention was 7.2–16.1 g N m−2 and 3.7–8.3 g P m−2, amounting respectively to 5% and 23% of the gross export from a 113 km2 catchment area. A fraction of the nitrogen export (2.4%) is permanently removed from the stream by denitrification. Most of the stored nutrients were released during high flow. The ecological relevance of the delayed release is that the nutrient flux becomes out of phase with the most productive period and highest demands in downstream lakes and estuaries. 10.2.4.4 Biological Retention Thin biofilms, covering the surfaces in streams, can have significant retention capacities, as has been shown by Battin et al. (2003). During 1 month of the progressive development of a diatom-dominated biofilm in large, experimental flumes, the transient storage of water increased by about 300%, and of particles by 120% in the quiescent zones in the biofilm, thus greatly influencing the energy flow and nutrient transformation. The transfer of water and particles into the biofilm is much facilitated by the biofilm’s convoluted surfaces, microchannels and voids which developed during the experiment. These structures favour advective transfer processes in addition to diffusion. A retention mechanism at a similar microscopic scale is in operation in the initial development of the biofilm on clean, smooth stones (Madsen 1972). Spores from Hyphomycete-fungi were trapped on smooth microscope slides submerged in a Danish headwater stream (Fig. 10.1), and in the developing network of their tiny hyphae, minute leaf fragments were trapped. In their turn, the leaf fragments become the basis of a biofilm, when they are colonised by fungi and other microorganisms. 10.2.4.5 Retention in the Floodplain Zone The catchment is a source of water and matter transferred to a stream, but its floodplain can act as a sink too. A vegetated floodplain acts as a barrier and a filter between the land and stream, which bars or reduces sediment and nutrients carried by overland flow.
10 The Stream and Beyond: Reinstating Natural Functions in Streams
169
Fig. 10.12 A wet meadow rich in organic carbon has a high denitrification capacity (From Madsen 1995)
Floodplains with wet, organic soils have a great capacity for eliminating nitrate (NO3), seeping out from the catchment. Bacteria living in the anoxic zone use the nitrate for the oxidation of organic carbon, and the nitrate molecules are transformed to inert nitrogen molecules, N2 (in the process called denitrification, Fig. 10.12). Danish studies, summarised by Kronvang et al. 2001, have proved that the process can be very efficient, eliminating an average 300–400 kg N-NO3 per hectare in meadows receiving nitrate-rich water seeping out from agricultural fields. Another efficient retention mechanism is the flooding of the floodplain. Water is stored here until it eases its way back to the stream, leaving an often substantial fraction of particulate and dissolved matter permanently, or for a prolonged period, stored. A study at the Gjern River at a stretch where it meanders through a 200–300 m-wide floodplain has revealed the impact such flooding can have on nutrient transport in streams (Kronvang et al. 2002). The study area was a 5,000 m2 floodplain along a 100 m river stretch, where gross and net deposition of suspended particles was measured. During the 1992–1993 winter, three overbank flooding events occurred and due to a special ditch configuration, they were supplemented by two irrigation events. Precipitation and discharge was about the average for the preceding 30-year period. During one 19-day flooding event, where the average residence time for the water was 1 h, the water lost 81% of its suspended material and 40% of its phosphorus. The entrapment efficiency, i.e. the ratio between deposition on the floodplain and the export from the catchment, was approximately 24% and 5% for suspended material and phosphorus, respectively. During the three flooding events within the study period, entrapment efficiency varied depending on the duration and height of flooding. The conclusions from these and other similar studies is that an inundation of even a small part of a floodplain can greatly reduce the transport of sediment and phosphorus out from a catchment, and the implications in abating eutrophication from phosphorus are promising. In the Gjern Å river case, one flooding of a 1-hectare floodplain can retain as much phosphorus as the stream receives from approximately 150 ha of agricultural field.
170
B.L. Madsen
10.3 Stepping Stones Towards Natural Stream Floodplain Relations Few, if any, streams and rivers in Europe have reached modern civilisation unchanged. The flow regime has been changed, often dramatically, with higher floods and smaller minimum flows. The connectivity from source to mouth has been interrupted by dams. Habitats in the streams and in the floodplains have been destroyed. Retention capacities have deteriorated, impoverishing the natural purification system. The challenge for today’s stream-valley restoration is to find a solution so that mankind can live with the rivers without destroying their natural functions, which ultimately benefits man too. For many streams, most notably the larger ones, a major impact on habitats and flow regime, come from dams. The ecological impact of large dams have for many years been subject to investigation, widely reported in the literature, e.g., Ward and Stanford 1995. As with habitats, the impact of dams span a wide range of scales – from huge dams on rivers to small mill dams and weirs on smaller streams. Both large and small dams, however, have in common that they bar fish migration. Danish practice has shown that weirs (see Fig. 10.13) and smaller mill dams can be replaced by riffles, while an alternative at some mill dams is a bypass stream with a discharge sufficient for fish migration (see Fig. 10.14). Using such methods, substantial parts of Danish stream systems have been reopened for migrating fish (Madsen and Debois 2006). But it is crucial that the inlet and outlet can be detected by the migrating fish (Aarestrup and Koed 2003). An additional ecological benefit of
Fig. 10.13 Weirs can be replaced by riffles facilitating free migration for fish and invertebrates
10 The Stream and Beyond: Reinstating Natural Functions in Streams
171
Fig. 10.14 When a dam must be preserved (e.g. for historical reasons), a bypass with a sufficient discharge can be a solution superior to conventional fish ladders
bypasses is that they are often constructed with stones and riffle-pool sequences to provide suitable salmonid habitats. In Denmark, an acceptable minimum flow for bypasses has been arbitrarily defined as a certain fraction of the discharge parameter ‘median-minimum’, i.e. a minimum discharge, below which the discharge will fall on average each second year, calculated from a time series of say 20 years. It is applied when granting licences for water intake to, for example, trout farms, where at present a fraction of at least 50% of the median minimum discharge must be left to the stream bypassing the ponds. Whether this is sufficient for the passage of migrating salmonids in the smaller streams remains a moot point. It is an administrative measure only.
10.3.1 A More Natural Flow Regime The most pressing flow-regime problems are the recent flooding events – those that are catastrophic in e.g. Central Europe. Flooding that puts the riparians and their property at risk is not new: floodmarks are witnesses of catastrophic floods in past centuries. But the magnitude and frequency of the catastrophic floods we see now are new. Our answers to flooding have been to do more engineering work to make the dykes higher and stronger, but that hasn’t tamed the flood. Time and again the
172
B.L. Madsen
streams have burst their man-made constraints. In 1879, Mark Twain said about the taming of the Mississippi: Ten thousand stream commissions, with all the mines of the world at their back, cannot tame the lawless stream, cannot curb or confine it, cannot bar its path with obstructions which it will tear down, dance over, and laugh at.
The truth in this intuitive insight has been obvious in the wake of the more and more violent catastrophic floods, we have seen in recent years. This has forced us to look for other solutions. Dykes are no longer the only answer to flood risk. The plumbing approach – to get rid of the water in a constrained stream channel – relieves some symptoms, but it is no cure. To search for a cure, we must focus on the causes of excess flooding. Before we cleared our forested valleys, water evaporated through a living, functioning vegetation, as was demonstrated by the experiments at Hubbard Brook (Section 10.2.2.1). The surplus water was stored in the voluminous litter layers, and it seldom took the overland flow route down the hill. Before we drained our agricultural fields, and ditched our forests, the excess water from rain and melt water trickled the long and slow way through the soil to the streams. Before we urbanised large areas in the floodplains, the stormwater took the slow route into the stream. This then changed when we constructed storm drains designed to get rid of the storm water fast. Before we straightened and deepened our streams, the torrents lost momentum when they ran over the riffles and through the meanders. Before we hemmed the stream in, the floodwater paused in the flooded meadows. The result of all our alterations in the catchments is an accelerated delivery of water to downstream parts of the stream channel. The causes of the present catastrophic flooding may seem more linked to our treatment of the stream-valley system than to increasing precipitation associated with the hype of a greenhouse effect. We begin to smell a rat: Have we, with all our engineering and plumbing skills, caused our own problems? When we divorced the stream from its floodplain, we removed the natural storage and brake of the water flow. In the natural functioning stream-floodplain system, river water loses volume, and excess forces are relieved when it overflows the banks and is temporarily stored within the floodplain. We are now realising that a key to prevent catastrophic flooding is to make alliances with nature’s own devices, and not rely on dykes alone to keep the flood at bay. The problems downstream must be solved upstream, where there is room to ease the flooding, and where its forces are no greater than we can cope with. The problems must be solved before the floodwater and their forces have accumulated to magnitudes we cannot cope with. The answers are: 1. We must reforest areas along the headwaters to divert a larger part of the water cycle to the atmosphere, and re-establish the water retention capacity in the soils to prevent overland flows. 2. We must delay the water flow in ditches and drains by abandoning them or by inserting retarding ponds.
10 The Stream and Beyond: Reinstating Natural Functions in Streams
173
3. We must rethink the storm plumbing systems in urban areas so they do not flush the stormwater into the streams without a delay comparable to the flooding of the meadows. 4. We must tap the current’s forces by re-introducing pools, riffles and meanders. 5. We must allocate floodplain areas to receive and store floodwaters – until there is room for it in the stream. This insight is a key element in a new European strategy on flooding initiated by the EC Council in the wake of recent catastrophic floods. The basic idea is to look for ‘soft’ ecological solutions rather than ‘hard’ engineering solutions, and to involve the whole catchment area. Details are described at www.ecoflood.org., and in Just et al. (2005). Such projects will not be the end of flooding – as said before, flooding is a natural phenomenon. But the perspectives are that they can be eased, so our necessary engineering work in flood-prone urban areas can withstand them. Going back to nature’s own flood control devices will not make the engineering control devices such as dykes obsolete.
10.3.2 A Better Place to Live While the rehabilitation of streams has traditionally focused on good water quality, it has become increasingly evident that good stream quality includes more dimensions: a sufficient base flow and a high degree of physical heterogeneity. In the dawn of modern Danish stream management and legislation the above three aspects were symbolised in the ‘stream quality trinity.’ Now that the basics have been acknowledged by the administration, the ‘stream quality trinity’ has been extended by two more dimensions: the upstream-downstream continuity, i.e. a free way for upstream and downstream migration, and the freedom for a stream to occupy its riparian areas (see Fig. 10.15). The quality and availability of habitats in the stream and its floodplain is intimately connected to the flow regime. The flooding shapes the floodplain and sets the conditions for its inhabitants. The base flow sets the life conditions in the stream during low-flow seasons, and it is crucial for the passage of migrating fish through bypasses at dams. When the floodplain was severed from its stream, and the meadows were drained and ditched, the wetlands along the stream disappeared and the base flow in the stream during summertime was reduced. Re-installation of good habitats in a stream-floodplain system should focus on a flow regime that approaches a natural one (Poff and Allan 1997), with some focus on regular floods that bring water back to the floodplain, and a base flow that is sufficient to maintain a diverse fish and invertebrate life in the stream. The physical dimension of stream habitat quality has been much in focus in the last 25 years of stream management in Denmark. Almost all of the country’s lowland streams had been seriously altered when they were straightened and deepened (see Fig. 10.13). Meanders and riffles had vanished, and any trace of a natural recovery was suppressed by regular dredging and weed-cutting to maintain the
174
B.L. Madsen
Fig. 10.15 The five crucial dimensions of stream quality: a lodestar for stream restoration
improved water discharge capacity. Fine-grain sediment, which before the straightening was deposited with the floodwater on the floodplain, remained in the stream, where it clogged up and buried the few remaining gravel beds. When high flows are not eased by water flowing over the banks, they can seriously scour the bank and stream bed, thus increasing the amount of sediment, which in the end results in a stream bottom dominated by migrating sand. This trend accelerated through the middle of the last century when agricultural exploitation of wetlands peaked, strongly subsidised by government grants. The poor life conditions in the monotonous streams became visible to the public when the great and expensive effort to clean up the polluted streams showed little or no sign of improvement for stream life. The habitats had vanished, the streams were wearisome gutters. The last 25 years or so have proved that it can be a simple and uncomplicated technical matter to restore lost habitats in these monotonous streams (Madsen and Debois 2006). Restoration has included a few larger projects recreating meandering course of the river such as Gels Å and Brede Å (Nielsen 1995) and Skjern Å (Madsen 2000). But even straightened streams with no re-meandering can regain good habitats when they are restored by a variety of methods, developed by the local stream authorities and their river keepers; examples are: introduction of new spawning beds, substitution of weirs by riffles, reopened culverts, and creation of bypasses at dams and other obstacles for migrating fish. However, far the most widespread improvement of habitats in Danish lowland streams is attributed to a ceased or modified weed cutting, where the river keepers (Fig. 10.16) in concert with the streams own forces mimic the current-channel (thalweg) and maintain a mosaic vegetation pattern known from natural lowland streams (for summary see Madsen 1995). When weed cutting and other kinds of maintenance cease or are reduced, the fluvial processes regain their freedom to reinstate shapes that with time approach the natural ones, governed by the present flow regime.
10 The Stream and Beyond: Reinstating Natural Functions in Streams
175
Fig. 10.16 The river keeper, a key species in maintaining good stream quality in lowland streams. Prudent weed cutting is the most cost-efficient to create and maintain good habitats in streams. The river keeper should work with and not against the stream’s own forces
Improvement of habitats applies to the floodplain too. Among the conspicuous victims of the vanishing wetlands in floodplains are the migrating water birds and geese, and this, in Europe, has paved the way to several projects. In Denmark, chains of habitats have been recreated as ‘stepping stones’ along the main migration route for the huge number of geese and water birds moving between the arctic and temperate or southern Europe each spring and fall. One of the recreated habitats is Skjern Å river project, where large, more or less permanently flooded wetlands have been reestablished by re-meandering the stream (Madsen 2000). In a nearby floodplain of Varde Å river, water has come back when drain tiles and ditches were abandoned (Madsen 2000). Now migrant birds find a rich larder, and other species find breeding places. Not surprisingly, birds rapidly occupy these new wetlands. The natural wetland vegetation, however, is expected to invade the recreated wetlands more reluctantly. One reason is that the former agricultural fertiliser and manure applications have increased the level of nutrients far beyond the natural level in fertile floodplains.
10.3.3 Nature’s Own Purification Plant An objective of catchment management is to reinstate the retention and transformation of the materials that are borne by the water flow. A concerted Danish effort to abate pollution of surface water with phosphorous and nitrate from agricultural sources
176
B.L. Madsen
incorporates a re-establishment of the retention capacity in riparian areas along streams. Buffer zones with grass or natural vegetation along streams in agricultural areas are traditionally regarded as ‘best management practice’ (BMP) to reduce the adverse effect of particulate matter washed with surface runoff into the streams. At all Danish streams of natural origin, a 2 m wide uncultivated buffer strip has been compulsory since 1992, without any economic compensation. How efficient matter retention these narrow margins offer, is still a moot point as water with sediments can leak through animal burrows, etc. Nevertheless the strips can introduce some kind of diversity in the monotonous agricultural landscape, and offer a habitat for the adult stages of aquatic insects. At certain designated streams, e.g. at erosion prone areas, wider and more efficient strips can be established, funded by EEC subsidies, and in 2004 it has been decided in Denmark that some set aside areas, which are agreed by the EEC agricultural policy (CAP), can be established as buffer zones along streams and lakes. By 2009 very few, if any such areas, has been established. 10.3.3.1 Raising the Water Table As mentioned in Section 10.2.4.5, nitrate can be efficiently eliminated if it seeps through anoxic, organic soils, which constitute large parts of many floodplains. In many regions in Europe, however, this capacity has been lost or reduced when floodplains were drained and converted to intensive farmland. The soil becomes aerated because of the lowered water table; however, this is a transient condition. The aerated organic soil is decomposed by oxygen-consuming bacteria at a rate of 1–2 cm soil per year. The life expectancy of a drained meadow with organic soil seldom exceeds a half century, before the ground surface is again approaching the water table, and the soil regains its former wet and anoxic state. This offers a huge potential for a natural re-installation of natural properties in many floodplain areas, where agriculture peaked around the middle of the last century. In Denmark, about 10% of the agricultural land, or approximately 200,000 ha, is prone to such subsidence (Madsen 1995). Denmark has recognised the importance of wet meadows as nitrate sinks, and in order to reduce the nitrate input to estuaries and the sea, the conversion of drained meadows into wet meadows has been accelerated. The water table is raised by cutting field tiles, filling ditches, re-meandering stream channels, introducing riffles and ceasing weed-cutting. The owners are paid compensation. Many projects have been conducted (Hoffmann et al. 2003; Kronvang et al. 2008), but of the expected 16,000 ha only 8,000 ha have been realised so far, and the timeframe far exceeded. Despite the slow pace, the result is an annual elimination of ca. 350 kg of nitratenitrogen per hectare, and the important side effect is an intact stream-floodplain system with improved habitats. Among the floodplain projects accomplished are: a 15-km-long stretch of the Sønderå near the Danish-German border, 5 km of Odense Å on Funen and 4 km of Halkær Å in Northern Jutland (Madsen et al. 2006). New projects are in progress, ranging from some small brooks to the largest rivers as, for example the River Gudenå.
10 The Stream and Beyond: Reinstating Natural Functions in Streams
177
Recreating former wet meadows containing organic soil have a significant potential for abating global warming. When such meadows, in Denmark ca. 200,000 ha, are drained for agricultural use, the organic soil decomposes, and CO2 is released. An annual release of 30 metric tonnes CO2 per hectare has been reported (Nielsen et al. 2009). Restoring the former wet state in the soil will greatly reduce further decomposition of the soil. A serious side effect of an artificially-lowered water table in some floodplains, often in regions which were not covered by ice during the last (Weichel) glacial period, is the oxidation of pyrite in the soil. As the water table is lowered, atmospheric oxygen penetrates the soil; this leads to oxidation of the pyrite, an iron-sulphur compound, and highly toxic dissolved iron and sulphuric acid is released into the stream water, where it is eventually oxidised to ochre. In some catchments in Denmark, entering the North Sea, about 25% of the streams are affected by these compounds to such a degree that they have no or very little fish and invertebrate life. A successful effort to restore these catchments included raising the water table, often accomplished by re-meandering the straightened stream, ceasing weed cutting, and raising the stream bottom with gravel (Madsen 2005, www.okker.dk, Prange 2007). The higher water table prevents oxygen from entering the layers of pyrite, and the iron-sulphur compound is locked firmly in the wet soil as insoluble pyrite. 10.3.3.2 Reinstating Regular Flooding Flooding is the big annual cleanser of stream water and the increased flooding potential along the re-meandered streams and degraded meadows will leave increasing amounts of sand, phosphorous-rich soil particles and ochre on the meadows. Re-instating flooding events can trap the fine sediment including particulate phosphorus on the floodplain, as is documented at the re-meandered stretch of Brede Å stream where 9.1% and 4.4% of the total export of sediment and particulate phosphorus was deposited during one winter’s floods (Kronvang et al. 1998). In some areas with severe ochre problems the stream’s water level is raised by weirs in wintertime, so the meadows are flooded, creating a shallow lake (‘winterlake’). Here dissolved iron is oxidised and ochre sedimented. In summer, the water level is lowered again, so the meadow can be grazed by cattle, mimicking the ancient seasonal shift in the flooding state of meadows. Substantial amounts, often 70–90% of the iron can be cleansed from the stream water in this way (Madsen 1995). Shallow lakes, often termed ‘meadow lakes’, were a common feature in many natural lowland streams, but they vanished when the water table was lowered. In Denmark, some of them are now re-appearing when the water level is raised – their cleansing capabilities for phosphorous and nitrate being well utilised. One example is in the catchment area around the largest lake in Denmark, Arre Lake, in North Zeeland. The lake was heavily eutrophicated by phosphorous from sewage plants and agriculture (through one stream alone came sewage with 90 t phosphorus per year). A big effort to reduce the phosphorous the “hard”, technological way in sewage plants
178
B.L. Madsen
from 90 t P to 10 t P per year had only a minor effect on the status of the lake. It has been estimated that the load to the lake must not exceed 6 t P per year, but this goal cannot be met by further technological cleansing. Reliance on ‘soft’ ecological methods is now preferred: five ‘meadow lakes’ have been re-established in the catchment and these ‘polish’ the treated sewage and take their share of the agricultural phosphorous load, so that the objective for the lake can be met (Madsen 2000).
10.3.4 In Dynamic Equilibrium with Streams In few other places has it been easier for man to subdue nature than in lowland streams and floodplains. But we also have the power and insight to reverse this process and bring nature back there again. Streams are resilient, and they have properties that make them cooperative partners. They can help us to reveal their hidden nature. The properties typical for streams are as follows: 1. One of the peculiar differences between still water bodies and flowing water is that the former are doomed to vanish with time. They change to swamps and bogs and eventually disappear as water bodies. But not so with streams: they are never at rest, they are continually disturbed and rejuvenated by the current and the floods. Abandoned meanders, oxbow-lakes, are left to vanish, but new meanders carve their young track through the floodplain. 2. It is the current in concert with the stream bed and bank that makes streams similar – and different! There is a structured, dynamic variability in the stream bed with regards to pools and riffles. There is an optimised variability in current velocity and flow direction in the meanders. And there is a random variability resulting from a host of heterogeneities of bank and bed conditions (Leopold 1994). 3. Streams are resilient against disturbances. If the term stability should be applied to streams, it must be defined as the ability to persist within a range of natural events, often termed ‘dynamic equilibrium.’ If a natural disturbance is permanent (e.g. a definite change in climate), the streams and floodplain will respond to the new situation with a new dynamic equilibrium. While, for many years, we focused on artificial re-creation of stream habitats, we are now ready to join forces with the stream itself. As White (2002) formulated it: ‘increasingly we see the value in a more natural approach employing the self-regenerative, self-sustaining capacities of streams’. Streams respond also to human disturbances. An equivalent to a climatic disturbance is the man-induced change in flow regime. Big dams have seriously changed the natural function of the downstream channel and its floodplain (Ward and Stanford 1995), and many of such disturbances may be regarded as permanent disturbances, at least on a human time-scale. The widespread human disturbances in the headwaters and the smaller streams and their floodplains, however, need not to be permanent. In straightened and deepened streams, a recovery from the disturbance is at work from the very beginning. When the Skjern Å river channel was abandoned in the year 2000,
10 The Stream and Beyond: Reinstating Natural Functions in Streams
179
traces of riffle and pools appeared in the dry bed (they had evolved in less than 40 years) and a very diverse insect life became established (Wiberg-Larsen et al. 2000). However, the methods of traditional stream management often led to situations where streams were prevented from making their own recovery. Riffles were regularly dredged and banks stabilised to prevent meandering. Macrophytes were cut thus preventing these ecological engineers from raising the water table and assisting the current in reshaping the vanishing varied form. If we leave the streams to their own devices they may eventually recover – but there are constraints. Only streams with their natural flow regime, slope and their natural bed material intact are expected to recover to something like their state prior to human disturbance. It is advisable to have in mind the statement made by Moss et al. (1996): Strictly speaking, we rehabilitate a habitat to acceptability rather than restore to some former state.
This was applied to lakes, but it is often valid in streams too. In streams we can often accelerate the recovery by excavating meanders and supplying gravel and stones. In principle stream restoration is simple, as expressed by the Canadian stream hydrologist Bob Newbury (1995): The real art of stream restoration is to mimic the streams own geometry, but some mystery still remains.
Stream restoration can be a serious affair when property can be at risk. Predictions of consequences are necessary and methods with a sound engineering foundation are advisable. Fortunately, a rich literature on the topic is emerging (e.g. Brookes and Shields 1996). When it comes to smaller streams, the fertile field of experiments that take full advantage of the stream’s powers of self-recovery provide us with the most valuable experience. Here, Steven Budiansky’s (1995) advice to ‘nature’s keepers’ seems appropriate: The best advice is to try. One does not have to be a scientist, or even have a high school diploma, in order to experiment. All it takes is common sense, perhaps backed up by a little intuition.
In the case of these smaller streams, no or little property is at risk while valuable habitats can be reinstated. Here is a key to the flood control of downstream reaches, when water’s natural brake comes to work again. Increasing public awareness has paved the road to many such small projects in many parts of Europe – a good omen for the future, where man and river again should coexist in dynamic equilibrium. And we have a beacon to guide us towards this goal. Francis Bacon wrote in 1620: Nature cannot be ordered about, except by obeying her. Epilogue The world is like a big river that runs along its bed, accidentally puts up sandbanks now here, now there and is forced by these, in turn, into a different course.
180
B.L. Madsen All this happens so nicely and easily and little by little, yet the water engineers find it hard when they try to work against its nature. (Wolfgang Goethe 1807)
References Aarestrup K, Koed A (2003) Survival of migrating sea trout (Salmo trutta) smolts negotiating weirs in small Danish rivers. Ecol Freshw Fish 55:169–176 Allan JD, Castillo MM (2007) Stream ecology. Structure and function of running waters. Chapman & Hall, London, 436 pp Battin TJ, Kaplan LA, Newbold JD, Hansen CME (2003) Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439–442 Brookes A, Shields FD (eds) (1996) River channel restoration: guiding principles for sustainable projects. Wiley, Chichester Budiansky S (1995) Natures keepers. The new science of nature management. Weidenfedl & Nicolson, London, 310 pp Dahl J (1962) Studies on the biology of Danish fishes. The food of Grayling. Meddelelser fra Danmarks Fiskeri- og havundersøgelser 3:199–264 Décamps H (1984) Towards a landscape ecology of river valleys. In: Cooley JH, Golley FB (eds) Trends in ecological research for the 1980s. Plenum, New York, pp 163–178 Giller PS, Malmquist B (1998) The biology of stream and rivers. Oxford University Press, Oxford Goethe W (1807) Goethes Gespräche, Hrsg. von Woldemar Freiherr von Biedermann, Band 1–10, Leipzig 1889–1896, Band 8, S. 298–299 Gomi T, Sidle RC, Richardson JS (2002) Understanding processes and downstream linkages of headwater systems. BioScience 52:905–930 Hildrew AG (1996) Whole river ecology: spatial scale and heterogeneity in the ecology of running waters. Arch Hydrobiol Suppl 113:25–43 Hoffmann CC, Baatrup-Pedersen A, Stamphøj EM, Fabricius K (2003) Re-establishing freshwater wetlands in Denmark. International Conference “Towards natural flood reduction strategies”. Warsaw Hynes HBN (1970) The ecology of running waters. Liverpool University Press, Liverpool Hynes HBN (1975) The stream and its valley. Verh Internat Verein Limnol 19:1–15 Hynes HBN (1984) Running waters and mankind. In: Lillehammer A, Saltweit SJ (eds) Regulated rivers. Proceedings of the second international symposium on regulated streams held in Oslo, Norway, 8–12 August 1982, pp 15–21 Jacobsen D (2000) Gill size of trichopteran larvae and oxygen supply in streams along a 4.000 m gradient of altitude. J N Am Benthol Soc 19:329–343 Jacobsen D, Friberg N (1997) Macroinvertebrate communities in Danish streams: the effect of riparian forest cover. In: Sand-Jensen K, Pedersen O (eds) Freshwater biology: priorities and development in Danish research. Gad, Copenhagen, pp 208–222 Jacobsen D, Sand-Jensen K (1992) Herbivory of invertebrates on submerged macrophytes from Danish freshwaters. Freshw Biol 28:301–308 Just T, Matoušek V, Dušek M, Fischer D, Karlík P (2005) Vodohospodářské revitalizace a jejich uplatnění v ochraně před povodněmi. (Stream restoration and its use in flood protection). Praha, 359 pp (in Czech) Kaushik NK, Hynes HBN (1971) The fate of the dead leaves that fall into streams. Arch Hydrobiol 68:465–515
10 The Stream and Beyond: Reinstating Natural Functions in Streams
181
Kawaguchi Y, Nakano S (2001) Contribution of terrestrial invertebrates to the annual resource budget for salmonids in forest and grassland reaches of a headwater stream. Freshw Biol 46:303–316 Keup LE (1968) Phosphorous in flowing water. Water Restor 2:373–386 Kjeldsen K (1996) Regulation of algal biomass in a small lowland stream: field experiments on the role of invertebrate grazing, phosphorous and irradiance. Freshw Biol 36:535–546 Kronvang B, Svendsen LM, Brookes A, Fisher K, Møller B, Otttosen O, Newson M, Sear D (1998) Restoration of the rivers Brede, Cole and Skjerne: a joint Danish and British EU_LIFE demonstration project, III: Channel morphology, hydrodynamics and transport of sediment and nutrients. In: Hansen HO, Boon PJ, Madsen BL, Iversen TM (eds) River restoration. The physical dimension. Special issue of Aquatic Conservation. Mar Freshw Ecosyst 8:1–264 Kronvang B, Hoffmann CC, Svendsen LM, Windolf J, Jensen JP, Dørge J (1999) Retention of nutrients in river basins. Aquat Ecol 33:29–40 Kronvang B, Jensen JP, Hofmann CC, Boers P (2001) Nitrogen transport and fate in European streams, rivers, lakes and wetlands. In: Follet RF, Hatfield JL (eds) Nitrogen in the environment: sources, problems and management. Elsevier Science BV, The Netherlands, pp 183–206 Kronvang B, Falkum Ø, Svendsen LM, Laubel A (2002) Deposition of sediment and phosphorus during overbank flooding. Verh Int Verein Limnol 28:1289–1293 Kronvang B, Andersen HE, Børgesen C, Dalgaard T, Bøgestrand J, Blicher-Mathiesen G (2008) Effect of policy measures implemented in Denmark on nitrogen pollution of aquatic environment. Environ Sci Policy 11:144–152 Lancaster J, Hildrew AG (1993) Flow refugia and the microdistribution of lotic macroinvertebrates. J N Am Benthol Soc 12:385–393 Leopold L (ed) (1964) Fluvial processes in geomorphology. W.H. Freemann, San Francisco, CA, 522 pp Leopold LB (1994) A view of the river. Harward University Press, Cambridge Likens GE (1984) Beyond the shoreline: a watershed-ecosystem approach. Verh IntVerein Limnol 22:1–22 Likens GE, Bormann FH, Johnson NM, Fisher DW, Pierce RS (1970) Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed ecosystem. Ecol Monogr 40:23–47 Lock MA, Wallace RR, Costerton JW, Ventulli RM, Charlton SE (1984) River epilithon: towards a structural-functional model. Oikos 29:1–4 Macneal KH, Peckarsky BL, Likens GE (2005) Stable isotopes identify dispersal patterns of stonefly populations living along stream corridors. Freshw Biol 50:1117–1130 Madsen BL (1968) The distribution of nymphs of Brachyptera risi and Nemoura flexuosa in relation to oxygen. Oikos 19:304–310 Madsen BL (1972) Detritus on stones in small streams. Mem Ist Ital Idrobiol 29 Suppl:385–403 Madsen BL (1995) Danish Watercourses. Ten years with the new watercourse act. National Agency of Environmental Protection, Copenhagen Madsen BL (2000) Brave new nature. Nature management in Denmark at millennium. Ministry of Environment, Denmark Madsen BL (2005) Ochre, A watercourse problem we can deal with. (www.okker.dk) Madsen S, Debois P (eds) (2006) River restoration in Denmark – 24 examples. Storstrøm County, Copenhagen, 97 pp Madsen BL, Bengtsson J, Butz I (1973) Observations on upstream migration by imagines of some Plecoptera and Ephemeroptera. Limnol Oceanogr 18:678–681 Madsen BL, Boon PJ, Lake PS, Bunn SE, Dahm CN, Langford TE, Zalewski M (2006) Ecological principles and stream restoration. Verh Int Verein Limnol 29:22045–2050 Mernild SH, Hasholt B (2003) A test of classification and regionalisation of Danish watercourses. Dan J Geogr 103(2):13–27 Meybeck M (1982) Carbon, nitrogen, and phosphorus transport by world rivers. Am J Sci 282:401–450
182
B.L. Madsen
Montgomery DR, Buffington JM, Peterson NP, Schuet-Hames D, Quinn TP (1996) Stream bed scour, egg burial depth, and the influence of salmonid spawning on bed surface mobility and embryo survival. Can J Fish Aquat Sci 53:1061–1070 Moss B, Madgwick J, Phillips G (1996) A guide to the restoration of nutrient enriched shallow lakes. Environment Agency, Broads Authority, Norwich, 180 pp Newbold JD, Elwood JW, O’Neil RV, Sheldon AL (1984) Phosphorus dynamics in a woodland stream ecosystem. Bioscience 34(1):43–44 Newbury RW (1995) Rivers and the art of stream restoration. In: Costa JE et al. (eds) Natural and anthropogenic influences in fluvial geomorphology. AGU Geophysical Monograph, 89, Wolman Volume. Washington, DC, pp 137–149 Newbury RW, Gaboury MN (1993) Stream analysis and fish habitat design. A Field Manual. Newbury Hydraulics Ltd, Gibsons, BC Nielsen MB (1995) Restoration of streams and their riparian zones – South Jutland, Denmark. In: Eiseltová M, Biggs J (eds) Restoration of stream ecosystems – an integrated catchment approach, vol 37. IWRB, Slimbridge, pp 30–44 Nielsen O-K, Winther M, Mikkelsen MH, Gyldenkærne S, Lyck E, Plejdrup M, Hoffmann L, Thomsen M, Fauser P (2009) Projection of Greenhouse Gas Emissions 2007 to 2025. National Environmental Research Institute, Denmark – NERI Technical Report no. 703, 211 pp Petersen RC (1986) In situ particle generation in a southern Swedish stream. Limnol Oceanogr 31:432–437 Poff N, Allan JD (1997) The natural flow regime. BioScience 47:769–784 Prange H (2007) Ochre pollution as an ecological problem in the aquatic environment. Solution attempts from Denmark. Edmund Siemens Stiftung, Hamburg Pringle CM (1985) Effects of Chironomid tube building activities on stream diatom communities. J Phycol 21:185–194 Rosgen DL (1985) ‘A stream classification system’, In: Riparian Ecosystems and Their Management, Interagency North American Riparian Conference, Gen. Tech. Rept. ROM-120, pp 91–95, Rocky Mt. For. and Range Expt. Sta., USDA Forest Service, Fort Collins, CO Rosgen D, Silvey L (1998) Field Guide for stream classification. Wildland Hydrology Books, Pagosa Springs, CO, 81147 Ruttner F (1926) Bemerkungen über den Sauerstoffgehalt der Gewässer und dessen respiratorischen Wert. Naturwissenschaften 14:1237–1239 (in German) Sand-Jensen K (1997) Macrophytes as biological engineers in the ecology of Danish streams. In: Sand-Jensen K, Pedersen O (eds) Freshwater biology: priorities and development in Danish Research. Gad, Copenhagen Schumm SA (1977) The fluvial system. Wiley, New York Søndergaard M (1997) Bacteria and disolved organic carbon in lakes. In: Sand-Jensen K, Pedersen O (eds) Freshwater biology: priorities and development in Danish research. Copenhagen, Gad, pp 138–161 Stanford JA, Ward JV (1993) An ecosystem perspective of alluvial rivers: connectivity and the hyporheic corridor. J N Am Benthol Soc 12(1):48–60 Statzner B, Higler B (1985) Questions and Comments on the river continuum concept. Can J Fish Aquat Sci 42:1038–1044 Thienemann A (1918) Lebensgemeinschaft und Lebensraum. Ein Vortrag. Naturwissenschaftlichen Wochenschrift. G. Fisher Verlag, Jena (in German) Thienemann A (1950) Verbreitungsgeschichte der Süsswassertierwelt Europas. Die Binnenge wässer, Band XVIII. Stuttgart (in German) Thorup J (1970) The influence of a short term flood on a spring brook community. Arch Hydrobiol 66:447–457 Vannote RL, Minshalll GW, Cummins KW, Sedell JR, Cushing CE (1980) The river continuum concept. Can J Fish Aquat Sci 37:130–137 Vogel, S. (1994). Life in moving fluids. Princeton University Press Wallace JB, Eggert SL, Meyer JL, Webster JR (1997) Multiple trophic levels of a forest stream linked to terrestrial litter inputs. Science 277:102–104
10 The Stream and Beyond: Reinstating Natural Functions in Streams
183
Ward JW, Stanford JA (1995) The serial discontinuity concept: extending the model to floodplain rivers. Regulated rivers. Res Manage 10:159–168 White RJ (2002) Restoring streams for salmonids. Where have we been? Where are we going? In: O’Grady M (ed) Proceedings of the 13th International Salmonid Enhancement workshop, Westport, Ireland, p 1–31 Wiberg-Larsen P, Brodersen KP, Birkholm S, Grøn PN, Skriver J (2000) Species richness and assemblage structure of Trichoptera in Danish streams. Freshw Biol 43:633–647 Winterbourn MJ, Rounick JS, Cowie B (1981) Are New Zealand ecosystems really different? N Z J Mar Freshw Res 15:321–328
Chapter 11
Floodplain Restoration of Large European Rivers, with Examples from the Rhine and the Danube Erika Schneider
Abstract This paper emphasises the relevancy of floodplains and their manifold functions. It shows that the European river floodplains have suffered dramatic changes over the years as a result of man’s interference. The straightening of the river, the floodplains cutting-off from the river by dykes, and drainage for different purposes have led to disturbances and alterations of the complex framework of ecological conditions in floodplain ecosystems. Moreover, all this has caused a radical decrease in the site-specific species and habitat diversity. The loss of floodplains, its consequences and basic principles for the restoration of functioning floodplain ecosystems are described using the two largest European rivers, the Rhine River and the Danube River as an example. The rising risks of floods and the necessity to provide retention areas to defuse the flood risks have brought about a change of minds. It has led to the development of integrated, large-scale concepts for the Rhine and the Danube rivers management which eventually has materialised in concrete projects dealing with flood protection through floodplain restoration. But aside from flood protection measures, an increasing number of projects aiming at the re-establishment of near-natural floodplain ecosystems is being implemented. In these projects priority is given to the manifold functions of the floodplains and their sustainable use. A number of restoration projects implemented on the Rhine and the Danube with special regard to the Lower Danube are presented. The knowledge acquired in these projects may be used elsewhere. Keywords Floodplain functions • Floodplain restoration • Human impact • Loss of floodplains
E. Schneider (*) KIT Karlsruhe Institute for Technology – University of Land Baden-Württemberg, Chair of the WWF-Institute for Floodplains Ecology, Josefstrasse 1, 76437 Rastatt, Germany e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_11, © Springer Science+Business Media B.V. 2010
185
186
E. Schneider
11.1 Introduction Floodplains – the low-lying areas along rivers subject to periodical flooding – are closely coupled to their river channels and as such are functionally interdependent. While they are usually clearly distinguishable from the surrounding landscape, their character and widths vary in different sections. The functional characteristics of floodplain ecosystems are dependent on the hydrological regime, the main factor being the fluctuation of water levels, i.e. the variations between high and low extremes reflected in the dynamics of the river discharge (Dister 1985, 1995; Schneider 2002a). A whole complex of ecological effects are determined by the hydrological regime, such as the input of nutrients into the floodplain, morphodynamical processes including the dynamics of soils and sediments, mechanical stresses produced by water, ice, bedload and sediments, the groundwater table dynamics, and the exchange and drift of organisms and diaspores (the reproductive units of plants). The dynamics of plant and animal communities all depend upon these various factors (Dister 1994, 1995; Schneider 2002a). The duration, height, seasonal occurrence and frequency of flooding bring changes to floodplains as well as determining ecological processes and the position of habitats along ecological gradients. In a natural floodplain, one characterised by dynamic processes, a whole range of microhabitats can be found – from permanently water-covered, permanently wet and temporarily wet, to dry and extremely dry habitats, with coarse to fine-grained sediments; together these various habitats support a large range of species (Dister 1985; Schneider 1996). The distribution of a floodplain’s vegetation, in its cross and longitudinal profile, also corresponds to the river’s dynamics and consequently to the size of deposited sediments (cf. Dister et al. 1989; Dister 1995; Schneider 2003). The complexity of processes occurring in natural floodplains and their interactions with the river result in floodplains fulfilling many important functions: Hydrological functions • • • •
Water storage/flood retention/moderation of flooding Sediment transport/balancing of coastal land-loss Groundwater and stream water recharge Maintenance of short water cycle / improvement of climate
Biogeochemical functions • • • •
C/N/P cycling Nutrient retention and recycling Retention of sediment and toxic substances (pesticides, heavy metals) Transformation of organic and inorganic pollutants
Ecological functions • Habitat for plants and animals (spawning, feeding, nesting, etc.) • Reservoir of biodiversity, storage of genetic resources
11 Floodplain Restoration of Large European Rivers
187
• Bio-corridor • Bio-productivity Thanks to these natural functions, floodplains and river deltas provide important resources and values that are useful to mankind. These include plant and animal resources with their direct utility value for fishery, forestry, hunting, agriculture, etc., and also floodplains’ indirect utility values, such as flood protection, groundwater supply, nutrient and toxicant retention, and regional climatic stabilisation. Furthermore, there are the more idealised values like the aesthetic value of landscape used for recreation and ecotourism from which local people can benefit (WWF-DCP and WWF-Auen-Institut 1999; Schneider 2002b). These resources, values and their use, taken together, constitute important socio-economic functions performed by floodplains and deltas (Schneider 2002a, b).
11.2 Human impact on Floodplains – The Case of the Rhine and the Danube Over human history, remarkable changes in Europe’s floodplain ecosystems have occurred, most of which have occurred over the last 150 years. These changes are mainly the result of hydraulic engineering measures that destroy or dramatically affect the entity of the river and its floodplain, often fundamentally transforming the floodplain ecosystem. The Upper Rhine represents a dramatic but not untypical example – a mere 6% of its former floodplain on the section between Basel and the Iffezheim dam has survived. Also the Danube, especially the Upper Danube in Baden-Württemberg, Bavaria and Austria up to Vienna, shows dramatic examples of human interventions with profound consequences for its floodplain.
11.2.1 The Rhine – From a Wild ‘Free’ River to Regulated Shipping Way Up until the early nineteenth century, the Rhine constituted a wild, free-flowing river. In its natural state (Fig.11.1), the upper part of the Rhine between Basel and Mainz (Fig.11.2), referred to as the Upper Rhine, showed three distinctive sections: • A braided zone on the stretch with the steepest slope and coarsest sediments • A meandering zone with a gentle slope and fine sediments • A transition zone between these two sections In the braided zone, the Rhine flowed in numerous ramified branches, both large and small, that constantly changed and displaced the river channel, and included islands of varying size; the floodplain areas extended to a width of 2–3 km, many characteristic species occurring in these habitats.
188
E. Schneider
Fig. 11.1 The wild Upper Rhine. Painting by Peter Birmann (1758–1844) – View of the Rhine from the Istein Cliffs. Kunstmuseum Basel
In the meandering zone, up to its confluence with the left-side tributary, the Lauter, the Rhine had a gentle slope and formed large meanders in a wide riverbed. During floods, the river could easily extend over the whole floodplain area, no major obstacles impeding its flow. During high floods, the wild Rhine represented a great danger for those people who chose to inhabit its plains. First regulation of the Rhine was performed in 1817–1880 following the plans of J. G. Tulla. Many small, ramified branches were joined together to form one main river channel with a width of 200–240 m, while in the meandering zone large meanders were cut off from the river. In this way the present large channel of the River Rhine was formed, whilst the Upper Rhine’s length was shortened by 80 km. Tulla’s engineering measures resulted in new settlement areas and agricultural land being gained, while those people living in the Rhine valley had higher protection against regular floods. However, the water retention area of the whole Upper Rhine from Basel to Bingen (downstream of Mainz) was reduced – following newer calculations based on geoinformatic data – from 2,038 to 296 km2 – the loss being about 1,742 km2 (Table11.1). In the early twentieth century, M. Honsell continued Tulla’s water engineering works. By putting up stone heaps along the banks of the Rhine – the so-called “groins” – the flow of the river was narrowed and the Rhine’s cross-section diminished. These measures were the beginnings of a permanent navigable waterway up to Basel. These engineering works resulted in major floodplain areas of the Upper Rhine being lost, and the frequency of flooding, which had a vital impact on adjacent areas, being reduced. Characteristic habitats and species that require regular floods and certain morphodynamics were dramatically reduced or even eliminated. An evolution from temporarily flooded areas into habitats that do not require regular
11 Floodplain Restoration of Large European Rivers
189
Main Mainz
Main
Worm Mäander
zone
s
Main
z
Mannheim
in
Ma
km 450 Worms
Germers
heim
Mannhei
m
ck Worms
km 400
Karlsru
he
Mainz
Ne
Germersh
eim
Rasta
szone
Strasbourg
Mannhei
m
Karlsruhe km 350
Ne
cka
r
Germersh
eim
Rastatt
Karlsruh
Furkation
e
km 300 Strasbo urg 1974
Breisach km 250 Breisach
Basel
1970
N
Basel
Vollaus
bau
Strasbourg
1967
Schlinge nlösung
1963 1961
km 200
Rastatt
1997
1959 1956 1952 1932
Breisach
Seiten kanallösung
Basel
Fig. 11.2 Situation of the Upper Rhine. From the left: before Tulla’s rectification in 1817, after the Rhine rectification in 1878, and following the ‘modern development of the Upper Rhine’ in 1977 (From Dister 1986)
flooding could be observed over broad areas. Despite these changes, major parts of the floodplain were still exposed to the dynamics of the river and could thus maintain
190
E. Schneider Table 11.1 Comparison of extent of morphological and recent floodplain along the Rhine (cf. Günther-Diringer 2003). The River Rhine’s total length is 1,320 km and its catchment covers 220,000 km2 Morphological Recent floodplain River stretch floodplain (km2) (km2) Loss (%) Total Rhine 8,617 1,483 83.79 Upper Rhine 2,038 296 85.5 Central Rhine 73 56 23.39 Lower Rhine 8,506 1,132 82.61
a near-natural aspect. Connectivity between upstream and downstream habitats still remained and parts of the floodplain were directly connected to the river. The Treaty of Versailles (signed in 1919 after World War I) provided a basis for further, systematic development of the Rhine, France being conferred the right to divert water from the river and produce hydroelectric power. Between 1928 and 1977, ten dams were built on the Upper Rhine, the last one being the Iffezheim hydropower plant near the town of Rastatt. Upstream of the Iffezheim dam (river km 334), floodplains are completely cut off from regular floods. As a result of modern development on the Upper Rhine’s southern section, 130 km2 of inundation area, i.e. 60% of the area still available in 1955, has been lost (Dister 1990). At present, upstream of the Iffezheim hydropower plant, two areas (90 km2), constructed flood retention polders with inlet and outlets are submitted to controlled flooding during high flood events, while downstream some floodplains, e.g. the Nature Reserve “Rastatter Rheinaue” (floodplain of Rastatt), a recent floodplain connected to the dynamic of the river, are submitted to regular, natural floods. However, other large areas are cut off from regular floods. On the northern Upper Rhine between Worms and Mainz, 300 km2 have been cut off from regular flooding; only 95 km2 remain under the river’s hydrological regime, but of this area, about 38 km2 have been used as agricultural polders (cf. Mock et al. 1991). The loss of floodplains along the Rhine is clearly visible if we compare the area of the original morphological floodplain to that of the remaining recent floodplain (Table11.1). Very high floodplain losses are particularly significant on the Upper and Lower Rhine. Along the Middle Rhine the extent of floodplains is naturally limited by the mountains of the Rhine Highlands (Rheinisches Mittelgebirge). To illustrate the historical development of the Rhine valley landscape, one can compare topographical maps from different periods, such as, for example, from the floodplain near Rastatt from 1838 (see Fig.11.3), 1872, 1940 and 1989 (see Fig.11.4). The reduction of open gravel and sand banks can be observed in the first period after Tulla’s straightening of the Rhine; during the same period water surfaces were reduced (compare the bars for the years 1838 and 1872 in Fig.11.5). This was caused by the lacking river dynamics and accelerated aggradation processes in a number of waters. A renewed increase in water bodies was the result of gravel extraction, allowing broad flooded extraction pits to emerge in the Upper Rhine’s landscape. As for the forest areas, an initial slight increase can be detected after the Tulla’s measures, as forests were also able to develop on newly-emerged land areas.
11 Floodplain Restoration of Large European Rivers
Fig. 11.3 The floodplain of Rastatt (middle stretch of the Upper Rhine) in 1838
191
192
Fig. 11.4 The floodplain of Rastatt (middle stretch of the Upper Rhine) in1989
E. Schneider
11 Floodplain Restoration of Large European Rivers Area [ha]
193
Cultivated land
2500 2250 2000 1750 1500 1250
Forest Open Water Sand banks Gravel banks
1000 750 500 250 0 1838 1872 1940 1989
1838 1872 1940 1989
1838 1872 1940 1989
Fig. 11.5 Changes in the surface of water bodies, forests and cultivated lands within the floodplain of Rastatt (middle stretch of the Upper Rhine); based on information from topographical maps from different periods (1838, 1872, 1940,1989) (From Günther-Diringer and Musall 1989)
Later on, forest areas were transformed into agricultural lands, resulting in less forest area in favour of grassland and cultivated land (Fig.11.5). Subsequently, the forest areas grew again thanks to afforestation (Günther-Diringer and Musall 1989). Moreover, changes took place within existing forest areas, near-natural floodplain forests being transformed into poplar monocultures. The maps also show a clear expansion in settlements (compare Figs.11.3 and 11.4), mainly occurring in those areas where flooding no longer took place. Comparable landscape development can also be seen by studying maps of other stretches of the Upper Rhine. The most dramatic changes have occurred in the area of the southern and central Upper Rhine below the Iffezheim hydropower plant. The former floodplain and its specific plant communities of softwood and hardwood floodplain levels have shifted towards communities that are now hardly influenced by changing groundwater levels (Hügin 1981; WWF-Auen-Institut 1993). The softwood stands of white willow that were characteristic of the functioning floodplain have suffered a remarkable degradation or even disappeared. Given that morphodynamic processes no longer occur and that no new sediment banks have emerged, the natural regeneration of softwood stands is no longer possible. Species characteristic of dynamic stands such as Myricaria germanica, Hippophae rhamnoides, Salix eleagnos have disappeared from the Upper Rhine. Moreover, communities of ephemeral therophytes that, together with their specific fauna, settle in water channels and on old river branches during low water levels, and are subject to the water dynamics, are strongly reduced as a result of the floodplain loss and river embankment (Table11.2). But the loss of floodplains is not only reflected in the degradation of habitats and the loss of both biodiversity and ecological functions. An old problem has arisen
194
E. Schneider
Table 11.2 Changes in floodplain habitats (between the years 1825 and 1980) along the Upper Rhine between Basel and Neuburgweier (right side of river, 13,600 ha) showing potential natural vegetation (After Hügin 1981) Plant community/year 1825 1980 Permanent water bodies, partly with macrophytes 2% 5% Shallow waters, periodically for short time water free with Limosella and 2% – Polygonum mud-communities (Nanocyperion) Floodplain typical reed communities 2% – Changed reed communities following the development – 2% White Willow (Salicetum albae) forests 12% 3% Oak-Elm floodplain forest (Querco-Ulmetum) with Cardamine pratensis 30% 4% Oak-Elm floodplain forests (Querco-Ulmetum) with Hedera helix 50% 5% Black Alder (Alnus glutinosa) swamp-forests – 1% Black Alder-Ash forests (Alno-Fraxinetum) – 2% Elm-Oak-Hornbeam (Ulmo-Querco-Carpinetum) with Ficaria verna – 7% Typical Elm-Oak-Hornbeam forest 2% 27% Elm-Oak-Hornbeam forest with Carex alba – 38% – 6% Oak forest (Quercetum roboris) with Carex alba
again – that of the risk of floods. The extensive loss of floodplains and the acceleration of the flood peak caused the floods reaching dangerous dimensions during the last two decades. Presently, it takes the flood peak 30 h to move from Basel to Karlsruhe compared to 65 h it took before the modern development of the Rhine. Moreover, before 1955, flood peaks of tributaries reached the Rhine before the flood peak of the main river arrived; nowadays, the flood peaks of both tributaries and Rhine coincide (cf. Dister 1990). This coincidence of flood peaks, together with the above-mentioned loss of retention areas, has dramatically increased the probability of high and catastrophic floods along the Rhine.
11.2.2 The Danube – Changes on the River with Extensive Floodplains Until the early nineteenth century, the Danube was mostly a near-natural river with floodplains of varying width that were subject to periodical rhythm of flooding and drying out. Its morphological dynamics was closely related to its hydrological dynamics. On the Upper Danube, from its spring to the ‘Door of Devin’ or Porta Hungarica (the Hainburg hills), a considerable bed load supply used to arrive from its alpine tributaries. As a result, the Upper Danube was inclined to form a widely ramified river bed spread over a broad alluvial basin. The river bed consisted of main and secondary channels and changed with almost every flood event. Huge talus cones had formed in the estuaries of the alpine rivers where characteristic pioneer species had settled according to the substrate’s grain size.
11 Floodplain Restoration of Large European Rivers
195
The Central Danube (downstream of Bratislava to the defile in the Carpathian Mountains at the Iron Gate) had characteristics of a lowland river with a small slope and a rich load of suspended matter. Two tributaries, the Sava and Tisza, were the main contributors to the Lower Danube’s supply of bed load and suspended matter. In the Lower Danube, with its average slope of 0.05 per mill, suspended sediments and suspended load led to the creation of numerous ramifications and islands (called ‘Ostrov’) consisting mainly of fine sediments, and also to the damming of tributaries with a low transport capacity, forming riparian lakes and so-called river ‘limans’ (Reg. Zus. Donauländer 1986; Schneider 1991). Initial river engineering work started on the Upper Danube already in the early nineteenth century. In Austria, the first Danube regulation measures were undertaken in 1882 for flood protection purposes and to ensure the river’s navigability. The great Danube regulation works of the late nineteenth century, the construction of flood protection dams, low water regulation and the later construction of an almost uninterrupted chain of hydropower plants on the Upper Danube (Fig.11.6) left little of the original character of the Danube and its floodplains. On the Central Danube, regulation works and the drainage of broad floodplain areas also occurred during the nineteenth century. For the better use of the Lower Danube as a waterway, the rapids situated next to the Iron Gate had been blasted during the second half of the nineteenth century. To ensure navigability for large freighters, the Danube bed was dredged immediately upstream of the delta, between Braila and the delta’s first ramification, the Ceatal Izmail. Moreover, in the late nineteenth century, a great cut-through was completed on the Sulina branch in the delta (near to Crisan village). River-bed dredging has also been performed on other sections of the Lower Danube. However, all these measures were not as grave in their consequences for the floodplain ecosystem as were the measures taken with regard to developing large areas for agricultural purposes in the early 1960s (Gâstescu et al. 1983). The first smaller dykes date back to the late nineteenth and early twentieth century. Dykes, with a low profile that could be overflowed, were constructed in the 1930s and 1940s to protect cultivated lands (Schneider 1991). With the land-use plan for the Danube floodplains and the delta put into force in 1962, the large-scale cutting-off of the floodplains from the river began. The objectives were the transformation of inundation areas into arable land, and also partly intensive fish-farming, rather than flood protection, given that the settlements of the Lower Danube are mainly situated on its terraces. The development of the Danube upstream of Vienna, and even of some of its tributaries, has resulted in the almost total absence of bedload transport and supply: sites with their characteristic dynamics and habitats have thus been dramatically reduced and or even erased. Along the whole Upper Danube, a mere 95 km2 of recent floodplain area have remained, the major part of it situated in the National Park Danube Floodplains, east of Vienna.
Fig. 11.6 The Danube River with a chain of weirs and hydropower plants
196 E. Schneider
11 Floodplain Restoration of Large European Rivers
197
Along the German Upper Danube, only small stretches still exist. In BadenWürttemberg, only 12 natural stretches, each a mere kilometre in length, have remained untransformed (Konold and Schütz 1996). As for the Bavarian Danube, along the last free-flowing stretch of 70 km between Straubing and Vilshofen, only a few areas of the recent floodplain have been preserved (Weiger 1994). The Central Danube only comprises 2,002 km2 of recent floodplain (most of which are situated between Gemenc, southern Hungary, and the mouth of the Drava river, Kopacki Rit), but a much larger area has been lost. On the Lower Danube 2,200 km 2 of floodplains do still exist; however, more than 4,500 km2 (excluding the Delta) are cut off from the river and thus from regular flooding (Table11.3, Fig.11.7). For the whole Danube about 72% of the floodplain area was lost (Schneider 1991; WWF-DCP and WWF-AuenInstitut 1999; Schneider 2002a). These data show the broad extension of the former floodplains on the whole Central and Lower Danube (WWF-DCP and WWF-Auen-Institut 1999). Comparing the area of morphological floodplain to that of recent floodplain regularly subjected to floods, the full loss of floodplains becomes clear and with it the scale of human intervention and changes made to the river. Cut off from a functioning ecological network, floodplain areas along the Lower Danube slowly lost their capacity to function. One of the first consequences related to dam construction has been the breakdown of fisheries upstream of the delta, due to the lack of lateral connectivity between the Danube and its floodplains, which functioned as important spawning areas for migratory fish (most of them Cyprinid species). Every spring during high water levels in the Danube, these species entered the large, temporarily-flooded areas (so-called ‘intinsura’) with partially shallow waters, for reproduction. Loss of these large spawning areas has resulted in a drastic decrease in fish stocks, not only on the Lower Danube but in the Danube Delta as well (Staras 1998). Furthermore, the construction of the two dams, Iron Gate I and Iron Gate II, limited the longitudinal connectivity of the Danube river and prevented the migration of sturgeons, which had a major impact on spawning results (Staras 1998). Fisheries on the Lower
Table 11.3 Comparison of the extension of morphological and recent floodplain along the Danube (WWF-DCP and WWF-Auen-Institut 1999). The total length of the River Danube is 2,840 km and its catchment covers 817,000 km2 Morphological Recent River stretch floodplain (km2) floodplain (km2) Loss (%) Central Danube 8,161 2,002 75 Danube Delta 5,402 3,799 30 Lower Danube 7,862 2,200 72 Upper Danube 1,762 95 95
198
E. Schneider
Fig. 11.7 Loss of floodplains on the Lower Danube (Schneider 1991, after Botan 1984)
Danube no longer provided a viable economic income for local people. In addition to the loss of spawning places, increasing eutrophication caused by wastewater discharge from industry and intensive agriculture along the Danube also created problems. The filtering capacity of large floodplain areas acting as biological filters for the Danube Delta has been dramatically reduced. The loss of floodplain wetlands and large floodplain lakes has also led to local climate changes. Due to the disturbed hydrological regime in certain parts of the floodplain, the so-called ‘Balta’ area, an increasing dryness has occurred, together with the desertification and salinisation of soils in some areas of the former floodplain (Schneider 2002a).
11 Floodplain Restoration of Large European Rivers
199
11.3 Fundamental Considerations for the Restoration of a Functioning Floodplain The full consequences of floodplain loss along the Rhine and Danube rivers as outlined above can only be evaluated by considering the importance of the floodplains’ natural and socio-economic functions. In this regard, the increased risk of high floods has led to a recent change of mind when considering floodplains and their relevance as retention areas during major flooding events. Moreover, their important function of nutrient retention and recycling, their ecological functions as habitats for plants and animals, reservoirs of biodiversity, biocorridors and sites of high bioproductivity are now being highly recognised. This has led to the elaboration of various ecological flood protection and rehabilitation concepts. As the hydrological regime represents a major ecological factor, its improvement is an essential prerequisite for the restoration of self-sustaining floodplain ecosystems and a top priority in all restoration activities. The first step in planning ecological restoration on the Rhine and the Danube rivers was to evaluate the present ecological status of the area in order to answer the following questions: 1. How far is the present floodplain from being a functioning floodplain ecosystem? 2. What condition are the habitats in with regards to their degree of degradation, and to what extent has their original biodiversity been preserved? 3. What kind of measures will be required to restore the ecosystem into a sustainable functioning state? 4. How can the restoration objectives be reconciled with the present land use interests? To be able to answer these questions it is necessary to determine the minimum requirements for a functioning ecosystem, the optimal solution, and the real chances for restoration and recovery of natural functions in a floodplain that has been changed in so many ways through the intervention of man. Another important question is to what extent the reconnected former floodplains situated in the backwater areas of hydropower plants may be restored. Apart from the lack of floods, the high and hardly fluctuating groundwater levels can be a problem. To restore a functioning floodplain, the groundwater levels would have to be lowered and connected to the dynamics of the river. Persisting high groundwater levels lead to alterations and shifts in vegetation, as can be seen in the example of the Greffern-Söllingen area, situated in the backwaters of the Iffezheim hydropower plant. As the dynamics of softwood floodplain communities do not occur, the site is turning into boggy land giving way to the formation of alder swamp forests (WWF-Auen-Institut 1993). If the site potential (as regards to microrelief, soil texture, nutrient supply, microclimate) and the biotic potential (reproduction units of plants and site specific organisms) still exist, then, combined with the restoration of the hydrological regime (ecological flooding and lowering
200
E. Schneider
of the ground water level), they allow for the restoration of conditions similar to those of the floodplain.
11.4 Integrated Programmes for Flood Protection and Ecological Restoration 11.4.1 Programmes and Concepts for the Rhine 11.4.1.1 The Integrated Rhine Programme of Baden-Württemberg The ever increasing threat of floods on the Upper Rhine has led the Federal State of Baden-Württemberg to address flood protection in an integrated way. Numerous studies were undertaken to evaluate the status of former and recent floodplains on the Upper Rhine (Dilger and Späth 1988; Ministerium für Umwelt Baden-Württemberg 1988) and to develop concepts for floodplain restoration (WWF-Auen-Institut 1989, 1992a, b). The Integrated Rhine Programme (IRP), adopted in 1996, has proposed the creation of flood retention areas (polders) in the former floodplain of the Rhine between Basel and Mannheim (Pfarr et al. 1996). The IRP combines both economic interests of flood safety and ecological interests for floodplain ecosystem restoration. Diverse interests, such as water resource management, agriculture, forest management, and recreation, all had been taken into consideration (Dister 1990). At increased water discharge of the river Rhine (the discharge rates have been agreed), the retention areas, or polders, are artificially flooded via intake structures. The controlled flooding process allows a constant movement of water through the polder – but delaying it before its return to the river through an outlet structure. As the length of time of artificial flooding and volume of flood retention may be, to a large degree, controlled, polders can be used highly efficiently to offset flood peaks on the main river. The best method, however, that enables a free connection between the river and its floodplain, is to relocate existing dykes further inland, i.e. to provide the river with large enough natural inundation areas. At high water levels, the river can overflow its banks without any hindrance and thus the rise of a flood wave is attenuated early on, i.e. preventing critical discharges. To restore adequate living conditions for floodplain biocoenosis and securing its long-term survival, the flood tolerance of the biocoenosis must be considered. The objective is to provide such hydrological conditions that a disastrous flood merely represents one event among many regularly-occurring smaller flood events. All floods, not only disastrous ones, must have the best possible access to the retention (floodplain) areas. Ecological flood protection means that the restored floodplains are flooded as nearnaturally as possible, above all, during the vegetation period, thus providing appropriate ecological conditions to the re-established floodplain vegetation. To secure this, an
11 Floodplain Restoration of Large European Rivers
201
open polder can be equipped with inlet and outlet structures (sluice gates) that allow the water level in the polder to be adapted to that of the Rhine with only a short delay. The characteristic longitudinal flow and water exchange in the floodplain can again occur. In the case of smaller flood events, this kind of open polder can be left with open inlets and outlets; it will fill and empty all by itself given the low differences in water level between the Rhine River and the retention area. As for average flood events, sluice gates are used to control the peak water levels with the aim of modifying them: the water level difference between the polder and the river is allowed to continually increase with the rise of the wave; when the peak arrives, the still large retention area is filled and, in so doing, the peak of the flood wave lowered. 11.4.1.2 Concepts for Flood Protection on the Northern Upper Rhine Downstream of Mannheim In 1990, the state government of Hesse also started a similar initiative for flood protection through ecological restoration with a large integrated programme on the Northern Upper Rhine. Studies undertaken by WWF Institute for Floodplains Ecology (WWF Auen Institut) and the Hydraulics Institute at Darmstadt Advanced Technology University, identified floodplain areas cut off from the dynamics of the river that are suitable to provide flood retention. These included areas proposed to be used as polders (after being reconnected to the Rhine’s dynamics via inlet and outlet sluices), and areas suitable for the so-called ‘floodplain extension concept’, i.e. after relocation of dykes the natural flood areas of the river will be enlarged (Fig. 11.8; WWF-Auen-Institut 1990). Unfortunately this large initiative was not implemented, but the existing proposals could still be implemented if landowner and user interests can be reconciled. 11.4.1.3 Other Plans and Programmes for the Rhine The many regional programmes, plans or concepts for flood protection and restoration on the Upper Rhine, including those on the French side of the Rhine (Gartner 1995; Schmitt 1995; Freydefont et al. 1995) and in Rhineland-Palatinate (WWF-AuenInstitut 2000; Ministerium für Umwelt und Forsten Rheinland-Pfalz 2000) together form, along with those described above, the integrated programme ‘Rhine 2020’ of the ICPR (International Commission for the Protection of the Rhine, IKSR). This sustainable development programme for the whole Rhine includes: 1. Improvement of the entire river ecosystem (restoration of typical biotopes and their connectivity, and to assure the connectivity of tributaries), from Lake Constance to the North Sea 2. Improvement of flood control and flood protection 3. Improvement of water quality 4. Groundwater protection
202
E. Schneider
Fig. 11.8 Areas on the Northern Upper Rhine proposed to be open for free flooding or used as polders
11 Floodplain Restoration of Large European Rivers
203
These objectives will only be achieved through a comprehensive strategy of communication and public information, based on easily understandable materials. Monitoring the success of projects is an essential part of this programme. Assessments of the state of the Rhine will be undertaken regularly – according to the EU Water Framework Directive’s standards (IKSR Internationale Kommission zum Schutz des Rheins/ICPRR International Commission for the Protection of the Rhine River 2001).
11.4.2 Programmes and Concepts for the Danube 11.4.2.1 The Integrated Danube Programme (IDP) in Baden-Württemberg Similar to the Integrated Rhine Programme, the Integrated Danube Programme was adopted for the Upper Danube (from the source rivers Brigach and Breg to the town of Ulm) by the federal state of Baden-Württemberg (1992). The programme objectives are to: 1. Combine flood protection and ecological restoration 2. Create more flood retention areas by reconnecting floodplain areas 3. Restore floodplain habitats that were lost after having been cut off from the river’s dynamics At the same time the programme aims to ensure connectivity along the river and within its tributaries for fish migration. In total, the IDP comprises 169 measures along the Upper Danube and its tributaries in Baden-Württemberg; these measures have partly been implemented by the Danube/Bodensee Water Directorate of Baden-Württemberg, in concordance with the efforts for European biodiversity conservation. An example of a successful restoration project realised within the framework of IDP is the restoration of meanders and floodplain area in the Blochinger Sandwinkel (Klepser 1994; Pfänder 1994; Gewässerdirektion Donau/ Bodensee 1999). The IDP remains an open concept with possibilities of enlargement with new measures. However, the whole programme’s success will only be possible with large political and administrative support and that of both community and private land-owners. 11.4.2.2 The Danube Floodplain Concept Neuburg-Ingolstadt (Upper Danube, Bavaria) Despite the Upper Danube between Neuburg and Ingolstadt having been already developed with dams and hydroelectric schemes, and the fact that floodplains have been cut off from the river, a large initiative was started in 1995 by the town of Ingolstadt to return some dynamics to the river’s floodplains between the hydroelectric power plants of Neuburg and Ingolstadt, an area of some 2,100 ha. Based
204
E. Schneider
on a feasibility study (WWF-Auen-Institut 1997) for the reconnection of former floodplains to the Danube’s hydrological regime, a larger ecological concept has emerged. This concept was supported as a Leader II Project of the European Union (Stadt Ingolstadt/Umweltamt 2001). Following the above-mentioned feasibility study, three major points had to be considered: 1. Reconnection of floodplains to the river to ensure regular flooding 2. The dynamisation/oscillation of the groundwater table which at present is permanently high (due to the neighbouring barrages) 3. Restoration of a biocorridor function by recreating connectivity for migratory fish
11.4.2.3 The WWF Green Danube Programme The Green Danube Programme developed in 1990s by the WWF-Auen-Institut has been focused on the conservation, restoration and sustainable management of the Danube River floodplains on the basis of exemplary projects in the Danube catchment. These included projects located at: the mouth of the River Isar (a right-hand tributary of the Danube in Germany), the Multilateral Park of the central Danube east of Vienna – the actual Danube Floodplains National Park (Nationalpark Donauauen) and the Lower March (Morava) River floodplain; the Danube-Drava National Park, several Danube islands in Bulgaria; and the Danube Delta (Fig.11.9). The Danube Delta Biosphere Reserve has constituted from the very beginning a key part of this programme, the first pilot restoration projects having been realised there.
11.4.2.4 The Lower Danube Green Corridor Based on the successfully-implemented restoration projects in the Danube Delta, and the evaluation of the ecological and restoration potential of the Danube floodplains (WWF Danube-Carpathian-Programme and WWF-Auen-Institut/WWFGermany 1999), a large-scale transboundary nature protection initiative was started under the auspices of the Danube-Carpathian Programme of WWF International. In 2000, Romania, Bulgaria, Moldavia and Ukraine signed a declaration entitled the ‘Lower Danube Green Corridor’. One year later, all Carpathian and Danubian countries committed themselves to protect and develop the Danube-Carpathian region in a sustainable way. The Lower Danube Green Corridor, situated between two large protection areas, the Nature Park ‘Iron Gate’ and the Danube Delta Biosphere Reserve, constitutes Europe’s largest wetland area. It comprises a network of protected areas, areas to be protected, and floodplain areas to be restored (Fig. 11.10). On the Bulgarian Danube, studies have been conducted for the restoration of Persina and Kalimok/Brushlen sites through opening of dykes with the aim to reduce water pollution of the Danube. In the Romanian section of the Danube
Fig. 11.9 WWF Green Danube Programme with model projects
11 Floodplain Restoration of Large European Rivers 205
206
E. Schneider
Fig. 11.10 The Lower Danube Green Corridor – an initiative for an ecological functioning network of protected areas, planned protected areas and restoration areas
Calarasi DEM with Dyke (based on GPS-Point-measurements) 24.10.2002 73 74 78
III
77
76
IV 75
V 6
79 80
7 8
VI 9 9.3 346
347 13m and higher
348 349
12m
350 351 352
I
11m 10m
13.00 - 17.18 12.00 - 12.99 11.00 - 11.99 10.00 - 10.99 9.00 - 9.99 8.00 - 8.99 7.53 - 7.99
352.5 Dyke
9m
Bank of Danube Main Channel
8m
0
II
Fig. 11.11 Restoration area on the Lower Danube, island Calarasi-Raul/Romania
0.5 1
1.5
2km
11 Floodplain Restoration of Large European Rivers
207
floodplains, the Calarasi-Raul project (Fig.11.11), financed by the Romanian government and the World Bank, intends to reduce the pollutant input originating from agriculture through restoration of the floodplain – 3,000 ha of former rice polders should be reconnected to the flood regime of the Danube by smaller and larger openings. As the costs of such restoration projects are relatively high, it is necessary to compare the costs of agricultural use of the area (i.e. water pumping, irrigation) and benefits of the area if reconnected to the Danube’s hydrological regime – the benefits of restoring the Danube’s biogeochemical functions such as the nutrient retention and recycling capacity of the restored floodplain. During the high floods that occurred on the Danube in 2006, the entire area of Calarasi-Raul polder had been opened in the spots proposed by the abovementioned project. Moreover, the upstream area of the island, i.e. the part that was intended for agricultural use to continue, was also reconnected and used for flood retention. Still in late autumn the inner parts of the island were covered by water. This backed the importance of this area for flood retention. Regrettably, in 2007 the dyke openings have been again closed. Discussions on the importance of this possible flood retention area do, however, persist. In the meantime a number of other projects have commenced within the Lower Danube Green Corridor. Among those is the Forestry Strategy for the Bulgarian Danube Islands (2001) which describes precise steps to be taken in order to restore the floodplain forests. It comprises the conversion of hybrid poplar monocultures into near-natural floodplain forests. A further project of the LIFE programme that is to be completed by 2010 also includes, among other activities, the conversion of poplar monocultures into near-natural floodplain forests. It involves a number of Danube islands near the city of Calarasi, upstream Calarasi-Raul island. Another project “Protection of wetlands of the Danube – a pilot project for the Cama-Dinu area (county Giurgiu)” has started in 2004 with PHARE financial support. The project consists of two parts: collection of biodiversity data in the area of Cama and Dinu islands (river-km 500–521) and evaluation of the ecological conditions and biodiversity of a 455 km long section of the Danube River (river-km 838–383). The outputs of this project are proposals for sites to be included in NATURA 2000 network and recommendations for restoration. The largest project within the Lower Danube Green Corridor is the project entitled “Ecological and economic floodplain re-dimensioning on the Lower Danube” financed by the Romanian Government/Ministry for Environment and Sustainable Development (resolution 1208/06.09.2006). Aerial laser survey has been conducted in 2007 by the Danube-Delta-Institute in Tulcea in cooperation with several other partners to provide fundamental data for further designation of floodplain restoration and flood protection areas within the Lower Danube. Apart from the benefits of restoring the hydrological, biogeochemical and ecological functions, a cost/benefit analysis of restoring the fishing grounds and other traditional agricultural practices, development of rural tourism should be done.
208
E. Schneider
11.5 Examples of Implemented Restoration Projects 11.5.1 Restoration Projects on the Rhine 11.5.1.1 Restoration of Floodplains on the Northern Upper Rhine – the example of the Kühkopf-Knoblochsaue Nature Protection Reserve Years before the concepts of flood protection through ecological restoration for the River Rhine were developed, a large-scale floodplain restoration started in the Nature Reserve ‘Kühkopf-Knoblochsaue’, situated on the northern Upper Rhine (historical evolution of this site see in Fig.11.12). The area constitutes one of the largest protected floodplain areas in central Europe. The central part of the Kühkopf, about 700 ha (of which 400 ha were intensively cultivated by a stateowned farm), was protected against the Rhine floods by a system of dykes until the 1983 flood when a dyke broke. The dyke was not reconstructed and the floodplain became again subject to high floods. By the decision of the Land Hessen authorities, the agricultural use came to an end. Initially a small area remained in agricultural use, but even this area was abandoned after the high and long summer flood of 1987. The hydrograph for the years of 1987–1991 is given in Fig.11.13. In this way, the conflict between the Nature Reserve’s conservation status (declared in 1952) and its intensive agricultural use was resolved. This was the beginning of a large-scale floodplain restoration in Europe (Fig.11.8). On the abandoned agricultural land, areas for restoration of floodplain meadows
Fig. 11.12 Historical evolution of the landscape on the northern Upper Rhine, area of Kühkopf around 1799 (left) and recent topographical map (right)
11 Floodplain Restoration of Large European Rivers
209
Fig. 11.13 Hydrographs on the gauche station Worms/Upper Rhine, upstream the area of Kühkopf-Knoblochsaue
and floodplain forests have been designated through close cooperation between nature conservation and forest authorities. When deciding the designation of areas for redevelopment of floodplain forest and floodplain meadows and the proportion between them, the restoration of the specific cultural-historical landscape was also taken into account. To improve scientific knowledge in the field of floodplain forest and meadows restoration, permanent plots of natural succession (evolution of floodplain forests) and controlled succession (evolution of floodplain meadows) were established and documented since 1986 (Fig.11.14). The experimental plots included sites with varying soil conditions (different sand content) and different micro-relief. To document the influence of game (wild boar, deer) on the course of succession, in particular for the evolution of hardwood floodplain forests but also for the meadows, fenced and non-fenced plots were established for every experimental area. A large network of experimental variants were thus established and monitored. For controlled succession, a total of 18 permanent plots (nine fenced and nine non-fenced) were monitored at five different sites. For natural succession without management, four (provisionally five) sites were monitored with 16 fenced and four non-fenced quadrats, each square including 25 small quadrats; hence a total of 500 quadrats were monitored to document the evolution of floodplain forests – with both number and proportion of different wooden species. The research also included a comparison between a plot seeded with seeds collected from a species-rich, extensively-used floodplain meadow existing in the area of Kühkopf-Knoblochsaue.
210
E. Schneider
Legend:
N
monitoring site - natural succession monitoring site - controlled succession
0
Water Reedbed Woodland Grasslands under traditional use Sites with natural succession Sites with controlled succession Settlement Nature Reserve border Winter dykes Summer dykes Path Water line
500 1000 m
Erfelden
Versuchsfeld Rindswörth Finkenwiese
Schafweide
Schwedenkirchhof
Plattenacker
Erlenwäldchen
Schwarzer Weiher
Ochsenlache Vor dem Eichwald
Stockstadt
Fig. 11.14 Restoration area in the Nature Reserve Kühkopf-Knoblochsaue with sampling/monitoring points
In succession-controlled plots (cutting either once or twice a year), the succession went through various stages: from the stage of annual therophytes (annuals-ephemerals) in the first and second year to that of ruderal hemicryptophytes (‘ground rosette-like’) in the third, fourth and fifth year, followed by grass stages with different grass-species and herbs from the sixth year onwards. The stage of ruderal hemicryptophytes was dominated by thistle (Cirsium arvense) in the third and fourth year (years 1985, 1986). As a consequence of the 1987 long summer flood and the selection between the species with respect to their flood resistance, the thistle was eliminated more rapidly. It was followed
11 Floodplain Restoration of Large European Rivers
211
by the establishment of a grass-stage with herbs, which developed microfacial structures showing ‘flowerbed’ aspects (Schneider 2002b). This structure remained for 4–5 years and changed, more or less, only due to the fluctuating water levels. Finally, regular cutting altered these microfacial structures and the area’s aspect became similar to natural floodplain meadows. In the eighth year, and even more in the ninth year, the grasslands growing in more elevated places (already a few centimetres playing a role) turned into Arrhenatherum elatius-like meadows, but on lower spots, the floodplain meadows were dominated by Alopecurus pratensis. In the lowest places, i.e. the more or less regularly-flooded sites, wet-meadows with Poa palustris occurred. Beginning from the 10th year, the meadow’s aspect became more and more typical that of floodplains, with a more or less stable species composition in the 15th year. Due to the flooding regime, the abundancedominance of species was subject to many changes. Due to the break-up of the vegetation cover by wild boar, within the large developed area of floodplain meadows, smaller areas with initial stages occurred. These pioneer communities were characterised often by small Chenopodium species, Hyoscyamus niger (very rare in Hessen land), and Mercuralis annua and constitutes a community of annual plants (Mercurialetum annuae) which has become in recent decades very rare on the Upper Rhine (Schneider 2002b). Frequently, different development stages were lying side by side, creating a mosaic which is characterised also by specific macroinvertebrates. In contrast to the areas of controlled succession, one may observe a slower although more continuous development on the natural plots. From the very beginning, white willows (Salix alba) and black poplar (Populus nigra) had settled along with other herbaceous pioneer plants on open virgin soils. In the following years, these pioneer species were complemented by species of hardwood forests (QuercoUlmetum). The first plants of hardwood floodplain forests were dogwood (Cornus sanguinea), hawthorn (Crataegus monogyna), oak (Quercus robur) and ash (Fraxinus excelsior). In the area where the dyke had broken, over a surface of about 4 ha, the soil of the former agricultural lands was covered with sand. This is where, uniquely in central Europe, the natural development of a softwood floodplain forest from a pioneer stage with Salix alba and Populus nigra and the transition to a hardwood forest with oak (Quercus robur), elm (Ulmus laevis, Ulmus carpinifolia) and ash (Fraxinus excelsior) may be observed. The results of this 20-year study can be summarised as follows: 1. It revealed that floodplain succession can show peculiarities that differ from a development in humid fallow lands beyond the floodplain. This difference is caused by the co-interaction of the flooding factor with the other ecological conditions. 2. The decisive factor was the restoration of a flooding regime that was as nearnatural as possible. As regards a succession towards a floodplain habitat, the more frequent small floods are just as important as the rarer but more significant floods. If the shorter and more frequent floods no longer take place due to human intervention, the succession may temporarily be inverted and lead to communities untypical for floodplains.
212
E. Schneider
3. The height, duration, period and frequency of floods played a determining role in the development of the area. Summer floods of long duration, such as the 1987 flood, make a selection between flood-resistant, flood-tolerant and floodintolerant species and are responsible for a shift towards typical floodplain habitats. Years without floods are responsible for a stagnation in the evolution; as a consequence of a series of drier years, a shift towards species communities preferring drier conditions took place. The alternation of wet and dry years is the basis for so-called ‘harmonic’-succession, with a changing shift between wetter and drier communities. 4. Succession progresses very rapidly in the restored floodplain zones and follows different stages from the annual pioneer therophytes, through a ruderal grassrich stage of hemicryptophytes, and to differently structured grass-stages. In the grass stages the dominance of different species is strongly dependent on the flooding regime. 5. Seeding accelerated the transition from the initial phase to the grass- and herb-rich stage, so that there was no ruderal stage. But after 6–7 years, the evolution of seeded and unseeded plots was the same and no differences were visible. 6. The seed (diaspore) bank of the floodplain soil plays an important role in the evolution of the area. A considerable genetic potential – with numerous rare and endangered species – can be activated under certain circumstances. On the Kühkopf, rare species for the entire Upper Rhine such us Hyoscyamus niger, Verbascum blattaria, Euphorbia falcata and even communities of therophytes and thero-hemicryptophytes which had been considered to have disappeared on the Upper Rhine such as the community of Mercurialis annua with Kickxia elatine could be found (Schneider 1995, 2001, 2002b). 7. Controlled succession proved to develop more rapidly as compared to natural succession and the regularly-cut meadows already reached a stable stage after 20 years, but are still lacking the high plant species diversity characteristic of extensively- and traditionally-used floodplain meadows. The regular mowing of the meadows is important for the extraction of nutrients remaining in the soil from the fertilisation of the former intensively-used agricultural lands. 8. The development of macroinvertebrate communities (biocoenoses) was distinctly related to that of vegetation structures. 9. The sediment dynamics that provide virgin soils was vital for the natural regeneration of softwood forests (this was demonstrated at the site near the breaking of the dyke, where a large-scale re-settling by black poplar and white willow took place). 10. High deer density substantially restricts oak regeneration as was shown in fenced and non-fenced plots (the situation has been considerably improved by regulation of game stock through hunting). 11. The wild boar activity, in particular in the meadows, played an important role in the development of the area. Due to the break-up of the vegetation cover on the open soils, so-called micro-succession developed – with often rare pioneer species, originally from the seed bank of the soil (Schneider 2002b).
11 Floodplain Restoration of Large European Rivers
213
Thanks to the reconnection of the floodplain to the Rhine’s water dynamics, redevelopment of a site’s typical mosaic of macro- and micro-habitats was possible. It is a good example of restoration of a functioning floodplain that shows that, if understanding exists, a large area of agricultural lands can be restored and turned into typical floodplain communities, for the benefit of nature and local people. It is beneficial for farmers, who are allowed to cut the meadows under certain conditions (not before 15th of June) and to use the hay, and also for the people of the surrounding areas for recreation. Furthemore, the Kühkopf-Knoblochsaue Nature Reserve plays a remarkable role in environmental education, presenting nature conservation issues in the information centre supported by the federal state. With the help of a functioning model (with possible flooding) and maps, flood protection issues can be explained.
11.5.1.2 Projects in the Nature Reserve Floodplains of Rastatt The Rhine floodplain near Rastatt on the middle part of the Upper Rhine downstream of the Iffezheim power plant, is one of the few existing near-natural floodplain areas connected to the Rhine and subjected to free and regular floods. This area has been declared as a Nature Reserve in 1984. Even though the Rhine floodplains near Rastatt are recent floodplains that underlie the direct dynamics of the Rhine River, conditions were in need of improvement in some sections. Numerous runoff obstacles occurred caused by tracks that were situated too high and thus had a dam effect. They hindered a regular flow-through as the connectivity of the flood channels had been disrupted. Moreover, in some places the pipe culverts were too small and led to impoundments at the moment of high water levels. And last but not least, gravel extraction and forestry (hybrid poplar monocultures) had also left its marks. For this reason the WWF-FloodplainsInstitute developed proposals for improving the hydrological and ecological situation of the area. A feasibility study to improve water circulation and ecological conditions in the Rhine floodplains between Rastatt-Wintersdorf (335.7 km) and Au a. Rhein (354.1 km) – a total area comprising about 1,680 ha was prepared between 1987 and 1991. This was followed by an intensive planning and implementation of the proposed measures. The project was part of the Integrated Rhine Programme. The main emphasis of the study was put on the examination of aquatic life communities as indicators for the quality of habitats. Their present state has been examined with respect to three major aspects: • The general limnological conditions (including physical and chemical parameters) • The aquatic macrophytes • The present hydrological conditions The latter have been studied from the point of view of both water management and ecology; to accomplish this, the hydraulic conditions have been studied in detail on specifically-chosen sites.
214
E. Schneider
Due to the situation in the transition area between the braided and the meandering zone of the former Rhine river (see Fig.11.2), a multiform geomorphological structure and very manifold water types are present in the area. These types are oxbow lakes, old smaller or larger river branches connected to the Rhine, groundwater streams, ponds with a groundwater supply, gravel extraction lakes, etc. All the different types of water bodies, during mean water levels, cover about 430 ha, i.e. 25% of the whole studied area. In these waters, physical parameters such as water temperature, conductivity, pH, and oxygen were measured. In addition, in the gravel-extraction lakes, temperature and oxygen stratification was studied. On the basis of their oxygen saturation, water bodies were categorised into four groups: • • • •
Waters with a changing saturation, from over-saturated to deficiency of oxygen Waters with more than 60% of oxygen with frequent over-saturation Waters with around and over 60% oxygen Waters with less than 60% saturation of oxygen
These different water types support a diversified macrozoobenthos fauna, which has been used to evaluate the ecological state of the water bodies. Macrophytes and their communities were mapped to evaluate the current ecological status of the sites and to make a prognosis concerning possible changes in vegetation after hydrotechnical improvement measures will have been implemented. By comparing the macrophyte samples with older vegetation data (from 1978), it was possible to see the degree of changes in some water bodies – from mesotrophic to eutrophic. The comparison also made it possible to see the increase of populations of Ceratophyllum demersum, Potamogeton pectinatus and Zannichellia palustris, as indicators of eutrophication – and the decrease of populations of Characeae species, Hippuris vulgaris, Groenlandia densa and Potamogeton friesii in waters which had a more mesotrophic or meso-eutrophic character and were shifting to more eutrophic conditions. The shift towards the species of more eutrophic conditions was observed also on the macrozoobenthos fauna. The eutrophisation of aquatic habitats has been explained mainly by the loss of natural floodplain areas along the River Rhine upstream the studied area, the canalisation and impoundment of the Upper Rhine directly upstream the studied area and the change in land use. This has resulted in increased flow velocities and higher inputs of fine sediments to the floodplain due to its location downstream of the Iffezheim hydropower plant and the wash-out of fine sediments from the dam. Furthermore, the floodplain’s natural flow gradient was distorted by trails and artificial gravel pits. This has led to the occurrence of stagnating waters and waterlogged areas on the one hand, while on the other hand other areas suffered from drying out. In the artificial slack water zones, the flow of the unpolluted flood wave during its normal runoff has been obstructed and thus sedimentation of a higher load of suspended matter increased. In the connected floodplain waters, the flood flow has been accelerated and erosion encouraged. Following the study, implementation of first restoration plans in the Nature Reserve “Floodplain of Rastatt” begun in 1990 (cf. Dister 1990). A gravel pit has
11 Floodplain Restoration of Large European Rivers
215
been separated by earth spill from a valuable oxbow lake called “Bärensee” with a remarkable population of Trapa natans, a very rare organism on the Rhine. The sites with Trapa natans had been endangered by flows of cold water from the gravel extraction lake and measures to protect the sites of this species were needed. At present, the situation of the oxbow lake with Trapa natans after these construction measures is stabilised and oscillations of the population are only due to flood events. In other parts of the recent floodplain near Rastatt, a number of dirt roads which had formed dam-like obstacles during floods have been lowered (Figs.11.15 and 11.16).
Fig. 11.15 Restoration measures for improvement of water circulation: a lowered trail which had dam effects
216
E. Schneider
Fig. 11.16 Flooding of a restored floodway (“Schlute”) during high water levels in the floodplain near Rastatt-Wintersdorf
The hydrological and ecological conditions of the recent floodplain have been considerably improved after its reconnection to the dynamics of the Rhine River. The shift of aquatic species composition towards the eutrophic zone is no longer observed. Instead species of meso-eutrophic waters such as Potamogeton densus or Hippuris vulgaris or small Potamogeton species such as Potamogeton berchtoldii occur more frequently again. Comparable circumstances may be observed for macrozoobenthos. Another project implemented between 1989 and 1994 was the re-establishment of near-to-natural forest management methods within the still-existing hardwood floodplain forests in the “Rastatter Rheinaue” Nature Reserve (850 ha). The objectives of this project included the replacement of hybrid-poplar monocultures by near-to-natural, structured floodplain forest with its characteristic biodiversity and close-to-natural water regime. Changes in the forest herbaceous layer as well as the dynamics of epigeic zoocoenoses and mezofauna have been documented. Restoration was undertaken in experimental plots on various floodplain sites (Fig.11.17) that differ by the height and duration of flooding and soil composition. Different forest management methods were applied including selective cutting of individual poplar trees. In the open spaces that occurred natural succession was encouraged and, in addition, oaks as well as wild fruit trees were planted. Some results of the 5-year study on the floodplain forest restoration can be summarised as follows: 1. The improvement of water circulation, i.e. improving the system dynamics, had positive effects on the forest ecosystem. Moreover, due to the improved connectivity, habitats for fish – including spawning places – were created.
11 Floodplain Restoration of Large European Rivers
217
Fig. 11.17 Sampling/monitoring points for floodplain forests restoration in the Nature Reserve “Rastatter Rheinaue”
2. The forest management measures had a positive effect on the development of a near-natural floodplain forest – increased species numbers and higher biodiversity was observed on the experimental plots (for more details, see WWF-AuenInstitut 1995).
218
E. Schneider
3. The flood tolerance of different species was tested – the flood tolerance of oaks planted in stands at the lower level of the hardwood forest (where hybrid poplars had been extracted) was very high (32–40 days). The highest tolerance at the hardwood forest level was found to have elm (Ulmus laevis, U. carpinifolia), oak (Quercus robur), chestnut (Corylus avellana), wild apple (Malus sylvestris) and wild pear (Pyrus pyraster); ash (Fraxinus excelsior) was less tolerant compared to these species. 4. The changed light conditions in the forest had a determining influence on the composition of the herb layer and had its largest effect on the dominant, invasive species that occurred in the poplar monocultures; for example, Himalaya Balsam (Impatiens glandulifera) was reduced as a result of the closed crown canopy and the enhanced structural diversity. The changed forest structure has also led to alterations in soil fauna and a shift towards a higher diversity of characteristic floodplain species. The results of the forest restoration project in the floodplain of Rastatt may serve as a basis for forest restoration projects on other European rivers: a similar project was initiated by WWF-Auen-Institut on Vardim island (downstream the town of Svistov) on the Lower Danube in Bulgaria. The experience gained, including the improvement of the hydrological regime, were also used for other proposals, including the “Strategy for the Protection and Restoration of Floodplain Forests on Bulgarian Danube islands” developed by the Bulgarian Ministries of Agriculture and Forests, and Environment and Water, WWF, Bulgarian experts and local NGO Green Balkans Bulgaria (2001).
11.5.2 Restoration Projects on the Danube 11.5.2.1 Restoration of the Floodplain of Regelsbrunn and Orth in the Danube National Park Austria The Danube floodplain near Regelsbrunn, on the right side of the Danube, and the floodplain of Orth, on the left side of the Danube, are part of the Danube National Park east of Vienna and have been, for a very long time, of great concern to nature conservationists. In this context, studies concerning the floodplain forest restoration and management were undertaken in the early 1990s (WWF-Auen-Institut 1992b). After an intensive campaign lasting over 13 years, WWF Austria secured the designation of these two floodplains as a National Park in 1996. In 1997, the Danube National-Park Authority together with WWF-Austria and the Austrian Waterway Administration completed the reconnection of the floodplains to the dynamics of the Danube. The objectives were the improvement of the hydrological regime dynamics and morphodynamics, re-creation of fish connectivity, creation of site typical habitats and the increase of characteristic biodiversity for the site (Schiemer et al. 1999). The success of the implemented project was recognised by the University of Vienna as regard the hydrological, biogeochemical and ecological functional aspects (Tockner et al. 1999). The realised projects demonstrate the possibilities for restoration of a
11 Floodplain Restoration of Large European Rivers
219
functioning floodplain ecosystem with typical habitats and biodiversity, when the main factor controlling the system, i.e. the hydrological regime, is restored. 11.5.2.2 Restoration Projects on the Lower Danube and the Danube Delta Restoration of polders that had been reclaimed for agriculture and fish farming but later abandoned started in the Danube Delta Biosphere Reserve, in close collaboration between the Danube Delta Biosphere Reserve Authority, the Danube Delta National Institute and WWF-Auen-Institut in Rastatt/Germany. After the successful implementation of two pilot projects – agricultural polders Babina (2,100 ha in 1994) and Cernovca (1,580 ha in 1996) in the north-eastern part of the Danube Delta on the Chilia branch – restoration continued with the reconnection of the fish polder Popina south (3,600 ha in 2000), situated downstream of the two aforementioned polders on the same branch, an more recently with the forest polder Fortuna (2,115 ha). The monitoring programme undertaken in 2006 and 2007 by the Danube Delta National Institute and WWF Institute for Floodplains Ecology has demonstrated the first successes of the implemented restoration measures. Another restored area is the fishpond of Holbina (5,630 ha) where monitoring activities are running. The restored areas in the Danube Delta altogether account to more than 15,000 ha making it the largest restoration area in Europe. Ten years of monitoring performed at the Babina island that was reconnected to the hydrological regime of the Danube in 1994 have confirmed that: • • • •
Reflooding supports the development of diverse macro- and microhabitats Biodiversity (species, habitats) has increased and is well-maintained The aquatic habitats offer sufficient food resources for fish The regression of halophilous vegetation and the development of grasslands and reeds offer possibilities for traditional land-use
The development of site-specific habitats for a large range of biological communities occurred quite rapidly after the opening of the circular dyke surrounding the island’s polder (Table11.4). The restoration is considered a big success that has brought local population back the traditional economic occupations – fishing, hunting, reed-harvesting and the use of seasonally flooded pastures for cattle grazing. The resource evaluation has demonstrated economic benefits to the local people in the area (Marin and Schneider 1997; Schneider et al. 2008).
11.6 Lessons Learned 1. To restore river floodplains we require a knowledge of the functioning of a natural floodplain ecosystem as a whole (the river and its floodplain inseparably coupled together) and an understanding of the various factors that influence how the system functions.
220
E. Schneider
Table 11.4 Overview of macrohabitats in the area of Polder Babina Before opening (1994) After reconnection 1994 Small patches of halophilous vegetation Large areas of halophilous vegetation and with their typical diversity of halophilous macroinvertebrates in the macroinvertebrates upstream part of the polder – Temporarily-flooded meadows Dry xerothermic vegetation of dams Dry xerothermic vegetation of dams – Pioneer vegetation on newly-developed sediment banks – Development of riparian softwood stands of different age classes Disturbed reed beds Well-developed reedbeds Poor aquatic macrophytes in the artificial canals Species-rich and abundant aquatic macrophyte communities Artificial canals Artificial canals – Revitalised old river channels (Gârla) with typical vegetation – Lakes – Stagnating temporary waters – Temporarily-flooded channels
2. An evaluation of the present status of each area to be restored, based on studies of habitat structure and functioning and the presence of indicator species, forms the basis for further planning: for determining the degree of naturalness, and the degree of modification. This also has importance for the implementation of the Water Framework Directive. 3. The first pre-requisite for successful floodplain restoration is the restoration of the site’s hydrological regime. Its degree of disturbance, as well as the extent of changes with respect to site conditions, habitat composition and structure, determines the type of measures to be selected for restoration. 4. Monitoring the success of a restoration is a very important component of every restoration undertaking. The questions to be asked are: “Is the system functioning in a sustainable way and are additional measures required for the system to operate in a self-regulatory manner?” 5. All aspects of the project, i.e. ecological and socio-economic aspects, and the balance between the ecological and economic benefits, have to be openly discussed with all interest groups. 6. A cost-benefit analysis is both helpful and necessary as a basis for the acceptance and implementation of restoration projects. For example, in the case of the Integrated Rhine Programme: the costs for its implementation were estimated at €500 million, whereas the damage caused by a big flood (flood with a returning period of 200 years) was estimated to amount to more than €6 billion. 7. To achieve a wider acceptance of floodplain restoration projects, a clear and easily comprehensible communication directed at all concerned is needed. This should include an explanation of the important functions of floodplains, the losses incurred by their degradation, and the advantages and economic benefits from their restoration. Such information will prove necessary at all levels: the local
11 Floodplain Restoration of Large European Rivers
221
population as well as local, regional and national authorities have to be involved. Local authorities play an important role in the discussions with all stakeholders. 8. The feeling of ownership is an important aspect in any restoration project. Experience has shown that restoration measures are often impeded because owners or tenants of the areas do not accept them. 9. Apart from the above aspects, and the useful experience gained for other restoration projects, the projects demonstrate that even on large rivers, or parts of them, which have been turned into artificial water ways – such as the Rhine and the Danube with their floodplains cut off from their hydrological dynamics – restoration work can still be done successfully. Though many aspects have to be taken into account, from both ecological and socio-economic points of view, if an understanding that improvements are necessary exists, than success can also be achieved.
References Botan M (1984) Apele in viata poporului roman, Ceres Bucuresti, pp 410 / The waters in the life of Romanian people Dilger R, Späth, V (1988) Rheinauenschutzkonzeption im Regierungsbezirk Karlsruhe. Materialien zum Integrierten Rheinprogramm, bd. 1: 1–178, BNL Karlsruhe (in German) Dister E (1985) Auenlebensräume und Retentionsfunktion. In: Die Zukunft der ostbayerischen Donaulandschaft, Laufener Seminarbeiträge ANL Laufen/Salzach 3:74–90 (in German) Dister E (1986) Hochwasserschutzmaßnahmen am Oberrhein. Ökologische Probleme und Lösungsmöglichkeiten. Geowissenschaften in unserer Zeit 4(6):194–203, VCH Verlagsgesellschaft Weinheim (in German) Dister E (1990) Floodplain protection in central europe. Gate questions, answers, information 3: 3–15, Deutsches Zentrum für Entwicklungstechnologien (in German) Dister E (1994) The function, evaluation and relicts of near-natural floodplains. Limnologie aktuell, vol. 2, Kinzelbach (Hg.), Biologie der Donau, Gustav Fischer Verlag Stuttgart, Jena, New York (in German) Dister E (1995) Die Ökologie der Flußauen und ihre Beeinträchtigung durch den Verkehrswasserbau. Das 2. Elbe-Colloquium, pp 56–64, Editor. Michael Otto-Stiftung für Umweltschutz, Edition Arcum (in German) Dister EP, Obrdlik E, Schneider Er. Schneider, Wenger E (1989) Zur Ökologie und Gefährdung der Loire-Auen. Natur und Landschaft 64(3):95–99 (in German) Freydefont M, Durand E, Ferry O (1995) La renaturation de la bande rhénane. Propositions de programmes des services déconcentrés et établissement publics de l’état de la région Alsace). Prefecture de la Region Alsace, Strassbourg, pp 13 (in French) Gartner K (1995) Inventaire des operations de restauration des anciens bras du Rhin. Document de présentation, Ecole Nationale Supérieure d’Agronomie et des Industries Alimentaires de Nancy, pp 76 (in French) Gâstescu, Zavoianu, Rusu (1983) Anthropic modifications of the hydrographic network. Geography of Romania, vol I, Physical geography, EDit. Academiei, Bucuresti, pp 309–310 Gewässerdirektion Donau/Bodensee (1999) Lebensraum Donau erhalten – entwickeln. Maßnahmen Stand Sept. 1999. Integriertes Donau-Programm, Heft 5, Riedlingen (in German) Gewässerdirektion Südlicher Oberrhein/Hochrhein (ed) (1997) The Integrated Rhine Programm. Flood control and restoration of former flood plains on the Upper Rhine, Lahr Gonnermann H (2002) Die Wälder des Naturschutzgebietes – von der Pappelwirtschaft zum Prozessschutz. %0 Jahre Naturschutzgebiet Kühkopf-Knoblochsaue, Regierungspräsidium Darmstadt, pp 28–42 (in German)
222
E. Schneider
Günther-Diringer D (2003) Aufbau eines online-Flussauenbewertungssystemes großer Flüsse Mitteleuropas. Rhein, Elbe, Oder und Donau. Dissertation Universität Salzburg, pp139 (in German) Günther-Diringer D, Musall H (1989) Landschaftshistorische Entwicklung der Rastatter Rheinaue. Karte, FH Karlsruhe, Studiengang Kartographie (in German) Hügin G (1981) The riparian woods of the southern Upper Rhine valley – changing and endagering by Rhine development. Landschaft und Stadt 11(2):78–91, Ulmer Verlag Stuttgart IKSR Internationale Kommission zum Schutz des Rheins (2001) Rhein 2020. Programm zur nachhaltigen Entwicklung des Rheins. Rhein-Ministerkonferenz 2001, pp 27 (in German) Klepser H (1994) Die naturnahe Umgestaltung der Donau bei Blochingen, Stadt Mengen, Landkreis Sigmaringen, Land Baden-Württemberg. In: Habitat Danube – an European EcoSystem, Proceedings of the international colloquium at Ulm. Beiträge der Akademie für Natur – und Umweltschutz Baden-Württemberg 17:240–247 (in German) Konold W, Schütz W (1996) Die Donau – Gefährdungen eines internationalen Flusses.In: Lozan JL, Krausch H (Hrsg) Warnsignale aus Flüssen und Ästuaren, Berlin (in German) Landesanstalt für Umweltschutz Baden-Württemberg, LfU (1999) Auswirkungen der ökologischen Flutungen der Polder Altenheim. Ergebnisse des Untersuchungsprogramms 1993–1996. Materialien zum Integrierten Rheinprogramm 9, Karlsruhe (in German) Landesamt für Wasserwirtschaft Rheinland-Pfalz (2000) Integriertes Raumnutzungskonzept für die Hördter Rheinniederung. Ministerium für Umwelt und Forsten Rheinland-Pfalz (in German) Marin G, Schneider E (1997) Ecological restoration in the Danube Delta Biosphere Reserve. Babina and Cernovca islands. ICPDD/Umweltstiftung WWF-Deutschland, pp 120 Ministerium für Umwelt Baden-Württemberg (1988) Biotopsystem Nördliche Oberrheinniederung. Bestandsaufnahme und Entwicklungsvorschläge. Materialien zum Integrierten Rheinprogramm 2:1–137, Bonn, Karlsruhe, Oppenheim, Wiesbaden (in German) Ministerium für Umwelt und Forsten Rheinland-Pfalz (2000) Integriertes Raumnutzungskonzept für die Hördter Rheinniederung, pp 4 Mock J, Kretzer H, Jelinek D (1991) Hochwasserschutz am Rhein durch Auenrenaturierung im Hessischen Ried F(lood control of the Rhine by remodeling alluvial meadows in the Hessische Ried). Wasser & Boden 43(3):126–130 (in German) Pfänder J (1994) Die naturnahe Umgestaltung der Donau bei Blochingen. Exkursionspunkt 5: Blochinger Sandwinkel. In: Habitat Danube – an European Eco-System, Porceedings of the international colloquium at Ulm. Beiträge der Akademie für Natur- und Umweltschutz BadenWürttemberg 17:299–308 (in German) Pfarr U, Kuhn S, Huppmann O, Klaiber G (1996) Rahmenkonzept des Landes Baden-Württemberg zur Umsetzung des Integrierten Rheinprogramms. Materialien zum Integrierten Rheinprogramm 7:1–96, Oberrheinagentur Lahr (in German) Regionale Zusammenarbeit der Donauländer (1986) Die Donau und ihr Einzugsgebiet. Eine hydrologische Monographie. München (in German) Schiemer F, Baumgartner C, Tockner K (1999) The Danube restoration project: conceptual framework, monitoring program and predictions on hydrologically controlled changes. Regul River 15:231–244 Schmitt L (1995) Approche methodologique pour une restauration des anciens bras du Rhin. Université Louis Pasteur I, UFR de Géographie de Strasbourg, Mémoire de Maitrise de Géographie Physique, pp 70 (in French) Schneider E (1991) Die Auen im Einzugsgebiet der unteren Donau. In. Erhaltung und Entwicklung von Flussauen in Europa. Lauefener Seminarbeiträge 4:40–57, Lauffen/Salzach (in German) Schneider E (1995) Zur Vegetationsentwicklung auf den aufgelassenen Ackerflächen des Kühkopfs und das damit verbundene Auftreten seltener Arten. Collurio Zeitzschrift für Vogelund Naturschutz in Südhessen 13: 67–78 (in German) Schneider E (1996) Pioniervegetation kurzlebiger Arten an der mittleren Loire und dem unteren Allier. In: Braunschweiger Kolloquium zur Ufervegetation von Flüssen in Braunschweiger Geobotanische Arbeiten 4:309–322, TU Braunschweig (in German) Schneider E (2001) Restoration of floodplain meadows and forests. Results of 15 years of monitoring in natural and controlled succession on re-flooded areas in the Nature Reserve Kühkopf-
11 Floodplain Restoration of Large European Rivers
223
Knoblochsaue/Upper Rhine. In: River restoration in Europe. Practical approaches, Proceedings Conference on River Restoration, RIZA rapport nr. 2001.023:197–199 Schneider E (2002a) The ecological functions of the Danubian floodplains and their restoration with special regard to the Lower Danube. Large rivers 13:1–2, Archiv Hydrobiol Suppl. 141/1–2:129–149 Schneider E (2002b) Vom Acker zur Auenwiese. 20 Jahre Grünlandsukzession auf dem Kühkopf. in: 50 Jahre Naturschutzgebiet Kühkopf-Knoblochsaue, Regierungspräsidium, pp 43–49, Darmstadt (in German) Schneider E (2003) Organisation longitudinale des forêts alluviales, le cas du Danube. In: Piégay H, Pautou G, Ruffinoni Ch (eds) Les forêts riveraines des cours d’eau, écologie, fonctions et gestionripisylves de l’Europe, chap13, Institut pour le developpement forestier, Paris, pp 272–285 (in French) Schneider E, Tudor MA, Staras M (eds) (2008) Ecological restoration in the Danube Delta Biosphere Reserve, Romania. Evolution of Babina polder after restoration works. WWF Germany, DDNI Tulcea, 80 pp Stadt Ingolstadt (2001) Donau Auenkonzept Neuburg-Ingolstadt. Dynamisierung der Donauauen zwischen Neuburg und Ingolstadt. Informationen zum Leader II-Projekt, pp 8, Ingolstadt (in German) Staras M (1998) Fishery in relation to the environment in the Danube delta Biosphere Reserve.In: Dealing with nature in deltas, Wetland management symposium, Proceedings, RIZA Nota nr. 99.011:157–168, Lelystad/The Netherlands Tockner K, Pennetzdorfer D, Reiner N, Schiemer F, Ward JV (1999) Hydrological connectivity and the exchange of organic matter and nutrients in a dynamic river-floodplain system (Danube, Austria). Freshw Biol 41:521–535 Weiger H (1994) Zum geplanten Ausbau der unteren deutschen Donau – rettet die Donau jetzt. In: Lebensraum Donau – Europäisches Ökosystem, Tagungsdokumentation des internationalen Kolloquiums in Ulm. Beiträge der Akademie für Natur- u. Umweltschutz Baden-Württemberg 17:123–152 (in German) WWF-Auen-Institut (1989) Konzept zur Verbesserung der wasserwirtschaftlichen und ökologischen Situation in der Rheinaue zwischen Rhein-Km 354,5 und Km 359,0. Retentionsraum Bellenkopf, Kastenwörth, Rappenwörth. Endbericht (Final report), Rastatt (in German) WWF-Auen-Institut (1990) Hochwasserschutz am Rhein durch Auenrenaturierung im Hessischen Ried. Teil 2 Landschaftliche Grundlagen und Konzept Auenerweiterung, pp 131 (in German) WWF-Auen-Institut (1992a) Untersuchungen zur Verbesserung der wasserwirtschaftlichen und ökologischen Verhältnisse in der Rheinaue zwischen Rastatt-Wintersdorf und Au a. Rhein. Schlussbericht (Final report), Rastatt, pp 158 (in German) WWF-Auen-Institut (1992b) Studie über die mittelfristige, naturnahe Waldbehandlung zur Renaturierung der Auwälder des Schutzgebietes “Regelsbrunner Au”, Rastatt, pp 39 (in German) WWF-Auen-Institut (1993) Polder Greffern/Söllingen. Untersuchung über die Umweltverträglichkeit. Im Auftrag des Ministeriums für Umwelt des Landes Badeb-Württemberg, Rastatt, pp 364 (in German) WWF-Auen-Institut (1995) Erprobungs- und Erforschungsvorhaben: Naturnahe Auenwälder am Oberrhein – Möglichkeiten der Renaturierung und naturnahen Bewirtschaftung, Scientific report (unpublished) pp 150 WWF-Auen-Institut (1997) Machbarkeitsstudie “Donau-Auen bei Ingolstadt”. Auenrenaturierung an der Donau zwischen den Staustufen Bergheim und Infgolstadt. im Auftrag der Stadt Ingolstadt (Project report), Rastatt, pp 57 (in German) WWF-Auen-Institut (2000) Aufstellung eines integrierten Raumnutzungskonzeptes für die Planung einer Hochwasserrückhaltung in der Hördter Rheinaue. Bericht zu Teilrpojekt 3: Bewertung der Auswirkungen auf Naturhaushalt und Naturschutz. im Auftrag des Landesamtes für Wasserwirtschaft Rheinland-Pfalz, Rastatt, pp 81 WWF Danube-Carpathian-Programme and WWF-Auen-Institut/WWF-Germany (1999) Evaluation of wetlands and floodplain areas in the Danube River Basin. Programme Coordination Unit/Danube Environmental Programme, UNDP/GEF Assistance
Chapter 12
Restoration of Streams in the Agricultural Landscape Lena B.-M. Vought and Jean O. Lacoursière
Abstract The advent of easily-accessible tile-drainage and the intensification of agricultural practices have brought widespread decoupling of streams and their riparian floodplains in agricultural streams over the last 150 years. Channelised and deeply incised in the landscape, these streams are more than often reduced to simple drainage ditches – when not simply running in underground pipes. This paper describes a modular approach to bring back the key functions and ecosystem services provided by lowland streams that still must perform their drainage purposes in an agricultural landscape. Five principal restoration measures, also referred to as “building-blocks”, are discussed: re-creation of buffer-strips, alteration of tile drainage, in-channel interventions, creation of riparian wetlands/ponds, and finally daylightening. The complete restoration of a stream ecosystem may not be the most acceptable option to farming communities. In most cases, however, the application of even the most basic measure (i.e. the buffer-strip) can significantly support the return of key ecosystem services provided by a stream; such as flow regulation, water purification and support to biodiversity. Cumulative implementation of the other measures at strategic points of the drainage basin will further ensure that a functional stream/river valley is reinstated. Keywords Agricultural streams • Stream restoration • Buffer-strips • Riparian wetlands • Ecosystem services
12.1 Introduction Initially proposed in 1990 (Petersen et al. 1990, 1992), the “Building Block Model” described a simple modular approach to bring back the key functions and ecosystem services provided by lowland streams; streams that still must perform their L.B.-M. Vought (*) and J.O. Lacoursière Sustainable Water Management Group, Kristianstad University, 291 88 Kristianstad, Sweden e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_12, © Springer Science+Business Media B.V. 2010
225
226
L.B.-M. Vought and J.O. Lacoursière
drainage purposes in an agricultural landscape. This practical approach, although set up to deal with the deeply-channelised and often imbedded agricultural streams of southern Sweden, has proven itself a useful tool in discussing stream and river restoration with managers, practitioners, the public and students alike.
12.2 Human Influence on Streams in the Agricultural Landscape Historical assessments of ancient stream systems show that many of today’s streams have little in common with those that existed prior to human influence (Sedell and Luchessa 1982; Wolf 1960). In Swedish lowland streams, for example, studies indicate that they were typically meandering, had close contact with their floodplains and passed through extensive areas of riparian wetlands/swamp forests (Wolf 1960). Although gradual changes brought by intense deforestation were initiated over a thousand years ago, it is the last 150 years that has seen the most drastic alterations of entire drainage basins with the advent of easily-accessible tile-drainage. Although intensive land drainage is an old practice going back to the Roman age, early drainage was limited to willow bundles and stone trenches. With the industrialised production of clay tiles starting around 1850, drainage of the land increased dramatically. For example, in Skåne, the southernmost province of Sweden, over 270,000 ha (or 25% of the province) were subsurface-drained during the short period of 1870– 1884 (Möller 1984). Because of the relatively low-relief landscape of the area, the streams receiving water from the drainage-tile networks had to be deepened and straightened to be able to transport the increasing water volume (Wolf 1960). The intensification of agricultural practices further saw the removal of riparian vegetation to the point where, in many locations, streams were seen as mere drainage ditches that needed to be put underground (i.e. running in pipes with manholes set at regular intervals). This brought a widespread decoupling of stream ecosystems and their riparian floodplain; i.e. a severe reduction of stream-land interaction which includes limiting the extent of the hyporheic zone. This also brought local and, in some instances, a regional decrease in groundwater levels. These changes led to: 1. Significant reductions in the nutrient-retention capabilities of streams and their riparian zones. The faster movement of both surface and subsurface waters to, and within a stream, has significantly reduced the self-cleaning capacity of the stream ecosystem; with the open-channels acting as little more than transport ditches, resulting in increased nutrient levels reaching water bodies and ultimately the sea (Kronvang et al. 2005). 2. Decoupling of land-water interactions. Surface runoff often enters the stream directly without passing through a riparian buffer-strip. Hence, large amounts of sediment, particle-bound phosphorus and other pollutants enter the stream during heavy rain or snowmelt (Syversen 2005). In subsurface-drained areas, subsurface flow is intercepted
12 Restoration of Streams in the Agricultural Landscape
227
by the drainage-tile networks and delivered directly to the open stream channel; hence limiting the biogeochemical processes taking place in the hyporheic zone (Vought et al. 1994). These alterations in runoff and subsurface flow patterns further increase the downstream transport of nutrients and other pollutants. 3. Changes in the stream hydrograph. The combined effects of (a) an increase in drainage density associated with subsurface drainage networks, (b) reduction in channel length associated with channelisation, (c) reduction in bank-storage and evapotranspiration potential associated with lowering of the groundwater table, and (d) a decrease in in-channel retention structures associated with the removal of woody-debris and the riffle-pool sequence, has resulted in larger peak-flows and reduced base-flows. 4. Changes in channel stability. Increased peak-flow in rectified channels characterised by steep banks, smooth bottoms and increased reach slope (i.e. shorter channel length for the same vertical drop), results in increased bank erosion and sediment transport. This instability is often compounded by riparian vegetation removal. Conversely, reduced intermediate flow (most of the water now flowing during peak events) and lower base-flow result in increased sediment deposition. These extreme events directly contribute to instability in channel geometry and therefore result in impoverished stream habitats. 5. Increase in solar radiation to the stream. With the removal of riparian trees and bushes, aquatic macrophyte production is enhanced, especially in nutrient-rich agricultural streams. Overall, the resulting dense macrophyte stands further increase sedimentation. Agricultural streams may therefore have to be dredged more often to avoid water saturation of adjacent sub-surface drained fields, as well as prevent local flooding events (further reducing nutrient-retention capacity and in-stream habitat complexity). In addition, increased solar-radiation penetration can cause elevated temperatures during summer, which can be detrimental to invertebrate and vertebrate life (e.g. trout/salmon populations). 6. Depletion of the flora and fauna within and around the stream. Subsurface drainage and channelisation allows riparian areas to be converted to farmland, removing the bio-diverse interface (i.e. the ecotone) between the stream and the surrounding land. In-channel removal of woody debris and rocks (riffles), compounded by the disturbances of high bed-load transport and sedimentation/siltation, further contribute to habitat homogenisation. Furthermore, the natural input of leaves, the main energy inflow (i.e. food source) in small headwater streams, is also significantly reduced. These changes result in lower biodiversity both around and within the stream ecosystem. To counteract these negative effects, stream enhancement projects are taking place in many parts of the world, slowly returning streams back to more functional and natural conditions. In the following sections, five key restoration measures (also referred to as “building-blocks”) are discussed for their capacity to regulate flow, reduce nutrient and sediment transport, and enhance flora/fauna around and within the stream. These are:
228
• • • • •
L.B.-M. Vought and J.O. Lacoursière
Re-creation of buffer-strips (defined as the foundation-block) Alteration of tile-drainage In-channel modifications Creation of riparian wetlands/ponds Daylightening
The actual measures selected will depend on the type of stream, the goals of the restoration project and the economic restrictions.
12.3 Methods for Stream Restoration 12.3.1 Re-creation of Buffer-Strips Developments in freshwater ecology have increasingly emphasised the strong links between terrestrial and freshwater environments and focused attention on the importance of the riparian ecotone: the transition zone between the terrestrial and the aquatic ecosystems. Depending on the study goals and the scientific backgrounds, the riparian ecotones have been referred to as: buffer-strips; riparian vegetated buffer-strips (VBS); vegetated filter strips (VFS); protective zones; shelterbelts; biogeochemical barriers; and the hyporheic zones when referring to sub-surface linkages. In the following text we will use the term buffer-strip to refer to a permanently vegetated zone between the agricultural fields and the stream. Most water entering or leaving a stream passes through the riparian ecotone. It usually does so via one of five pathways: surface runoff; seepage; shallow subsurface flow (i.e. inter-flow); deep subsurface flow (i.e. groundwater recharge/discharge); and through drainage tiles. The unique physical and biogeochemical properties of a riparian ecotone influence the flux of water, nutrients and other exogenous substances from the catchment areas to the stream, as well as within the stream and its immediate surrounding. The riparian ecotone also has a strong influence on a stream’s energy input, since most small headwater streams receive the major part of their energy from leaf-litter (Vannote et al. 1980). The type and diversity of riparian vegetation has a direct influence on the benthic community structure and the life cycle strategies present in agricultural streams. Because the riparian ecotone is an active and critical area of any stream-ecosystem, an appropriate restoration of this area – as for example a grassland (Fig. 12.1b), scrub or native woodland buffer-strip (Fig. 12.1c) – is a particularly valuable part of river restoration (Syversen 2002; Vought et al. 1994).
12.3.1.1 The Effects of Buffer-Strips on a Stream’s Hydrograph The presence of buffer-strips along the stream channel reduces both the volume and timing of runoff water entering a stream; hence lessening peak flows. Interception
12 Restoration of Streams in the Agricultural Landscape
229
Fig. 12.1 Base building-block for restoration: the buffer-strip. (a) As it commonly looks today in the agricultural landscape; (b) a 10 m wide grass buffer-strip; (c) a 10 m wide tree buffer-strip (Drawn by L. Widing)
(i.e. the portion of the rain that is caught by a structure and returned to the atmosphere by evaporation – and therefore never reaches the ground) is increased by leaf-litter, plants and trees, which in turn slow down runoff velocity by increased ground roughness. This roughness-detention (i.e. water held in small topographic irregularities) also favours increased water infiltration (enhanced by root systems and invertebrate activities). The presence of vegetation further reduces the volume of water contributing to the hydrograph through increased evapotranspiration.
12.3.1.2 Nutrient/Pesticide/Sediment Removal in Buffer-Strips One of the most significant effects of the reintroduction of riparian ecotones along the margins of a stream is the reduction of the amount of nutrients, pesticides and sediment entering streams by surface and subsurface flow (Syversen 2005). Sediment and sediment-bound particles enter streams mainly through surface flow. The trapping of sediments and attached pollutants will depend on factors such as slope, water transit
230
L.B.-M. Vought and J.O. Lacoursière
Fig. 12.2 Reduction of total phosphorus in surface runoff and nitrate in subsurface flow (Data compiled from literature)
time, vegetation and season (Syversen 2002, 2005). Literature sources provide fairly consistent results, where most of the phosphorus is removed within a buffer-strip of 10 m width (Fig.12.2). The retention efficiency of total phosphorus in surface runoff is often between 50% and 90%, where the long-term efficiency of P-retention is related to the sorption properties of the soil (Hoffmann et al. 2009). Nitrogen mainly enters streams through subsurface flow and, again, studies show that the major part of the nitrate is removed within a 10 m buffer-strip (Fig.12.2). This efficient removal of nitrate may be explained by the low oxygen levels and the large amount of organic material in a riparian ecotone favouring denitrification (Blicher-Mathiesen 2000; Hoffman 1998; Duff and Triska 1990; Triska et al. 1989, 1993). In addition, water-table elevation seems to control nitrogen cycling in riparian wetlands. Average water-table levels 10–30 cm below the ground surface favour denitrification while, at higher levels, ammonification is the main process. At lower water-table levels, nitrification dominates (Hefting 2003). Pesticides, mainly those binding with particles, such as glyphosate, propiconazole and fenpropimorph, can also be removed from surface runoff through vegetative buffer zones (Syversen 2005). Average removal efficiency was 51% for particles, with 48%, 85% and 34% for glyphosate, propiconazole and fenpropimorph, respectively. Furthermore, no significant difference was found in removal efficiency (%) between winter and summer (Syversen 2005). 12.3.1.3 Forested Buffer-Strips The reintroduction of trees along streams (Fig.12.1c) can be a valuable technique in stream restoration, as extensive root systems can stabilise stream banks, and provide habitats for fish and invertebrates. They also can provide shading and hence reduce water temperatures and in-stream macrophyte growth during summer. Through time, trees will also add large woody debris to the stream; hence creating new habitats and enhancing biodiversity.
12 Restoration of Streams in the Agricultural Landscape
231
Many lowland streams have dense macrophyte vegetation due to high nutrient and light levels. In Denmark, experiments to reduce plant growth suggest that to be effective trees along river banks should be of such extent that light levels are reduced to about half of that in un-shaded sections (Dawson and Kern-Hansen 1979). Since most headwater streams are dependent on leaf litter input as a food source for aquatic invertebrates (Vannote et al. 1980), the size and composition of the riparian forest will influence in-stream processes. Although, decomposition rates and macro-invertebrate functions are not necessarily reduced in streams without riparian vegetation, there is a change in benthic community structure and the life cycle strategies present. The presence or absence of trees along smaller streams affects macro-invertebrate communities, either indirectly through physical changes or temperature alterations, or through a change in detritus source (Sweeny 1993; Vought et al. 1998). Tree species and composition of tree communities further alter the species composition of stream invertebrates. A mixture of tree species will generally lead to a higher diversity and abundance of macro-invertebrates compared to monocultures along stream banks (Vought and Lannerstad 2001). This in turn can stimulate in-stream litter decomposition (Sanpera-Calbet et al. 2009). There are some suggestions that alder (Alnus spp.), one of the most common riparian trees in many parts of Europe and North America, may act as a significant nitrogen source in nutrient-poor streams and lakes (Dugdale and Dugdale 1961). Alder possesses an endophytic actinomycetal fungus in its root nodules and, like leguminous plants, is able to fix nitrogen. Rates of nitrogen fixation of up to 225 kg N/ha/year, or 22.5 g N/m2/year, have been measured (Wetzel 2001). However, it is likely that such rates predominantly occur in riparian areas naturally low in nitrate. Preliminary results from riparian alder woodlands in Sweden, for example, show no evidence of elevated nitrogen levels (Mander and Vought 1997). It seems likely that, in waters with high nitrogen levels, alder uses dissolved nitrogen in preference to the more energy-costly nitrogen coming from nitrogen fixation. 12.3.1.4 The Effects of Buffer-Strips on Fish The reported effects of marginal shading and tree cover on fish are variable. Work on forested streams in Western USA showed that clear-cutting areas adjacent to streams increased both the biomass and the density of cutthroat trout (Salmo clarkii) by a factor of 2 (Aho 1977). This increase was explained by the greater food abundance in the unshaded clear-cut sections (Hawkins et al. 1983). Streams in the USA had, however, very low nutrient levels and very dense tree cover compared to most European streams. Early work done by Boussu (1954) and Burton and Odum (1945) have shown the reverse trend; i.e. that the removal of natural cover decreases populations of trout because it allows summer temperatures to rise too high. Similar results have been obtained by Barton et al. (1985) who, in describing the habitat and physiological requirements of trout, found a 10 m buffer-strip to be optimal for supporting trout populations. Combining information from the literature, the following is suggested for the installation of buffer-strips:
232
L.B.-M. Vought and J.O. Lacoursière
• Considering faunal requirements, nutrient-reduction prerequisites and the farmers’ need for land, a minimum riparian buffer-strip width of 10 m seems a realistic suggestion for agricultural streams. • The vegetation along the buffer-strip should ideally provide moderate shading to the stream. • One of the main problem with bushes and trees is that roots can penetrate and clog drains situated under a buffer-strip. The solution is either to replace the tile drains with a solid pipe, or to dig-up the drain-pipe and replace it with an open ditch. If the drain-tile outflows are far apart, the area above these pipes could be left without trees. The key benefits from restoring buffer-strips are: • Lessening of peak flows through reduction of runoff volume and runoff velocity • Reduction of the amount of nutrients entering the stream channel • Reduction of the amount of sediment entering the stream channel through runoff • Reduction of the amount of pesticides entering the aquatic system through aerosol drift or runoff • Improved stream-bank stability through roots systems and reduced runoff • Decreased stream temperature during warm summers through shading • Decreased light penetration to the stream and thereby reduced macrophyte growth in the stream channel • Enhanced environment for fauna and flora around and within the stream • Wind protection • Increased movement of flora and fauna in the landscape through the creation of corridors Re-establishing buffer-strips can therefore be seen as the foundation to the restoration of key ecosystem services, such as: flow regulation, nutrient cycling, water purification, support to biodiversity, pollination and insect pest control.
12.3.2 Alteration of Tile-Drainage Many agricultural areas in northern Europe and temperate North America were originally developed from floodplain wetlands. To facilitate early seeding, fields have been underlined by drainage-tile networks and stream channels often had to be lowered to ensure efficient water-table draw-down. These drains now carry nutrient-laden waters below the floodplain to empty directly into the streams. In doing so, they effectively create numerous point-sources for nutrient pollution by bypassing the natural filtering systems surrounding streams. Furthermore, by intercepting water that would have recharged the groundwater, these subsurface drainage networks significantly increase the amount of water entering the stream; hence contributing to higher peak flows.
12 Restoration of Streams in the Agricultural Landscape
233
One approach used to decrease these agricultural point-sources of pollution is the opening up of drainage pipes before they enter the stream. This can be done in one of several ways depending on the landscape topography. 12.3.2.1 Opening Pipes Where valley slopes are moderate, pipes can be opened at the valley top-edge to allow drainage water to filter through the riparian zone before entering the stream (Fig.12.3a). They can also be opened so as to allow water to flow to an interception ditch running parallel to the top-edge of the valley, which will distribute the drainage water along the length of the valley and allow it to trickle through the bufferstrip towards the stream. This method of using the riparian buffer-strip as a trickle filter is similar to that of flooded meadows which, for centuries, have taken advantage of the nutrients in stream water to fertilise the land. The main concern using this method is to ensure that the water is dispersed and retained, and that drainage water does not simply run across the floodplain creating preferential flow paths. 12.3.2.2 Riparian Wetland Horseshoes Another solution is to use riparian wetland horseshoes (Fig.12.3b). They are semicircular shaped excavations within the buffer-strip to expose each drainage-tile outflow just before it enters the stream. Generally dug 8 m into the bank, it allows the drainage water to flow through a mini-wetland (i.e. organic-rich soil, hydrophilic vegetation)
Fig. 12.3 Building-blocks for restoration: riparian measures. (a) Drainage tile opened-up at the top of the stream valley; (b) drainage tile opened-up into a horseshoe built within the buffer-strip (Drawn by L. Widing)
234
L.B.-M. Vought and J.O. Lacoursière
before it reaches the stream. An alternative can be a deeper excavation, creating a small pond-like area along the side of the stream. Both methods will enhance the nutrient retention capacity of the riparian area, reducing the impact of pointsource entry of both nitrogen and phosphorus. This method can be used in areas where the topography is too flat to allow pipes to be opened along the edge of the valley. Although each interception area is small, the cumulative effect of each interception along the entire length of a stream can be significant. The effect is greatest during late fall and early spring when most of the nitrogen and phosphorus is exported from catchments (Petersen et al. 1987). The key benefits of intercepting and mitigating point-source drainage from agricultural fields are: • Lessening of peak flows through increased infiltration, evapotranspiration and water detention time • Reduction of the amount of nutrients entering the stream • Stimulation of growth of wetland plants along the stream valley Opening sub-surface drainage networks before they reach a stream further support the restoration of key ecosystem services as previously described, with a significant enhancement of nutrient cycling, water purification and support to biodiversity.
12.3.3 In-channel Modifications In channelised streams, water velocity is high since structures which would have created roughness and detention have been removed (i.e. woody debris, riffles and meanders). Erosion and consequently bank failures are a major source of sediment. Inputs can be so great that channelised water courses might have to be dredged every few years to maintain drainage capacity and flood prevention. Inputs of sediment can also considerably increase phosphorus levels since most phosphorus enters streams bound into, or adhering to, particulate matter. 12.3.3.1 Side Slope Reduction In many channelised streams banks are steep, usually approaching 1:3 (ca. 70°). Reducing these slopes to a minimum of 1:1 (ca. 45°) and stabilising them with vegetation has several benefits (Figs.12.4a and 12.5). Firstly, a reduced bank slope lowers the frequency of bank failure and consequently the amount of soil directly entering the channel. Secondly, slope reduction increases the width of the stream corridor, creating an area that functions like a narrow floodplain. In this area, a reduced water velocity minimises erosion and enhances particle deposition, hence reducing sediment transport into downstream receiving waters.
12 Restoration of Streams in the Agricultural Landscape
235
Fig. 12.4 Building-blocks for restoration: in-channel measures. (a) Side slope reduction; (b) riffle-pool section; (c) meandering section (Drawn by L. Widing)
12.3.3.2 Riffle-Pool Sequences A significant increase in physical complexity in a stream channel is achieved through the return of riffle-pool sequences (Fig.12.4b). The presence of riffles and larger stones in a stream also increases bottom roughness and turbulences; which in turn reduces overall water velocity within the channel. In natural streams, riffles and pools occur at more or less regular intervals, usually with a frequency of five to seven times the stream full-bank width (Madsen 1995). In some cases, riffles have simply been constructed from excess stone material collected from the surrounding agricultural lands. Exact placements are not critical since the stream re-sorts the material over the next 5–10 years. They have long been one of the main stream-management tools for trout and other fishes. Others were provision of shelters, deeper pools and small barriers to diversify the stream bed so as to support a greater abundance of invertebrate prey (Tarzwell 1935). Hynes (1970) highlighted the importance of riffle-pool sequences as one of the basic steps in stream management and regretted that this practice may have gone out of fashion. More recently, as streams have become increasingly valued as landscape features, the importance of riffle-pool systems are again being realised.
236
L.B.-M. Vought and J.O. Lacoursière
Fig. 12.5 Side-slope reduction and wetland-horseshoe measures after excavation on a channelised agricultural stream; Helsingborg Municipality, Southern Sweden (Picture: L. Vought 2004)
12.3.3.3 Recreating Meanders Where streams run across lands of low gradients one of the most stable channel configurations is the meander (Fig. 12.4c). Meanders also contribute to the control of water velocity by a lower slope compared to the rectified channel (i.e. longer channel length for the same vertical drop), as well as providing an increased area of friction (i.e. bed roughness). Reconstruction of meanders is not always easy, particularly since the original straightening of channels was often accompanied by channel deepening. Sometimes the groundwater level can be raised, which greatly simplifies the process. Reconstruction of meanders is normally a major undertaking and requires expert hydrological and geomorphological advice in order to select widths and amplitudes appropriate to the substrate and flow regime. Roughly, the wavelength of a full meander is 10–14 times the full-bank width of the watercourse (Madsen 1995). Where possible, re-establishment of the stream-valley meandering should be considered (Fig.12.6). The key benefits of in-channel modifications are: • • • •
Lessening of peak flows through reduction in water velocity Decreased bank erosion Decreased suspended load by sedimentation on the floodplain Increased retention of large particulate organic matter due to increased stream complexity, which increases available food for macro-invertebrates • Increased complexity of the stream bottom which enhances habitats for macroinvertebrate and fish fauna • Increased water exchange with the hyporheic zones through creation of riffles and meanders • Increased water retention time in the stream valley
12 Restoration of Streams in the Agricultural Landscape
237
Fig. 12.6 Daylightening of an underground stream, Klammersbäck, Simrishamn Municipality, Southern Sweden. (a) Meandering and riffle-pool sequences after excavation in the upper-part of the 1.3 km restored section; (b) downstream-end of the restored section where the end-riffle (right) returns the restored stream to the level of the actual channelised stretch where the old stream-pipe outflow (left) remain active as a security measure (Pictures: L. Vought & J.O. Lacoursière 2002)
Side-slope reduction and reintroduction of riffle-pool sequences and meandering further support the restoration of key ecosystem services with a significant enhancement of flow regulation, nutrient cycling, water purification and support to biodiversity.
12.3.4 Creation of Riparian Wetlands and Ponds Wetlands and ponds created along the stream valley are an economical and multipurpose restoration measure (Fig.12.7a, b). They increase hydraulic storage, retention time and sediment trapping, which significantly reduces particle-bound phosphorus. In addition, organic material build-up enhances denitrification in these areas.
238
L.B.-M. Vought and J.O. Lacoursière
Fig. 12.7 Building-blocks for restoration. (a) Detention pond; (b) riparian wetland/swamp-forest (drawn by L. Widing)
Quite often it is not possible to increase the groundwater level to the required height since the impact would be too large on the surrounding fields. Instead, ponds can be excavated along the stream where stream water can be diverted from an upstream location and returned to the stream after residing some time in the pond (Figs.12.7a,12.8). In addition, such ponds can also collect water from surrounding sub-surface drainage networks. In southern Sweden, these types of ponds are commonly used to combat the eutrophication of not only the receiving water bodies, but also of the Baltic Sea. Compilation of data available for these ponds show that, although highly variable, the average retention value for nitrogen amounts to 1,770 kg N/ha/year and average retention value for phosphorous amounts to 124 kg P/ha/year (Vought and Lacoursière 2000). Many channelised agricultural streams harbour riparian fields that are seasonally wet, difficult to plough, or less valuable as farmland due to the mineralisation of peaty soil. These areas, often of significant size, are usually indications of former wetlands or swamp forests. When restored, they are valuable sites for peakflow attenuation, nutrient retention and biodiversity enhancement (Fig. 12.7b). This can be done by raising the stream bottom or by setting a series of riffles, which in turn raises the groundwater table. The height of the stream bottom and riffle crests must however be set to ensure that upstream and lateral increases in groundwater levels do not permanently impact the adjacent fields. Vegetation in the restored area promotes a rapid build-up of organic matter which, together with the low oxygen levels associated with water-saturated soils, provides conditions ideal for denitrification. The use of riffles also promotes exchanges with the surrounding floodplain, thus sustaining a hyporheic zone (both around and within the
12 Restoration of Streams in the Agricultural Landscape
239
Fig. 12.8 Riparian wetland/pond system set parallel to a small agricultural stream, Hörby Municipality, Southern Sweden. (a) Channelised stream diversion-wetland inflow; (b) sediment trap; (c) first wetland; (d) connective riffle; (e) second wetland; and (f) wetland outflow-return of treated water to the channelised stream (Picture: K. Olsson 2001)
stream channel) which can be an extremely active site for denitrification processes (Hoffmann 1998). The other benefit of wetland/pond creation is the local increase in biodiversity in a landscape that will have seen drastic reductions in flora and fauna. When strategically combined with vegetated buffer-strips, both wetlands and ponds significantly contribute to the effectiveness of corridors in maintaining biodiversity at a regional level (Tiere 2009). Considering nutrient cycles as a whole, the most cost-effective place to create wetlands would be in the lower part of a drainage basin (close to the sea in the case of southern Sweden) where the nutrient concentrations are at their highest. However, given the magnitude of the problem, a wetland corridor along the stream may be more effective and have considerably greater overall benefits. Wetlands and ponds placed in headwaters can reduce peak flow and increase base-flow through groundwater recharge (i.e. retain water for dry periods). Key benefits of the wetlands and ponds are: • Lessening of peak flows through increased hydraulic storage and reduction in outflow discharge by increased evapotranspiration • Reduction of phosphorous through sedimentation of fine particulate matter • Reduction in nitrogen through increased denitrification • Creation of habitats suitable for wildlife such as ducks, other birds, frogs and fish • An increase in the length of time that water remains in the valley which will aid the self-cleaning capacity of the stream ecosystem • Growth of wetland plants and development of wetland habitat which will improve the overall aesthetic value of the valley The re-establishment or creation of wetlands/ponds within the drainage basin further supports the restoration of key ecosystem services, with a significant
240
L.B.-M. Vought and J.O. Lacoursière
enhancement of flow regulation, nutrient cycling, water purification, support to biodiversity and landscape enhancement.
12.3.5 Daylightening As mentioned in the introduction, many agricultural streams have been put underground. Lately, however, with the increasing awareness among the farming community, changes in land use or a need for replacement of old pipes, some of these streams are being reopened (called daylightening i.e. brought back to daylight). Although some streams are left deeply incised in the landscape (Fig. 12.1a), the stream is normally brought back closer to its original bottom and riffles-pools are created (Fig. 12.4b,) with often some level of meandering (Fig. 12.4c). In some instances, the former channel is dug out and the stream regains its original path (Fig. 12.6a). In some cases, the pipe is left intact (Fig. 12.6b), and measures are taken to avoid water losses from the recreated stream channel. The reason for this can either be as a security measure for the farmers where, if water saturation of the restored floodplains should impact too much the adjacent active fields the pipe could be partially re-activated, or to ease the legal aspects regarding water courses as it is often a problem in agricultural areas of Sweden (i.e. where the bottom of a channelised stream is set in a legal document). Key benefits of daylightening are: • Restoration of the stream-flow regulation function • Restoration of the stream self-cleaning capacity • Re-establishment of the stream as a functional ecosystem When the starting condition of a stream restoration is an underground channel, daylightening must be seen as the first step for the restoration of key ecosystem services supplied by an agricultural stream.
12.4 Concluding Remarks Since in most agricultural landscapes lowland streams must still perform their drainage function, the complete restoration of a stream ecosystem may not be the most acceptable option. In most cases however, it is possible to significantly support the return of ecosystem services even if only the most basic of the measures presented here (i.e. the buffer-strip) is applied along the length of the stream. Cumulative implementation of the other measures at strategic points of the drainage basin (e.g. at locations of high nutrient levels, high runoff, high erosion) will further ensure that a functional stream /river valley is reinstated.
12 Restoration of Streams in the Agricultural Landscape
241
References Aho RS (1977) A population study of the cutthroat trout in a shaded section of stream. M.S. thesis, Oregon State University, Corvallis, 87 pp Barton DR, Taylor WD, Biette RM (1985) Dimensions of riparian buffer-strips required to maintain trout habitat in southern Ontario streams. N Am J Fish Manage 5:364–378 Blicher-Mathiesen G (2000) Nitrogen removal in riparian areas. Ph.D. thesis, National Environmental Research Institute, Silkeborg, Denmark, 94 pp Boussu MF (1954) Relationship between trout populations and cover on a small stream. J Wildlife Manage 18:229–239 Burton GW, Odum EP (1945) The distribution of stream fish in the vicinity of Mountain Lake, Virginia. Ecology 26:182–194 Dawson FH, Kern-Hansen U (1979) The effect of natural and artificial shade on the macrophytes of lowland streams and the use of shade as a management technique. Int Revue ges Hydrobiol 64:437–455 Duff JH, Triska FJ (1990) Denitrification in sediments from the hyporheic zone adjacent to a small forested stream. Can J Fish Aquat Sci 47:1140–1147 Hawkins CP, Murphy ML, Anderson NH, Wilzbach MA (1983) Density of fish and salamanders in relation to riparian canopy and physical habitat in streams of the northwestern United States. Can J Fish Aquat Sci 40:1173–1185 Hefting MM (2003) Nitrogen transformation and retention in riparian buffer zones. Ph.D. thesis, Utrecht University, The Netherlands, 200 pp Hoffmann CC (1998) Nutrient Retention in Wet Meadows and Fens. Ph.D. thesis, National Environmental Research Institute, Silkeborg, Denmark, 134 pp Hoffmann CC, Kjaergaard C, Uusi-Kämppä J, Hansen HCB, Kronvang B (2009) Phosphorus retention in riparian buffers: review of their efficiency. J Environ Qual 38:1942–1955 Hynes HBN (1970) The ecology of running waters. University of Toronto Press, Toronto, 555 pp Kronvang B, Jeppsen E, Conley DJ, Søndergaard M, Larsen SE, Ovesen NB, Carstensen J (2005) Nutrient pressures and ecological responses to nutrient loading reductions in Danish streams, lakes and coastal waters. J Hydrol 304:274–288 Madsen BL (1995) Danish Watercourses/Ten Years with the New Watercourse Act. Miljønytt nr. 11. Danish Environmental Protection Agency, 208 pp Möller J (1984) Dikning i Skåne. Ale 2:14–28 Petersen RC, Petersen LB-M, Lacoursière JO (1990) Restoration of lowland streams: the building block model. Vatten 46:244–249 Petersen RC, Petersen LB-M, Lacoursière JO (1992) A building-block model for stream restoration. In: Boon PJ, Calow P, Petts GE (eds) River conservation and management. Wiley, Chichester, pp 293–309 Petersen RC, Madsen BL, Wilzbach MA, Magadza CHD, Paarlberg A, Kullberg A, Cummins KW (1987) Stream management. Emerging global similarities. Ambio 6:166–179 Sanpera-Calbet I, Lecerf A, Chauvet E (2009) Leaf diversity influences in-stream litter decomposition through effects on shredders. Freshw Biol 54:1671–1682 Sedell JR, Luchessa KJ (1982) Using the historical record as an aid to salmonid habitat enhancement. In Armantrout NB (ed) Symposium on Acquisition and Utilization of Aquatic Habitat Inventory Information. Proceedings of a western division American Fisheries Society, Bethesda, MA, pp 244–255 Sweeny B (1993) Effects of streamside vegetation on macroinvertebrate communities of White Clay Creek in Eastern North America. Proc Acad Nat Sci Phil 144:291–340 Syversen N (2002) Cold-climate vegetative buffer zones as filters for surface agricultural runoff. Ph.D. thesis, Agricultural University of Norway, pp. 12 Syversen N (2005) Cold-climate vegetative buffer zones as pesticide-filters for surface runoff. Water Sci Technol 51:63–71 Tarzwell CM (1935) Progress in lake and stream improvement. Trans Am Game Conf 21:119–134
242
L.B.-M. Vought and J.O. Lacoursière
Tiere G (2009) Biodiversity and ecosystem functioning in created agricultural wetlands. Ph.D. thesis, Department of Ecology, Lund Unversity, ISBN 978-91-7105-295-7 Triska FJ, Kennedy VC, Avanzino RJ, Zellweger GW, Bencala KE (1989) Retention and transport of nutrients in a third-order stream in northwestern California: hyporheic processes. Ecology 70:1893–1905 Triska FJ, Duff JH, Avanzino RJ (1993) Patterns of hydrological exchange and nutrient transformation in the hyporheic zone of a gravel-bottom stream: examining terrestrial – aquatic linkages. Freshw Biol 29:259–274 Vannote RL, Minshall GW, Cummins KW, Sedell JR, Cushing CE (1980) The river continuum concept. Can J Fish Aquat Sci 37:130–137 Wetzel RG (2001) Limnology; lake and river ecosystems, 3rd edn. Academic/Elsevier, San Diego, CA, 1006 pp Wolf Ph (1960) Land drainage and its dangers as experienced in Sweden. A study in soil erosion with particular reference to agriculture and fisheries. John Sherratt, Altrincham, 73 pp Vought LB-M, Dahl J, Lauge Pedersen C, Lacoursière JO (1994) Nutrient retention in riparian ecotones. Ambio 23:342–348 Vought LB-M, Kullberg A, Petersen RC Jr (1998) The effect of riparian structure, temperature and channel morphology on detritus processing in channelized and natural woodland streams in southern Sweden. Aquat Conserv Mar Freshw Ecosyst 8:272–285 Vought LB-M, Lacoursière JO (2000) Constructed wetlands for treatment of polluted waters: Swedish experiences. In: Mander Ü, Jenssen PD (eds) Constructed wetlands for wastewater treatment in cold climates. Wessex Institute of Technology Press, Ashurst, New Forest Vought LB-M, Lannerstad M (2001) The structure of the riparian ecotone and its implication for stream macroinvertebrate community. Verh Int Verein Limnol 27:1357–1360
Chapter 13
Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands: A Quantitative Assessment S. van der Schaaf, M.J. van der Ploeg, S.H. Vuurens, and M.M.J. ten Heggeler
Abstract The development of the acrotelm on two Irish Midland raised bogs, Raheenmore Bog and Clara Bog East, where drains were blocked in 1995–1997 is described. The acrotelm of both bogs was surveyed in 1991 and 1992 before the blocking and again in late 2002 and early 2003, when effects seemed visually clear. To quantify the effects, both acrotelm depth and transmissivity were measured and compared in Raheenmore Bog and transmissivity in Clara Bog East. Because acrotelm transmissivity depends on the flow rate, a direct comparison was not possible. It was made by comparing field data with theoretical values based on upstream flow path length, surface slope and specific discharge of water from the bog. The results show a significant positive effect on both acrotelm depth and acrotelm transmissivity. Keywords Bog hydrology • Acrotelm • Transmissivity • Bog restoration • Ireland
13.1 Introduction This chapter describes the results of hydrological fieldwork carried out in the winter of 2002/2003 and a comparison with similar results obtained in 1991 and 1992 to asses acrotelm recovery on Clara Bog East and Raheenmore Bog, Co. Offaly, Ireland (Fig. 13.1) resulting from drain blocking measures, taken around 1995/1996. Clara Bog and Raheenmore Bog are two of the few bogs in the Irish Midlands that still have many features of a living raised bog system. However, peat cutting along their margins and internal drainage have caused damage, both to the bogs themselves and to the laggs. The laggs have been destroyed almost entirely. Most other bogs in
S. van der Schaaf (*), M.J. van der Ploeg, S.H. Vuurens, and M.M.J. ten Heggeler Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700 AA Wageningen, The Netherlands e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_13, © Springer Science+Business Media B.V. 2010
243
244
S. van der Schaaf et al.
Fig. 13.1 Location of Clara Bog and Raheenmore Bog in Ireland
Raheenmore Bog Clara Bog
the Midlands are either being extracted or have already disappeared entirely as a result of industrial peat extraction. Conservation methods consisted mainly of drain blocking and in the case of Raheenmore Bog the construction of a few larger peat dams, two of which collapsed within a few years after construction. The third remaining one, being the last one made, still holds. The works were carried out from 1995 to 1997. From field observations it seemed that on both Clara Bog East and on Raheenmore Bog the acrotelm had recovered during 1996–2002, because a layer of fresh Sphagnum material of an estimated average thickness of 10–15 cm had formed since the measures had been taken, whereas much less change had been observed during previous years.
13.2 Description of the Bogs 13.2.1 Size, Damage and Restoration Methods Clara Bog comprises nearly 500 ha of bog. It is bisected into Clara Bog East and Clara Bog West by a road, which was probably built in the early nineteenth century. It has caused considerable subsidence over large parts of the bog (Van der Schaaf 1999, 2000). Clara Bog East was drained for industrial peat extraction in the early 1980s. This drainage system consisted of approximately 60 cm deep parallel drains, spaced about 15 m. Clara Bog became an official nature reserve in 1987. Hence industrial peat extraction never took place. Instead, the drains were blocked
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
245
provisionally between 1987 and 1989. A survey carried out in 1990–1993 in the framework of the Irish–Dutch bog project showed a severe decay of the acrotelm (Van der Cruysen et al. 1993; Van der Schaaf 1996). A more effective drain blocking operation was carried out in 1995/1996 by cutting peat monoliths, mostly from the drain itself, which thus was locally widened, and putting them into the drain with the vegetation side upwards to get the least permeable part of the block on the ditch bottom. The horizontal distances between successive blocks were such that the surface level between them would change by no more than about 10 cm, in order to keep the system effective in more sloping parts of the bog. Most of the design was made by Jan Streefkerk of Staatsbosbeheer, The Netherlands. Raheenmore Bog comprises about 130 ha of bog. In the northeast, it has been drained by a V-shaped system of drains (Fig. 13.2), which is probably between 100 and 200 years old. In 1989, the drains had terrestrialised to a large extent. However, the young Sphagnum peat in the drains had such a large hydraulic conductivity that a considerable discharge still occurred, particularly during wet periods. The system was blocked in 1995–1996. The blocking technique was similar to the one applied on Clara Bog East, except that most blocks were taken rather randomly from the bog, leaving artificial deep pools. In 2002, most of the pools were terrestrialising rapidly. Some additional damage may have been caused by a margin drain which lies around almost the entire bog. It could not be blocked because it also discharges water from surrounding agricultural land. Damage had also occurred as a result of a private peat cutting in the north that protruded into the bog. A large peat dam was built there in 1997 as part of the restoration measures (Fig. 13.2). A survey of the acrotelm in 1991 (Van ‘t Hullenaar and Ten Kate 1991) showed that it was in a rather poor state. A comparison of their results with a theoretical potential situation by Van der Schaaf (1999) confirmed their conclusion.
N g dam
Blockin
Partly terrestrialised drains (now blocked) Raheenmore Bog
500 m
Fig. 13.2 Raheenmore Bog with blocking dam and position of old terrestrialised drains
246
S. van der Schaaf et al.
The underlying theory will be developed in Section 13.3.2. The bog originally had a large number of small pools (up to about 40 m2), mostly in its central part. In 1989 all the pools had terrestrialised.
13.2.2 Position and Type Clara Bog lies between approximately 55–61 m above mean sea level, Raheenmore Bog at 100–107 m. Both bogs are located in the atlantic-subatlantic ombrotrophic mire province (IV.1a) as defined by Succow and Joosten (2001), but do not meet their description of being flat and shallow. Raheenmore Bog is clearly convex and has a largest peat depth of about 15 m (Van der Schaaf 1999) and Clara Bog has a peat depth of about 10.50 m over much of its area (Bloetjes and Van der Meer 1992). In this sense, both Clara Bog and Raheenmore Bog rather resemble the bogs of Succow and Joosten’s area IV.1c, which is the area of ombrotrophic bogs in the Netherlands and northwestern Germany. Although older, the Irish classification (Moore 1964; Hammond 1981) seems more adequate. In this classification, both bogs belong to the True Midland sub-type of Irish Midland raised bogs. The other sub-type is called Transitional. It is transitional between raised bog and blanket bog and should not be confused with the mire type that is transitional between fen and bog. The distinction between the two Irish sub-types is based on climate, bog morphology and botanical composition. The 1,000 mm per annum isohyet serves as an approximate geographic boundary, with the True Midland sub-type on the drier side and the Transitional sub-type on the wetter. Indicative of the True Midland sub-type are Vaccinium oxycoccus and Andromeda polyfolia, whereas the Transitional type has Pleurozia purpurea and Campilopus atrovirens as distinguishing species. True Midland raised bogs are clearly convex. In their natural situation they usually have a rather steep margin of 1–3 m high and a relatively flat central area.
13.2.3 Meteorological Setting The mean temperatures in the area of Clara and Raheenmore Bog over 1951 through 1980 are 4.5°C for January and 15.3°C for July (Rohan 1986). This shows that seasonal variation is relatively small with little winter frost and cool summers. Mean precipitation data are available of the stations Birr, nearest to Clara Bog, and Mullingar, nearest to Raheenmore Bog. An estimation for Clara Bog and Raheenmore Bog, based on data series over 1951–1980 of Birr and Mullingar and data series over 1989–1993 from the same stations and from direct measurements on the bogs, yielded an estimated long-term mean of 842 mm for Clara Bog and 956 mm for Raheenmore Bog (Van der Schaaf 1999, 2002). Figure 13.3 shows monthly precipitation means of the stations Birr and Mullingar. The rainfall is distributed rather evenly over the year, with the smallest sums in late winter and spring and the largest values in the autumn and the first half of the winter.
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
247
120
Birr Mullingar
100
mm
80 60 40 20
ay Ju ne Ju l A y Se ugu pt st em b O er ct o N ov ber em D ec ber em be r
ril
M
ch
Ap
ry
ar M
ua br
Fe
Ja n
ua
ry
0
Fig. 13.3 Mean monthly rainfall sums for the stations Birr and Mullingar, based on data over 1951–1980
120 100
Birr Mullingar
80
mm
60 40 20 0 −20
Ja
nu Fe ary br ua ry M ar ch Ap ril M ay Ju ne Ju l A y Se ugu pt st em b O er ct N obe ov em r D ec ber em be r
−40
Fig. 13.4 Excess precipitation means for Birr and Mullingar (Van der Schaaf 1999, 2002)
Figure 13.4 gives monthly means of the excess precipitation for the same stations. It shows a small deficit for May through July. The applied potential evapotranspiration is equal to 0.75× the result of the unmodified Penman equation and is supposed to apply to grassland (Connaughton 1967). This climate with relatively small seasonal variations and a small mean precipitation deficit in May through July is typical for
248
S. van der Schaaf et al.
the part of the Irish Midlands that lies roughly east of the river Shannon. Further west, there is usually no mean precipitation deficit for any month.
13.2.4 Geological Setting The Irish Midlands have a young glacial landscape that has been shaped in the Weichselian glaciation, in Ireland known as Midlandian. The underlying bedrock is mostly limestone and calcareous shale of Lower Carboniferous age (Warren et al. 2002). Below and around both bogs, the limestone is covered by glacial till of varying granular composition and thickness ranging from a few meters to about 20 m. Midland bogs have mostly developed from terrestrialisation of glacial lakes, formed in local basins after the retreat of the land-ice. Hence, the bogs are partly underlain by lacustroglacial clay, often with lake marl containing shell remnants between the clay and the lower boundary of the peat. The lower part of the peat is a rich fen peat and the upper part is bog peat. The transition in bog profiles is usually rather sharp, often within 1–2 dm. The proportions of bog and fen peat may differ between bogs. In Raheenmore Bog, the fen peat comprises 20% of the total amount of (dry) organic matter in the centre to about 50% at the margins. In Clara Bog these values are 20–50% and about 50%, respectively (Van der Schaaf 1999).
13.2.5 Microtopography, Vegetation, Ecotopes Probably due to the rather even climatic conditions, the differentiation of the bog surface into hummocks, hollows and pools is not extreme. Hummocks and hollows mostly have diameters up to a few metres. The pattern of hummocks and hollows is approximately random and there is usually no clear orientation related to flow direction. Differences in surface level between hummocks and hollows are mostly less than 40 cm, although locally higher hummocks of up to 1 m occur. A feature of Clara Bog is the occurrence of so-called soaks, which are excessively wet areas with vegetation that indicates minerotrophic influences. This was an important reason for co-ordinated research work in this bog, which began in 1989. The vegetation of both bogs was described in-depth by Kelly and Schouten (2002). They distinguished two types of soaks: minerotrophic and rheotrophic. The latter occurs on Clara Bog West and is probably the result of subsidence caused by drainage associated with the bog road (Van der Schaaf 1999). Clara Bog East has some larger minerotrophic soak systems, one with a bog lake, Lough Roe, which has now fully terrestrialised. Indicative species of the minerotrophic soak ecotope are for example Myrica Gale, Empetrum nigrum, Carex rostrata, Menyanthes trifoliata and Drepanocladus fluitans. Measures to restore the original situation are under investigation. Clara Bog East also has a series of smaller pools, some of them several m deep. Many are now terrestrialising.
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
249
Kelly and Schouten compared vegetation types with ecotopes, which were based on hydrology and surface morphology and found a good relationship. Their most important ecotopes of the high bog and assumed position on a schematised bog dome, but not including the soak and soak-related ones, are shown in Fig. 13.5 and described in Table 13.1. Central
Sub-central
Sub-marginal
Facebank
Marginal
Fig. 13.5 Schematic positions of high bog ecotopes (descriptions in Table 13.1) Table 13.1 Description of ecotopes shown in Fig. 13.5 (Van der Schaaf and Streefkerk 2003, based on Kelly and Schouten 2002) Ecotopes Properties Face bank Abiotic No hummocks or hollows. Mean phreatic level up to about 1 m below the surface, seasonal fluctuation several dm. No inundation, except by surface runoff at peak rainfall. No functioning acrotelm. Biotic No peat forming plant communities; vegetation usually dominated by Calluna vulgaris. Marginal Abiotic No hummocks or hollows. Mean phreatic level usually 1–4 dm below the surface, seasonal fluctuation up to 3–4 dm. No inundation, except some by surface runoff at peak rainfall. Acrotelm absent or poorly developed (£5 cm deep). Biotic Few or no peat forming plant communities; vegetation dominated by Calluna vulgaris and Scirpus caespitosus. Sub-marginal Abiotic Some differentiation between hummocks and hollows; hollows inundated up to 5% of the year. Mean phreatic level up to about 1 dm below the surface in the hollows; seasonal fluctuations up to 3–3.5 dm. Acrotelm absent or thin, nearly always £10 cm. Biotic Hollows dominated by Narthecium ossifragum and Sphagnum tenellum; hummocks resemble those in Central and Sub-central ecotopes. Sub-central Abiotic A microtopography of hummocks, hollows and lawns, no pools. Lawns are dominant. Inundation of lowest parts of lawns and hollows up to 70% of the year. Mean phreatic level around or a few cm below the average lawn surface; seasonal fluctuations generally about 2 dm. Acrotelm depth variable; locally well developed, up to about 4 dm deep, but also locally absent. Biotic Lawns dominated by Sphagnum magellanicum; hummocks with Calluna vulgaris, S. capillifolium, S. magellanicum and S. fuscum, locally S. imbricatum. Abiotic A microtopography of hummocks, hollows and pools. Seasonal Central fluctuations of the phreatic level up to about 2 dm. Acrotelm moderately to well developed; depths up to 50 cm, rarely absent. Biotic Hollows and some pools dominated by Sphagnum cuspidatum; hummocks as in Sub-central ecotope.
250
S. van der Schaaf et al.
13.3 Methods 13.3.1 Introduction The work described is mainly based on the theory that the transmissivity of an acrotelm adjusts itself to flow rate and surface slope. It was published in Russian in the 1950s by Ivanov and partly in English later (Ivanov 1965, 1981). It is derived in Section 13.3.2. The acrotelm transmissivity itself was measured using the method described by Van der Schaaf (2004). Acrotelm depth was also measured. The applied definition is given in Section 13.3.8. All data were compared with results obtained earlier in 1991–1993 from the same methods.
13.3.2 Theory Let us consider a section of flow path on a bog, e.g. as shown in Fig. 13.6. The time-invariant differential equation for the acrotelm discharge Qa [L3T−1] through the acrotelm at a position s [L] along a flow path of width w(s) [L], assuming a spatially variant specific discharge1 U, is dQa = Uw( s) ds
(13.1)
dH Now let us introduce the hydraulic head H [L] and the hydraulic gradient ds dH [1] by applying Darcy’s law, written as Qa = −Ta w( s) , where Ta is the acrotelm ds
Q aL
Qa0
w0
w
L
s
s=0
Fig. 13.6 Flow path section
1
Discharge in volume per area per time, which is equivalent to length per time, e.g. mm d-1.
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
251
transmissivity2 [L2T−1]. H decreases with increasing s, hence the minus sign. By writing Qa 0 at s = 0 as the product of path width w0 and flux per flow path width qa 0 [L2T−1], so Qa 0 = w0 qa 0 , we obtain the solution for Eq. (13.1): L
w0 qa0 + ∫ Uw(s )ds dH 0 =− ds wL Ta
(13.2)
In an acrotelm, H is practically equal to the level of the water table, or briefly dH phreatic level. Hence, is practically equal to the slope of the phreatic level,which ds dH ≈ − I . Note that in turn is almost equal to the surface slope I [1], so we can write ds I is defined positive. Substituting dH in Eq. (13.2) and placing Ta explicit yields: ds L
Ta =
w0 qa0 + ∫ Uw(s )ds 0
wL I
(13.3)
Equation (13.3) demonstrates that acrotelm transmissivity is a function of flow rate and surface slope. The mechanism behind this is the fluctuating water table and the strong upward increase of the hydraulic conductivity in the acrotelm (Ivanov 1957, cited by Romanov 1968). With a rising water table, more water will flow through the more conductive upper part of the acrotelm and with a falling water table the upper part of the acrotelm becomes unavailable to the flow. For Raheenmore Bog and Clara Bog, a change in Ta by up to two orders of magnitude at level fluctuations of only 15 cm was found (Van der Schaaf 1999). If the flow rate is too large to be accommodated in the acrotelm, surface flow will partially take over the discharge process. In practical applications, Eq. (13.3) is not very useful. However, it can be simplified easily. If U is averaged over the flow path upstream of s = L, one obtains the mean specific discharge va produced over the path area where 0 ≤ s ≤ L. Now va may be written before the integral. The remaining part of the integral is identical to the area of the flow path where 0 ≤ s ≤ L. If we let the flow path begin at the water divide, i.e. where Qa 0 = 0 , we obtain L
Ta =
va ∫ w(s )ds 0
wL I
=
va Au wL I
(13.4)
where Au [L2] is the area of the flow path upstream of s = L until the water divide. For approximately parallel or radially diverging flow, Eq. (13.4) can be simplified
2 Transmissivity is the integral of hydraulic conductivity k over the depth of the aquifer, often, but incorrectly, described as the product of k and aquifer depth (kD).
252
S. van der Schaaf et al.
further, because for parallel flow Au = LwL and for radially diverging flow Au = 1 LwL . Thus we obtain the general expression 2
Ta =
va Lu fI
(13.5)
where the suffix u expresses that Lu it is the full upstream flow path length from the water divide and f [1] expresses the flow pattern. For parallel flow f = 1 and for radially diverging flow f = 2. For flow patterns intermediate between both, intermediate values for f should be used. For converging flow 0 < f < 1. From Eq. (13.4) or (13.5) a theoretical value of Ta can now be estimated for each specific discharge at known va, I and flow pattern. In theory, a simple comparison with measured Ta shows to what extent the real acrotelm transmissivity approaches the theoretical value. In practice, this is rather more complicated, as the results of the measurements on Raheenmore and Clara Bog will show.
13.3.3 Surface Slope The surface levels of both bogs have been surveyed in a 100 × 100 m grid in 1991/1992. The survey was repeated in the winter of 2002/2003. For both Clara Bog East and Raheenmore Bog the differences between the results of both surveys were negligible. As severe subsidence had occurred on Clara Bog East previously (Van der Schaaf 1999), this result shows that the drain blocking had a positive effect on surface level stability.
13.3.4 Rainfall On both Clara Bog East and Raheenmore Bog a recording raingauge was installed. Both were of the tipping bucket type with a resolution of 0.1 mm and a recording interval of 30 min.
13.3.5 Phreatic Levels Phreatic levels were recorded at four sites on Raheenmore Bog (Fig. 13.7) and at three on Clara Bog East (Fig. 13.8), using transducer-based dataloggers with a 30-min recording interval.
10 10 2.5 2.0
Measuring weir & collector drain
Recording wells
J6
253
100.0
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
Rainfall recorder
107.0
0 4. 10 .5 3 10 .0 3 10
105.0
10 6.0
.5
N
106
P6 106.5
5.5
O6
10
106.0 105.5 10 5 10 . 0 4. 5
10
5 4.
L13 K6 L11 L12 L9L10 L8 L7 L5 L6 L3 L4 L2 L1 M6 L-1 L0 N6
Q6
500 m
Bog pools
Aa
59
.5
Ab
Drain
5
Rainfall recorder
60.
Bog
road
Fig. 13.7 Raheenmore Bog with surface level contours (dotted ) in m above sea level, position of the measuring weir, recording wells and the acrotelm transmissivity sites (open circles)
.0
60
Bc
59.5
58.5
59.0
58.0
Drain
Bb
Recording wells
Drain
Bog pools
Lough Roe
Cc Cb
Bog pools
.0
59
N
.0
58
59.0
58.5 58.0
500 m
Fig. 13.8 Clara Bog East with surface level contours (dotted ) in m above sea level, positions of the recording wells and the acrotelm transmissivity measuring sites (open circles)
254
S. van der Schaaf et al.
13.3.6 Discharge For application of the theory developed in Section 13.3.2, information on specific discharge is required. The discharge could only be measured on Raheenmore Bog, where a 90° Rossum type V-notch measuring weir (Boiten 1985) with stage recorder was installed. From the previous fieldwork period in 1989–1993, its position on the bog margin was known as an outlet of an internal bog catchment of 28 ha (Van der Schaaf 1999). The collector drain of 1989 was still available, but blocked. The drain blocks were left in place, but opened using PVC pipes that could later be closed with caps. The mean specific discharge va was calculated from the discharge over the weir and the catchment size. In comparisons of acrotelm transmissivity Ta, obtained at different spots, this implied the somewhat questionable assumption of a spatially invariant U. Because at each site several measurements of Ta at different discharge conditions were averaged, differences caused by spatial variation of U were unlikely to have a considerable effect on the result. On Clara Bog East, the discharge could not be measured directly. Therefore an indirect method was applied. Rainfall data of Raheenmore Bog and Clara Bog East, recorded with a tipping bucket rain gauge at 30-min intervals, were compared. From the discharge data of Raheenmore Bog the delay between rainfall peaks and discharge peaks was derived. From rainfall and phreatic level data from automatic recordings with 30-min intervals, it could be concluded that the delays between rainfall and water level peaks hardly differed between Raheenmore Bog and Clara Bog East (Ten Heggeler et al. 2005). The fall of the phreatic level after peaks suggested that the tail recession of the discharge from on Clara Bog East would be rather similar to the one of Raheenmore Bog. In this way an acceptable estimate of va on Clara Bog East was obtained from phreatic levels of Clara Bog and discharge data of Raheenmore Bog.
13.3.7 Acrotelm Transmissivity The measuring sites for acrotelm transmissivity on both bogs were the same as during the fieldwork in 1989–1993 to obtain maximum comparability. The measuring method was as described by Van der Schaaf (2004). On Raheenmore Bog each site consisted of two square pits of approximately 25 × 25 × 40 cm deep, one as much as possible in a hollow position and one in a hummock position. In 1991 only one pit per site was installed. The sites were located in two perpendicular transects (L1–L13 and J6–Q6, Fig. 13.7). On Clara Bog East only six sites were available (Fig. 13.8). These were earlier test sites of the fieldwork of 1989–1993 and had been selected on wetness, where sites ‘A’ were wet, sites ‘C’ were relatively dry and sites ‘B’ were of intermediate wetness. At each site shown in Fig. 13.8, six measuring pits were installed in an area of approximately 12 × 12 m2, three as much as possible in a hollow and three in a hummock. The same set-up was applied in 1992.
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
255
Transmissivity measurements were done four times on Clara Bog East and five times on Raheenmore Bog at different levels of the specific discharge va.
13.3.8 Acrotelm Depth A survey of acrotelm depth and a comparison with results of a similar survey in 1991 was conducted on Raheenmore Bog only. A problem in measuring acrotelm depth is that ‘acrotelm’ is a concept rather than a well-defined layer, such as a soil horizon. Therefore a reproducible ad-hoc definition was adopted, being the depth measured from the surface, over which the degree of humification was 3 or less on Von Post’s scale (Von Post 1922; Van der Schaaf 1999). Acrotelm depth was thus ‘measured’ at all points of the 100 × 100 m grid of Raheenmore Bog (Section 13.3.3), again in a hummock and in a hollow position.
13.4 Results 13.4.1 Acrotelm Transmissivity on Raheenmore Bog Somewhat surprisingly, a Wilcoxon Signed Ranks test (Sachs 1982) on the transmissivity data showed that the differences in transmissivity between hummock and hollow positions were not statistically significant (a ≈ 0.5 for the null hypothesis of no difference). This might support the impression that the upper part of the acrotelm has formed in recent years, during which a differentiation of hydraulic conductivity (and hence transmissivity), resulting from different rates of production, aeration and decay, might have had insufficient time to develop. In Section 13.5 it will be shown that new acrotelm material has indeed formed rather abundantly between 1991 and 2003. Some estimated flow paths are shown in Fig. 13.9, together with surface slopes and flow directions, based on the 100 × 100 m grid. Table 13.2 gives surface slopes for 1991 and 2002, flow path areas and lengths. Because the flow paths of 1991 and 2002 differed insignificantly, only one value is given for flow path area and length. Site L9 was discarded, because it appeared to be positioned in one of the terrestrialised drains. Table 13.3 shows the measured and theoretical values of Ta for 1991/1992 and their ratios. The bottom line contains the geometric means3 of the ratios obtained from both flow path length and upstream area. The usage of the geometric mean may require some explanation. It implies the assumption of a lognormal distribution of the data, which is about as arbitrary as assuming a normal distribution. The geometric mean is defined as the n-th root of the product of n numbers. It can also be found by calculating the arithmetic average of their logarithms and obtaining the inverse logarithm of the result. 3
S. van der Schaaf et al.
100.0
256
N
10 10 2.5 2.0 106.0 105.5 .0
.5 106
10 6
107.0
10 5.5
10 5 1 0 .0 4. 5
L-1
105.0
L11 L8
5 4. 10 0 4. 10 .5 3 1 0 .0 3 10
J6
106.5
P6
Legend
0
500 m
Estimated flowpath boundary Approximate boundary of nature reserve 105.0 Surface level contour (m above sea level) Slope of surface level. The dot gives the site, the line size and direction
Fig. 13.9 Surface slopes and some flow paths of 100 m width at the measuring site for some acrotelm transmissivity sites on Raheenmore Bog, based on the levelling of 1991 Table 13.2 Surface slopes I in m per m, flow path areas Au in ha (based on a width of 100 m at the site) and flow path lengths L in m on Raheenmore Bog in 1991 and 2002 Site I (1991) I (2002) Au L Site I (1991) I (2002) Au L L-1 0.0093 – 2.5 310 L11 0.0046 0.0054 6 1090 L0 0.0081 0.0072 2.3 200 L12 0.0067 0.0067 7 680 L1 0.0058 0.0052 1.3 140 L13 0.0089 0.0089 6 740 L2 0.0018 0.0014 0.7 130 J6 0.0043 0.0045 2.5 540 L3 0.0031 0.0014 0.8 240 K6 0.0039 0.0033 2 540 L4 0.0016 0.0008 1 270 M6 0.0019 0.0016 2 460 L5 0.0012 0.0008 1 380 N6 0.0013 0.001 2 460 L6 0.0027 0.0023 2 460 O6 0.0008 0.0013 1.5 500 L7 0.0043 0.0043 3 540 P6 0.0008 0.0016 2 570 L8 0.0026 0.0031 4.5 660 Q6 0.0025 0.002 3 480 L10 0.0032 0.0031 6 940
The advantage of a lognormal distribution is that negative confidence limits do not exist and that the effect of extremely large values on the result is smaller than if a normal distribution were assumed. The means of the ratios obtained from flow path length and from flow path area differ insignificantly. Their 5% confidence limits, based on a two-tailed t-distribution of the s.e.m.4 for the results based on both flow path area and flow path length are shown in Table 13.3.
4
Standard error of the mean, see, for example Sachs (1982).
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
257
Table 13.3 Ratios of potential and measured acrotelm transmissivity Ta on Raheenmore Bog, based on flow path area and flow path length for 1991/1992. Estimated shape factor f between 1 and 2, depending on flow pattern. Specific discharge va ranged from 1.2 to 2.7 mm day−1
Ta potential Site L-1 L0 L1 L2 L3 L4 L5 L6 L7 L8 L10
n 3 3 3 3 3 4 4 4 4 4 1
Ta measured
flow path area 2.7 18 11 1.1 0.4 16 0.6 3.5 26 3.6 4.8
Ta potential
Ta potential Ta measured
flow path length 1.7 7.9 5.9 1.1 0.6 21 1.1 4 23 26 3.7
Site L11 L12 L13 J6 K6 M6 N6 O6 P6 Q6
n 4 3 2 3 3 3 3 3 3 3
Geometric mean 5% confidence interval of mean
Ta measured
Ta potential Ta measured
flow path area 2.8 2.8 1.6 10 4.4 1.7 5.6 3.1 6.6 1.5
flow path length 2.5 1.4 1.0 11 6.0 1.9 6.3 5.1 9.4 1.2
3.6 2.2–5.9
3.8 2.3–6.4
Table 13.4 Ratios of potential and measured acrotelm transmissivity Ta on Raheenmore Bog, based on flow path area and flow path length for 2002/2003. Estimated shape factor f between 1 and 2, depending on flow pattern. Specific discharges va ranging from 0.27 to 2.3 mm day−1. The number of data n is from two pits per site
Ta potential Site L-1 L0 L1 L2 L3 L4 L5 L6 L7 L8 L10
n 10 10 10 10 10 10 10 10 10 10 10
Ta measured
flow path area 0.9 1.5 0.1 0.5 6.9 17 1.4 0.8 1.1 1.7 2.4
Ta potential Ta measured
flow path length 0.5 0.7 0.05 0.5 10 23 2.6 0.9 1.0 1.2 1.9
Geometric mean 5% confidence interval of mean
Ta potential Site L11 L12 L13 J6 K6 M6 N6 O6 P6 Q6
n 10 10 10 10 10 10 10 10 10 10
Ta measured
Ta potential Ta measured
flow path area 1.0 0.7 2.4 1.1 3.8 3.6 2.8 2.7 0.6 2.8
flow path length 0.9 0.3 1.5 1.2 5.2 4.0 3.2 4.5 0.9 2.2
1.6 1.0–2.5
1.8 0.8–2.7
Table 13.4 shows the same values of Ta for 2002/2003. Again, the means of the ratios obtained from flow path length and from flow path area differ insignificantly. The 5% confidence intervals of the mean, based on a two-tailed t-distribution of the s.e.m., is also shown for both methods. Because the intervals for 1991/1992 and 2002/2003 hardly overlap, the results are reasonably hard evidence of a considerable
258
S. van der Schaaf et al.
improvement of the acrotelm since 1992. A Wilcoxon Signed Ranks test gave a similar result with a level of significance of 0.002 (Ten Heggeler et al. 2005). These results mean a confirmation of the theory developed in Section 13.3.2 and indicate that the mean actual acrotelm transmissivity is approaching its theoretical potential value. The large variability of the ratios of measured and potential Ta may be explained for a large part by differences in acrotelm development at a spatial scale of a few metres. The unequal old bog surface, which probably caused a locally impeded horizontal outflow of water, might also have caused a – probably temporary – ‘overshoot’ in acrotelm development, which may explain some of the ratios <1 shown in Table 13.4.
13.5 Acrotelm Depths on Raheenmore Bog The results of the acrotelm depth surveys of 1991 and 2003 are shown in Fig. 13.10. At most sites the depth has increased, at much less a decrease could be concluded from the map. However, some critical remarks should be made. In 1991, the survey included one check per grid point and it was unclear whether this had been a hummock or a hollow. In 2003, at all sites a hollow and a hummock position were sampled and the average depth of both was taken as a result. This means that the 2003 data are more reliable and generally smoother than those of 1991. The overall picture is one of an increased acrotelm dept. This confirms the differences between Tables 13.3 and 13.4. A clear improvement was found in the area of the blocked drains in the northeast, as expected, but also in more central parts of the bog. Little improvement had occurred in the western part of the bog. This part has short flow paths and hence the acrotelm is rather poorly developed. Most likely, the drains in the east have caused subsidence in their surroundings, causing the apex of the bog to shift westwards. This resulted in increasing flow path lengths in the east, causing favourable conditions for acrotelm recovery there after the drains had been blocked.
13.6 Acrotelm Transmissivity on Clara Bog East As mentioned, acrotelm transmissivity on Clara Bog East was measured at six sites, two on each of the three plots marked A, B and C (Fig. 13.8). Since little or no subsidence had occurred between 1992 and 2003, flow paths had not changed significantly. Figure 13.11 shows the position of the flow paths at the six sites with 100 m width at the site. The presentation of the results in the tables below is similar to the one for Raheenmore Bog in Section 13.4.1. Because of the small number of sites, no map is given. Table 13.5 shows surface slope and flow path data. Although the slopes have slight differences between 1992 and 2002, the flow path data for both years have remained unchanged.
1
?
0 0
1
0
2 1
1 1
1
0
1
0
3 1
1 1
2
0
2
?
1
0
5
0
1 4
500 m
1
0
1
0
1
0
2
0
2
1 1
2
?
2 1
0
1
0
3
0
2 1
3 3
2
1
0
1
5 3
3
0
0
2
0
0
1
0
2 4 2 2 3 4
2
0
1
0
1
0
3
0
1
0
2 1 2
0
3
0
2
0
3
3
0
1 3
0
2 3
2 3 3 3 4
3
0
3 3
3
0
3 2
3
0
0
1
0
? 0
3 2 3 3
3
0
2 3
2
0
5 3
3
0
3 2
2
0
1
0
4
0
2 3
1 3
2 2
1
2 3
1
?
3 3
0
5 3
1 3
4
0
2 3
3
0
5 3
3
0
2 3
3 2
3 1 2 2
2
0
3
0
1 1
3 3
1 2
3 2
1
0
3 1 2
0
3 2
0 0
2 1
1 3
1
0
2 3
1
0
2 3
2
0
2 3
1 2
3
0
2
0
3 3 2
0
1
0
2
0
? 0
? 0
2
0
1
0
1
0
2
0
? 0 1 2 3 4 5
no data 0 cm 0-5 cm 5-10 cm 10-20 cm 20-30 cm >30 cm
Depth class
2
0
1
0
1
0
1
0
Fig. 13.10 Development of acrotelm depth on Raheenmore Bog between 1991 and 2003 with area of blocked drains shown
0 0
1
0
0 0
1 1
0 0
0 0
1
0
0 0
1
0
Site Situation 1991 Situation 2003 Site position
Legend
1
0
N
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands 259
260
S. van der Schaaf et al.
Cla
ra
N
60. 5
Bog pools 59
.5
Ab
59.5
58.5
road
Bog pools
Lough Roe 59.0
58.0
Bog
.0 60
Aa
Bc Bb
drain
s
.0
59
Cc .0
58
Cb
59.0
Legend 58.5 58.0 58.5
0
500 m
Cluster of measuring sites for acrotelm transmissivity Estimated flowpath boundary Approximate boundary of nature reserve Surface level contour (level in m. OD) Gradient of surface level.The dot gives the site, the line size and direction
Fig. 13.11 Flow paths on Clara Bog East with contours and surface slopes indicated
Table 13.5 Surface slopes I in m per m in 1992 and 2002, flow path areas Au in ha (based on a width of 100 m at the site) and flow path lengths L in m on Clara Bog East Site I (1992) I (2002) Au L Site I (1992) I (2002) Au L Aa Ab Bb
0.0018 0.0031 0.0030
0.0027 0.0025 0.0039
3.5 2.5 3.0
350 350 520
Bc Cb Cc
0.0020 0.0055 0.0061
0.0033 0.0076 0.0080
3.5 0.6 1.5
460 100 210
Table 13.6 shows the transmissivities and ratios of Ta for 1992/1993; those for 2002/2003 are shown in Table 13.7. Although the transmissivity values for hollow positions tended to be somewhat higher than those for the hummocks, there was no highly significant difference. This result may be attributed to the recent redevelopment of the acrotelm since only 1996, as a result of which differences in hydraulic conductivity due to more aeration of hummocks compared to hollows, had little time to form. For this reason, all the data per measuring site were averaged to a single value.
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
261
Table 13.6 Ratios of potential and measured acrotelm transmissivity Ta on Clara Bog East, based on flow path area and flow path length for 1992. Estimated shape factor f between 1 and 2, depending on flow pattern. The number of data n is from three pits per site. Specific discharges ranged from 0.5 to 3.6 mm day−1
Site Aa Ab Bb
n 11 14 14
Ta potential
Ta potential
Ta measured
Ta measured
flow path area 5.2 6.5 7.5
flow path length 5.2 6.5 6.8
Ta potential Site Bc Cb Cc
n 16 14 12
Ta measured
flow path area 10 5.1 6.9
Ta potential Ta measured
flow path length 8.6 4.4 4.8
Table 13.7 Ratios of potential and measured acrotelm transmissivity Ta on Clara Bog East, based on flow path area and flow path length for 2002/2003. Estimated shape factor f between 1 and 2, depending on flow pattern. The number of data n is from three pits per site. Specific discharges ranged from 0.2 to 1.0 mm day−1
Ta potential Ta measured
Ta potential
Ta potential
Ta measured
Ta measured
Ta potential Ta measured
Site
n
flow path area
flow path length Site
n
flow path area
flow path length
Aa Ab Bb
23 24 24
0.18 0.13 0.31
0.17 0.13 0.31
20 20 20
0.45 1.5 1.0
0.45 1.6 0.7
Bc Cb Cc
The differences between the ratios of potential and measured Ta in both tables are extremely large compared to those of Raheenmore Bog, shown in Tables 13.3 and 13.4. All, except those of site Cb which is the driest of the six, even show ratios below 1 which is theoretically impossible. This may be attributed to the presence of the old uneven surface, which may to some extent impede flow at the relatively low discharges encountered during the measurement period, which was in February to April 2003, when the amount of rainfall were relatively small. This may create a temporary ‘overshoot’ in local Ta, which will probably reduce at a later stage when the acrotelm becomes more homogeneous. Regardless of the explanation of the ‘overshoot’, a comparison of Tables 13.6 and 13.7 shows an extremely rapid recovery of the acrotelm on a bog surface where it had been largely destroyed as a result of recent drainage (Van der Cruysen et al. 1993; Van der Schaaf 1996)
13.7 Discussion and Conclusions The combination of the applied acrotelm transmissivity measuring technique, combined with the theory on acrotelm transmissivity as a function of specific discharge, flow path length and surface slope as developed in this chapter proved successful in assessing the recovery brought about by the described drain blocking on Clara
262
S. van der Schaaf et al.
Bog East and Raheenmore Bog and the construction of a dam on Raheenmore Bog. This does not necessarily mean that the applied techniques will be successful anywhere. For example, the Irish Midland type of bogs usually does not show the strong differentiation in micro-landscapes as present on bogs in a more continental climate. Hence the flow system in the latter kind of bog is more complex and the simple relationship as derived in this chapter may prove less easy to handle than in the Irish True Midland type. The drain blocking itself resulted in a rather spectacular, quantifiable and statistically highly significant recovery of the acrotelm on both bogs, particularly on Clara Bog East, where the damage was more recent than on Raheenmore Bog. This suggests that acrotelm recovery from recent damage may be faster than from damage that was caused much longer ago. No effort was made to investigate a possible cause, but the phenomenon – if it really exists – might well be related to the presence of more easily degradable young organic matter on a recently damaged bog surface than on one damaged several decades or longer ago. The mechanism may be the production of CO2 from the relatively fresh litter, resulting from the decayed acrotelm that existed before the drain blocking of the 1990s. Some literature (e.g. Smolders et al. 2001) suggests that CO2 production from the substrate stimulates Sphagnum growth and the younger the substrate, the larger is its production of CO2. The results of the transmissivity method were supported by the acrotelm depth survey on Raheenmore Bog, which showed a statistically significant average increase of the acrotelm depth between 1991 and 2003. The applied ad-hoc method to measure acrotelm depth, based on degree of humification, might not be applicable anywhere, but it proved useful during the surveys of 1991 and 2003. Acknowledgement The work described in this chapter was funded by National Parks and Wildlife, Department of the Environment, Heritage and Local Government, Dublin.
References Bloetjes OAJ, Van der Meer JJM (1992) A Preliminary Stratigraphical description of peat development on Clara Bog. Irish–Dutch Peatland Study. Fysisch Geografisch en Bodemkundig Laboratorium, Universiteit van Amsterdam, The Netherlands Boiten W (1985) De Rossum-stuw. Polytechnisch Tijdschrift/Civiele Techniek 40(2):1–8 Connaughton MJ (1967) Global Solar Radiation, Potential Evapotranspiration and Potential Water Deficit in Ireland. Technical Note 31, Meteorological Service, Dublin Hammond RF (1981) The Peatlands of Ireland. To Accompany Peatland Map of Ireland, 1978. Soil Survey Bulletin 35, 2nd edn. An Foras Talúntais, Dublin Hammond RF (1984) The classification of Irish peats as surveyed by the National Soil Survey of Ireland. Proceedings of the 7th International Peat Congress Dublin, June 18–23 1984, vol 1, pp 168–187 Ivanov KE (1965) Fundamentals of the theory of swamp morphology and hydromorphological relationships. Selected Papers. Sov Hydrol 4:224–258 Ivanov KE (1981) Water movement in Mirelands. Academic, London
13 Effects of Drain Blocking on the Acrotelm of Two Raised Bogs in the Irish Midlands
263
Kelly L, Schouten MGC (2002) Vegetation. In: Schouten MGC (ed) Conservation and Restoration of Raised Bogs; Geological, Hydrological and Ecological Studies. Department of the Environment, and Local Government, Staatsbosbeheer, Geological Survey of Ireland, Ireland, The Netherlands, Dublin/Driebergen, pp 110–169 Moore JJ (1964) Die regionale Verteilung der Moore Irlands. Ber. 8. Internat. Kongr. Univ. Mooru. Torfforschung, Bremen, pp 87–89 Rohan PK (1986) The climate of Ireland, 2nd edn. The Stationery Office, Dublin Romanov VV (1968) Hydrophysics of bogs. Israel Program of Scientific Translations, Jerusalem Sachs L (1982) Applied statistics. A handbook of techniques. Springer Verlag, New York/ Heidelberg/Berlin Smolders AJP, Tomassen HBM, Pijnappel HW, Lamers LPM, Roelofs JGM (2001) Substratederived CO2 is important in the development of Sphagnum spp. New Phytol 152:325–332 Succow M, Joosten H (2001) Landschaftsökologische Moorkunde, Zweite Auflage Schweizerbart, Stuttgart Ten Heggeler MMA, Van der Ploeg MMJ, Vuurens SH, Van der Schaaf S (2005) Subsidence of clara bog west and acrotelm development of Raheenmore Bog and Clara Bog East. Wageningen University, Sub-Dept. of Water Resources, Report 121 Van der Cruysen JHG, Grent A, Van Wolfswinkel R (1993) Acrotelm survey of clara bog. Irish– Dutch Peatland Study. Wageningen Agric. Univ., Dept. of Water Resources Van der Schaaf S (1996) Acrotelm conditions in two Irish Midlands raised bogs as affected by surface slope and superficial drainage. In: Lüttig GW (ed) 10th international peat congress. Peatlands use – present, past and future, vol 2. Schweizerbart, Stuttgart, pp 121–127 Van der Schaaf S (1999) Analysis of the hydrology of raised bogs in the Irish Midlands. A case study of Raheenmore Bog and Clara Bog. Diss. Wageningen University Van der Schaaf S (2000) Subsidence along disturbed bog margins and its expansion into bogs. In: Rochefort L, Daigle JY (eds) Sustaining our peatlands. Proceedings of the 11th international peat congress Québec, vol I, pp 262–268, Canada Van der Schaaf S (2002) Bog types, climate and land forms. In: Schouten MGC (ed) Conservation and restoration of raised bogs; geological, hydrological and ecological studies. Dept. of the Environment and Local Government, Staatsbosbeheer, Geological Survey of Ireland, Ireland, The Netherlands, Dublin/Driebergen, pp 11–15 Van der Schaaf S (2004) A single well pumping and recovery test to measure in-situ acrotelm transmissivity in raised bogs. J Hydrol 290:152–160 Van der Schaaf S, Streefkerk JG (2003) Relationships between biotic and abiotic conditions on Clara Bog (Ireland). In: Järvet A, Lode E (eds) Ecohydrological processes in Northern wetlands. Selected papers of International Conference & Educational Workshop Tallinn, Estonia 30 June–4 July 2003, pp 35–40 Van’t Hullenaar JW, Ten Kate JR (1991) Hydrology of Clara and Raheenmore Bogs. Evapotranspiration, storage coefficients, lateral flow in the acrotelm, catchment definition, test of the piezometer method for hydraulic conductivity. Irish–Dutch Peatland Study. Wageningen Agric. Univ., Dept. of Hydrology, Soil Physics and Hydraulics Von Post L (1922) Sveriges Geologiska Undersöknings torvinventering och några av dess hittils vunna resultat. Svenska Mosskulturföreningens Tidskrift 36:1–27 Warren W, Smyth M, Van der Meer JJM, Hammond RF (2002) Geology. In: Schouten MGC (ed) Conservation and restoration of raised bogs; geological, hydrological and ecological studies. Dept. of the Environment and Local Government, Staatsbosbeheer, Geological Survey of Ireland, Ireland, The Netherlands, Dublin/Driebergen, pp 16–31
Chapter 14
Self-Recovery of Cut-over Bogs: Summary from Case Studies Elve Lode, Lars Lundin, and Mati Ilomets
Abstract Investigations in 1996–1998 of one Estonian and three Swedish hand-cut peat sites, abandoned 20–50 years ago, showed five main types of natural or seminatural surface cover in peat pits, with prevailing development of: (1) Sphagnum spp.; (2) Eriophorum vaginatum with Sphagnum spp.; (3) Eriophorum vaginatum together with dwarf-shrubs, Pinus sylvestris and Betula pubescens tree species; (4) ³50% muddy peat surface; and (5) open water and/or ponds. Main reasons for development of different pit environments located in comparably similar climatic and field environment conditions, including their acid and nutrient-poor hydrochemical environment, were local differences in inundation depths and surface water regulation conditions. Those pits dominated by a re-established Sphagnum carpet had a relatively large range of inundation water depths (0.2–0.6 m above soil surface) but comparably small water-level fluctuations (SD = 1.1–2.6 cm). Pits dominated by Eriophorum vaginatum were on average less inundated (up to 30 cm), but with more fluctuation (SD = 3.7–4.8 cm). Study pits characterised by dwarf-shrubs with E. vaginatum had average water levels around 4–20 cm below the peat-soil surface, with comparably-average fluctuations (SD = 2.3–6.9 cm) – quality and quantity of woody growth being related to site wetness. Both highlyfluctuating water levels (SD = 12.1–18.9 cm) in comparably deeply-inundated pits, and the degree of surface soaking in non-inundated pits, were the main reasons for non-establishment of plant cover. Keywords Bog hydrology • Cut-over peatland • Mire ecology • New-grown peat • Wetland restoration
E. Lode (*) and L. Lundin Department of Soil and Environment, Swedish University of Agricultural Sciences, P.O. Box 7001, SE-75007 Uppsala, Sweden e-mail:
[email protected];
[email protected];
[email protected] M. Ilomets Institute of Ecology, Tallinn University, Uus-Sadama 5, 10120 Tallinn, Estonia e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_14, © Springer Science+Business Media B.V. 2010
265
266
E. Lode et al.
14.1 Introduction Historically, Finland, Sweden and Estonia had the largest extension of mires within the Baltic Sea Basin – about 32% of Finland, 23% of Sweden and 22% of Estonia (Joosten and Clarke 2002); currently, about 40% of mires in Finland, 47% in Sweden, and 32% in Estonia, are in an undisturbed state (Vasander et al. 2003). Although most mire disturbances are caused by agriculture and forestry, there is also a large extension of mires influenced by peat extraction – 57,000 ha in Finland, 15,000 ha in Sweden and 18,000 ha in Estonia. Furthermore, abandoned cut-away areas are estimated to about 3,000 ha in Finland, 300 ha in Sweden, and 15,000 ha in Estonia (Ibid). In all these countries, fen and transitional fen mires are the most affected mire types and have been drained for agricultural and forestry use (Paal et al. 1998). As bog landscapes usually consist of mire systems with bogs at the centre, surrounded by fen sites on their margins, a large number of Estonian bogs have also been affected by the drainage of their marginal fen sites. Drainage activity in bog lagg areas in Estonia was comparably extensive in the 1950s and 1960s (Ilomets et al. 1995). The development of peat extraction technology and its methodology have been favoured from the outset by the manner of a peat deposit’s origin, which in bogs usually consists of bog peat on the top, transitional peat in the middle and fen peat at the bottom. Usually, geologists distinguish four peat deposit groups – bog, mixedbog, transitional or mixed, and fen-peat deposits. Extracted peat from these peat deposits were and are crucial for the horticultural and fuel peat market, and is thus reflected by the extension and depth of peat extraction in peat fields. Hereby in Assessing the peat resources, the deposit is considered to be industrial or commercial if the thickness of the deposit is over 0.9 m for fen peat or 1.1 m for mixed or transitional peat or 1.2 m for bog peat and if the area of that kind of deposit is over 10 ha (Ibid).
In the above-named countries, the first hand-cutting and, to some extent, also machine-cutting of peat (Ilomets et al. 1995), is recorded from the middle of the nineteenth century (Vasander et al. 2003). The purpose of the earlier hand-cutting of peat on nutrient-poor bogs was mainly for litter harvesting. Pits or small transects that remained after this ‘wet’ hand-cutting of peat are still recognisable today. The comparably small areas, undrained water conditions and slow rate of peat extraction were all favourable ‘advantages’ for the subsequent spontaneous self-recovery of these areas (Wheeler and Shaw 1995). However, ‘commercially’ organised hand peat-cutting later turned out to be as extensive as the machine block-cutting, leaving comparably large cut-over peat fields with varying sizes of peat pits – and peat beds in between pits. Today’s peat fields, excavated by machinery – where the peat is either milled or cut as blocks or sods – are deeplydrained and large in extension. The unavoidable urgent need to restore such excavated former bog areas, especially in a European context, have quite often resulted in attempts which turned out to be very costly and unfortunately also unsuccessful (Wheeler and Shaw 1995, Lode 2001, Schouten 2002, Blankenburg and Tonnis 2004).
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
267
Apart from the definition of mires as ecologically peat-producing landscapes, the restoration of peat-accumulating conditions in previously-exploited bogs would initially mean the establishment of ‘peat-forming’ plant communities. But, in practice, initiating the redevelopment of fen or bog peat ‘producing’ landscapes to at least an average thickness of 30–40 cm peat (Stanek and Worley 1983) is a multidimensional and long-term process, over a period covering four to six times or more the average human lifespan. Even the re-establishment of ‘peat producing’ vegetation landscapes may take several decades. Within current human lifetimes, therefore, a criterion of successful cut-over peat-field restoration is the establishment of ‘permanent’ key species and communities (Joosten 1992) which have the ability to build up mire-type landscapes of continuously-accumulating peat. Thus the most time- and cost-effective restoration management might be achieved by intertwinning both man-made eco-restoration measures with site-specific almost natural or seminatural self-recovery processes. But, in many cases, it is difficult to precisely determine the complexity of factors that could be supported by man in order to give the successful start and continuity for a bog environment’s rehabilitation. The aim of the investigations of the semi-naturally or almost-naturally recovered peatlands presented in this paper was to distinguish different peatland self-recovering ‘scenarios’ – or possible ways for the establishment of peat-accumulating environments – after the cessation of peat-cutting. The results of the 5-year field investigations were obtained from three Swedish and one Estonian self-recovering mire landscapes where peat-cutting had terminated 20–50 years ago. Therefore, the stage to which these sites have already developed provides restoration development results in the field which lies on a time scale at least double that of comparable results from man-introduced peatland restoration projects carried out during recent decades (see also Lode 2001).
14.2 Study Areas An inventory of a number of exploited peatlands, undertaken in the summer of 1996, showed that only hand-cut peat sites of 20–50 years of age exhibited successful natural or semi-natural re-vegetation. Following these findings, three study sites in central Sweden, namely the Läsarmossen, Björnmossen and Hultamossen bogs, and one in northern Estonia – the Viru bog, were selected for detailed research. All these sites are located between 59°01¢ and 59°30¢ N, West and East of the Baltic Sea (Fig.14.1). At that time, the 50 ha hand-cut peat field of the former Läsarmossen bog (Fig.14.2) provided the widest range of self-recovering surface coverage in former hand-cut pits along with the most comprehensive archive material of geological survey before the start of peat cutting that was available. Therefore, the current selfrecovery study was concentrated on the former Läsarmossen hand-cut pits (59°01¢ N, 14°41¢ E), and integrated into the individual pit surface cover studies being made at the other Swedish and Estonian hand-cut areas (Fig.14.3).
268
E. Lode et al.
Fig. 14.1 Geographical location of the Swedish and Estonian bogs (open circles) with selfrecovering hand-cut sites, adapted to an image of the Baltic Sea drainage basin with wetland land cover (Baltic Environmental Atlas – http://www.grida.no/scripts)
Fig. 14.2 View from the air of the Läsarmossen hand-cut area in central Sweden in 2000 (Photo – AS Hasselfors garden/S.-O. Pettersson)
14.2.1 Management History of Study Sites Archive material of the Swedish Geological Survey (SGU) from 1939 shows that prior to peat extraction, the Läsarmossen bog was covered on average by a 3.7 m deep peat layer, with maximum depths of up to 5 m. The main interest in
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
269
Fig. 14.3 Diagrams of the extent of the Swedish and Estonian natural bogs (dotted lines on diagrams), and study sites in the hand-cut areas
peat harvesting was for ‘horticultural peat’, hence only the depth of the so-called horticulture peat was recorded; the average thickness of Sphagnum-dominated and poorly-humified peat (H = 3–5 in von Post scale) on surveyed area was 2.4 m with a maximum of 3 m in several places. Hand-cutting was the only method used for the excavation of peat blocks during the whole peat-cutting period of 1940–1985. In 1989, the abandoned drainage system was partly blocked and the spontaneous re-colonisation of the site by vegetation started in a considerable variety of ‘pit surface environment’ conditions. The other self-recolonising sites studied in Sweden and Estonia turned out to be similar in the cutting method used. The estimated ‘age’ of peat pit-recovering environments since cutting had ceased were about 25 years old at Hultamossen, and 40–50 years old at the Björnmossen and Viru bogs.
14.2.2 Peat Fields Topography The common topography of the former hand-cut peat fields was a broken one, consisting of hundreds of long rectangular pits lined up in six rows (Fig.14.2). The usual size of extracted-peat pits was about 0.2 ha (100 × 20 m) in Swedish sites and about 0.6 ha (300 × 20 m) in the Estonian Viru bog. Peat-pit rows were separated by 10–20 m wide peat beds and/or by 5–10 m beds between single pits (Figs.14.2–14.4). In spontaneously-inundated peat-pits, water levels were regulated by different through-flow channels and ditches in the peat beds. The peat pit ‘bottom’ surface
270
E. Lode et al. II pit row I pit row
Shallow drainage Deep drainage
P1
3.8
3.9
3.9
3.9
3.8
3.9
3.9
3.9
3.9
3.7
3.9 3.7 3.8 3.8 3.9
4.1
4. 4
4. 5
4.1
3.9
4.0
10-50
3.5
60-75 80-100
4.4
4. 5 4.7 P 10 4.6 P 11 5. 3
4.3
0
3.7 3.9
4.3
4.1
3.9
3.6
4.0
4.0
Muddy surface, 1-30% Open water in the pit, %
3.7 4. 0
3.8
4. 1
Tree species, 10-60%
3.6
4. 0
3.8
4.3
4.3
100 m
3.9
Eriophorum species, 1-10%
P5
3.9
4. 1
P 9 4. 3 4.3
Calluna vulgaris, 10-50% Eriophorum species, 11- 85%
3.8 P 4 3.7
3.6 P6
P 7 4. 0
3.9
Scale
3.9
4.0
4. 0 4. 1
0
3.8
4. 0
3.9
3.7
N
3.7
3.6
3.8
Sphagnum species, 10-99%
3.9
3.8
3.7
3.8
P8
4.0
3.8
3.8
3.7
3.9
4. 0
4.0
3.9
3.8
P 2 4. 0
3.9
3.9
3.7
3.9
4.0
3.9
Plant cover in the pit, %
IV pit row
III pit row
P3
Partly or noninvestigated pits
3.8 3.8 4.2
3.9 4. 3
3.9 3.8
3.9 P12
pH value in surface water Pits with piesometric tubes Discharge gauging station
P 12
3.9
3.4
Fig. 14.4 Distribution of the self-recovering pit surface coverage in Swedish Läsarmossen handcut area surveyed in summer 1997
often had a convex cross-profile, giving differences in pit ‘bottom’ heights from the higher centre to the deeper border of about 0.5 m. Comparison of all study sites showed that the area of cut-over peat fields in the Swedish Björnmossen and Läsarmossen bogs formed about 60% of the natural bog area, and in Hultamossen almost 35%, while the self-recovery area of Viru bog formed only about 15% of the former bog area (Fig.14.3).
14.3 Study Methods 14.3.1 Monitoring of Abandoned Peat Pit Environments Investigation of the abandoned pit environments, regarding topography, peat stratigraphy, plant cover, hydrology and hydrochemistry, started in summer 1996. Study installations were consistently carried out in peat pits, mainly with ‘typical’ types of surface cover, i.e. pits with prevailing: (1) Sphagnum spp. coverage (‘Sph. covered’); (2) Eriophorum vaginatum with Sphagnum spp. coverage (‘Erioph. covered’); (3) Eriophorum vaginatum together with dwarf-shrubs; Pinus
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
271
sylvestris and Betula pubescens tree-species coverage (‘Tree covered’); (4) ³50% mud bottom coverage (‘Mud’); and (5) open water coverage and/or ponds (‘Pond’). During the summers of 1996–1998, about twenty-five study plots were set up for the investigation of self-recovering processes (Fig.14.3). In autumn 1997, a detailed survey was carried out in 145 Läsarmossen peat pits (Fig.14.4) which included detailed plant-cover descriptions complemented by inundated water depths (measured from pit ‘bottom’ border lines) and surface-water pH measurements from the same places.
14.3.2 Geophysiography and Peat Deposit Stratigraphy Coring of residual peat was carried out in pits with groundwater (GW) (piezometer) installations (Fig.14.3). Peat deposits were cored with a Byelorussian (White Russian) peat sampler (50-cm long and 7-cm diameter), and newly-grown biomass sampling carried out using a 50-cm long and 15-cm diameter sharp-edged metallic tube. Composition of the residual plant species was determined in the field together with the degree of peat decomposition on a percentage scale (Lode 2001).
14.3.3 Installation of Hydrological Measurements During the summer periods of 1996–1998, about 25 GW measurement stations were established in peat pits with ‘typical’ self-recovering surfaces. GW stations mainly consisted of three piezometers installed at three levels: (1) in the mineral soil under the peat deposit, to measure the basin water pressure (‘Basin’); (2) and (3) at depths of 50–70 cm (‘Peat’) and 10–30 cm (‘Surface’) from the residual peat surface, to measure GW level fluctuations in the peat (Lode et al. 1998). A total of seven GW stations were established in representative natural areas of Hultamossen and Viru bogs, and 11 stations in a cut-away milled peat field neighbouring the Viru bog self-recovery area (Fig. 14.3). Water level measurements were made manually once or twice a month, except for frozen periods of the year.
14.3.4 Water Sampling for Chemical Analysis Water for chemical analysis was sampled from 22-mm plastic tubes, similar to that used for water level measurements. Total sampling occurred on two to four occasions. Surface water from inundated pit ponds was sampled on up to 13 sampling occasions. The first pH determination of sampled water was made directly in the field. Repeated determination of pH took place in the laboratory as soon as collected water samples arrived from the field, i.e. 2–3 days after sampling. Other parts of
272
E. Lode et al.
water samples were either treated with acid or frozen for conservation. Base cations and metals were analysed by ICP-OES, sulphates by ion-chromatography, DOC on a Shimadzu TOC-5000, nitrogen by flow injection analysis (FIA) and phosphorus by molybdate-blue detection on an auto-analyser after digestion with peroxodisulphate. In the survey of Läsarmossen bog, pH measurements were made in the field with a calibrated Orion pH-meter and a low ionic strength electrode N32 BNC.
14.4 Results and Discussion 14.4.1 Stratigraphy of Abandoned Peat Pits A comparison of the residual peat stratigraphy in the field with historical geological survey data showed that in all study sites only the upper layer of Sphagnum peat had been extracted. The average thickness of remaining peat layers in pits at all Swedish sites was approx. 2 m, with about 1 m of bog Sphagnum and 1 m of Carex fen peat (Lode et al. 1998). At the Viru self-recovering area in Estonia, the thickness of the residual peat in pits was about 5 m, with about 4 m of Sphagnum and 1 m of Carex peat. In a reference part of the natural Swedish Hultamossen site, the total peat thickness at the location of the GW stations was 2.9 m, with 2.5 m of Sphagnum and 0.4 m of fen peat, while in the natural Viru bog it was about 3.4–3.7 m, with about 3 m of Sphagnum peat. The maximum thickness of new-grown, mainly S. cuspidatum and S. magellanicum, biomass in pits was determined in the Viru bog self-recovering site to be up to 36 cm, and in the most favourable pits in Hultamossen, Björnmossen and Läsarmossen sites up to 30 cm (incl. 1–3 cm of new grown peat) (Fig.14.5).
14.4.2 Characteristics of Peat Pits Surface Cover A clustering matrix for the area covered by all Läsarmossen pits consisted of 13 main variables – 11 plant species (selection criterion being the occurrence of species in at least nine pits out of all 145 pits), and two non-plant variables, i.e. percentage area of open water (‘Open water’) and percentage of non-vegetated muddy peat (‘Mud’) within the pits. The tree diagram in Fig.14.6 illustrates the principal relationships between clustered variables. The second or ‘perpendicular’ clustering of the same matrix divided all pits into two main classes: ‘Open water-dominated’ and ‘Vegetation-dominated’ pits (Fig.14.7). In ‘Open water-dominated’ pits, open water area varied mostly between 90% and 100% of the pit surface (Graphs 1, 1a and 1b in Fig.14.7), while plant cover usually did not exceed 30%, except for pits with Bryales species where coverage reached up to approx. 60% (Graph 1a in Fig.14.7). In the ‘Vegetation-dominated’ pits, open water area varied between 10%
273 0.08 0.06 0.04
Thickness of newgrown biomass, cm
0.02 0
0 -10
Bulk density of newgrown biomass, g/cm3
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
-20 -30 -40 Lä P1 Bj I
Bj II
Bj III
Hu I
Hu II Hu III Hu IV
Fig. 14.5 Thickness of the new-grown biomass (cm from the top of the surface/microform) and its dry bulk density (‘bulk density’ in the graph) from the Swedish Läsarmossen (Lä P1), Björnmossen (Bj I, Bj II and Bj III) and Hultamossen (Hu I, Hu II, Hu III and Hu IV) abandoned hand-cut peat pits. Determinations were made in summer periods of 1996–1998 700
Euclidean linkage distance
600 500 400 300 200 100 0
Open water Trees Mud P. sylv. V. ulig. S. cusp. Tot. Sph. Bryales C. vulg. B. pub. L. pal. E. ang. E. vag.
Fig. 14.6 Tree diagram of 11 prevailing variables for 145 Läsarmossen self-recovering peat pit environments, where ‘Tot. Sph.’ = Total Sphagnum, ‘S. cusp.’ = Sphagnum cuspidatum, ‘E. ang.’ = Eriophorum angustifolium, ‘E. vag.’ = Eriophorum vaginatum, ‘C. vulg.’ = Calluna vulgaris, ‘L. pal.’ = Ledum palustre, ‘V. ulig.’ = Vaccinium uliginosum, ‘B. Pub.’ = Betula pubescens, ‘P. sylv.’ = Pinus sylvestris, ‘Open water’ = % of the open water coverage and ’Mud’ = % of the non-vegetated muddy surface coverage (see also Fig. 14.4)
and 40% (Graphs 2, 2a and 2b in Fig.14.7), whereas the main dominant plant species were Sphagnum spp. and Eriophorum vaginatum (Graphs 2a and 2b in Fig.14.7), usually accompanied by Oxycoccus palustris, Rubus chamaemorus, Empetrum
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
110
2a) VegD with Sphagnum and E. vag. n=24
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
110
Fig. 14.7 Box-plots of the classified surface cover variables for the 145 cut-over Läsarmossen peat pits, where n = number of variables (for variable abbreviations, see legend to Fig. 14.6)
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
110
2) Vegeation dominating (VegD) n=51
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
Percentage of coverage Percentage of coverage
110
1a) OpWD with vegetation n=22
Percentage of coverage Percentage of coverage
Percentage of coverage
Percentage of coverage
1) Open water dominating (OpWD) n=46
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
110
2b) VegD with E. vag. n=27
-10 Trees Mud Tot. Sph. E. vag. C. vulg. V. ulig. S. cusp. E. ang. Bryales L. pal. Open water
10
30
50
70
90
110
1b) OpWD with poor vegetation n=24
274 E. Lode et al.
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
275
nigrum, Drosera rotundifolia, Rhodococcum vitis-idaea, Rhynchospora alba, Myrica gale, Utricularia minor, and Andromeda polyfolia, and also by some Hepaticae and Lichenes species. In pits with peaty-muddy bottoms (Graph 2b in Fig.14.7), Eriophorum angustifolium was the dominant plant species besides Sphagnum spp., Calluna vulgaris, Bryales, Pinus sylvestris and Betula pubescens.
14.4.3 Peat Pit Surface Cover and Water Inundation Characteristics Measurements of the open water depth over all Läsarmossen pits showed a clear relation between deep and shallow inundation and the occurrence of typical ‘wetter’ or ‘dryer’ plant species. The range of water-level variations to the developed pit surface coverage, driven by the differences in surface water regulation giving the water conditions in individual pits, was later found to be important. Thus, pits with comparably low drainage ‘activity’ had smaller water level fluctuations with a standard deviation (SD) of 2.2–3.4 cm, compared with pits with a SD of 4.5–18.9 cm, i.e. higher drainage ‘activity’ (Table14.1). At the same time, differences in pit inundation depths were related to the relative heights of remaining pit ‘bottoms’ and also to the depths of connecting channels between pits and the drainage depths of main ditches between pit rows. For this reason pits with a similar inundation Table 14.1 Main statistics of observed surface water levels (WL, in cm above the ground surface) in Läsarmossen pits (Fig. 14.4), classified by the inundation water depths and intensity of water fluctuations for 1998 and 1999 vegetation periods, where ‘GW’ = groundwater, ‘–’ = the GW level below the ground surface, ‘SD’ = standard deviation and ‘Surface cover’ corresponds to the clustering results in Fig. 14.7 Observed WL, cm Confid. interv. of 95% Surface Pits No of cases Mean Min.–Max. Min.-Max Range SD (cm) cover I. Pits with average surface water level >65 cm; less regulated by the drainage P2 28 111 105÷116 110÷113 2.5 3.3 1b P9 25 96 88÷101 95÷98 2.8 3.4 1a P5 25 81 72÷85 78÷82 3.7 4.5 2a P10 25 64 55÷70 62÷65 3.1 3.8 1b II. Pits with average surface water level of 40–65 cm; less regulated by the drainage P7 25 63 58÷70 62÷65 2.3 2.8 1a P4 24 59 51÷65 58÷60 2.6 3.1 2a P8 25 43 37÷47 42÷44 1.8 2.2 2a III. Pits with average surface water level of 50–60 cm; strongly regulated by the drainage P11 25 51 34÷72 46÷56 10.0 12.1 1b P12 25 56 32÷86 49÷63 18.9 1a 14.8 IV. Pits with average water level below pit surface; strongly regulated by the drainage P1 28 0 –7÷6 –1÷1 2.1 2.7 2b P3 24 –6 –15÷1 –7÷-4 3.7 4.5 P6 25 –18 –32÷-10 –20÷-16 4.4 5.3 2b
276
E. Lode et al.
water depth on average had different ranges of water-level fluctuations and a correspondingly different pit surface cover. Hence, in ‘pits with average water level >65 cm’, the relatively low water-level fluctuation (SD = 3.3–3.4 cm) characterised less-regulated water conditions, compared with the SD of 4.5 cm in pit P5 (Table14.1). The same pattern was also observed for those pits with an average water depth of 40–65 cm and SD of 2.8–3.1 cm. Because of the comparably-deep open-ditch drainage, together with the comparably high average inundation depths, the surface water-level fluctuation of SD of 12.1–18.9 cm was the highest for the Läsarmossen area (pits P11, P12 in Fig.14.4). In those cases where the drainage system still functioned – pits with a small inundation depth – the average GW level fluctuated close to the pit peat surface and this was evidently suitable for the growth of Eriophorum vaginatum together with Sphagna and some small Pinus sylvestris individuals. In pit P6, with dominating Eriophorum vaginatum, Calluna vulgaris and relatively high trees, the average GW level was 18 cm below the pit ‘bottom’ surface, the lowest among all pits investigated. Although this P6 pit was well-drained, GW levels fluctuated only in the range of SD = 5.3 cm, compared with SD = 12–19 cm in the well-drained but comparably deeply-inundated (50–60 cm) P11 and P12 pits. Special attention was given to one pit with a very unfavourable plant-growth recovery and which was up to 80% covered by the soft peat mud (Pit 3 in Table14.1 and Fig.14.4). The average GW level in this pit was 6 cm below the peat surface, with corresponding fluctuations of SD = 4.5 cm. It seems that the nature of this surface-close GW level fluctuation was unsuitable for the development of almost any expected plant species because of the freezing and swelling of the upper part of the peat surface in the autumn-spring seasons and the drying out during the summer season. The dominant plant cover developed in pits also seemed greatly determined by the breadth of the plant species’ ecological ranges, giving the availability of each plant’s development in various inundation depths and changing water-level conditions. It was a common feature that pits with well-developed Sphagnum cover had limited drainage conditions and consequently a less-fluctuating surface-water level. However, in pits with good Sphagnum growth ‘supported’ by dominating Eriophorum vaginatum, the fluctuation of water levels was quite high: compare the SD = 4.5 cm in pit P5 with the SD of 3.4 cm in pit P9 that had a well-developed floating Sphagnum carpet (Table 14.1 and Fig. 14.4). Other study pits with similar types of dominant Sphagnum cover showed that Sphagnum may be dominant in inundation conditions from only 2 up to 60 cm water depth, but, however, with comparably low water-level fluctuations of SD = 1.1–2.6 cm (Lode 2001). Pits dominated by Eriophorum vaginatum together with Sphagna had an average water level fluctuation of SD = 3.7–5.0 cm at an inundation depth of 10–80 cm above the pit bottom. In comparably ‘dry’ pits, with dominating E. vaginatum together with some dwarf-shrubs, Pinus sylvestris and Betula pubescens, the GW level was mainly found to be 4–20 cm below the peat surface, with a range of variation SD = 2.3–6.2 cm. Water levels in ‘muddy’ pits were found to be rather close to the peat surface, i.e. from 11 cm below and up to 2 cm above the whole year round (SD = 2.3–4.9 cm).
277
20
Min-Max 25%-75% Median value
10 0 −10 −20
Viru-Ridge
Viru-Ridge
Viru-Hollow
Viru-Hummock
Viru-sr.-Depression
Viru-sr.-Depression
Viru-Hollow
Viru natural bog
Viru selfrecovered bog Viru-sr.-Hummock
−40
Hultamossen natural bog Hulta.-Hummock
−30
Hulta.-Hollow
Water levels related to the mire surface, cm
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
Fig. 14.8 Main statistics of groundwater levels (‘Water level’ in the graph) for the different microforms, monitored during vegetation period of 1998 in natural Swedish Hultamossen bog (in the graph – Hulta.), and self-recovering Estonian Viru bog site (in the graph – Viru-sr.), and natural part of Viru bog (in the graph – Viru)
GW levels in representative natural bog sites and in the Viru bog self-recovering area generally remained below the ground surface, but there were remarkable differences in average GW levels and their fluctuations for different bog microforms. Hence, in parts of the studied natural bogs both in Sweden and Estonia, the GW level in hollows fluctuated on average between 3 cm below and up to 7 cm above the bog surface (SD 2.6–3.0 cm; Fig.14.8). In self-recovering hollow-type depressions the GW level on average was slightly lower. For positive microforms (ridges and hummocks) of the natural study sites, average GW levels were about 9–17 cm below the ground surface, with a higher fluctuation range (SD = 3.2–5.4 cm) compared to depressions.
14.4.4 Peat Pit Surface Cover and Three-Depth Piezometric Levels Analysis of three-depth (‘Basin’, ‘Peat’, ‘Surface’) piezometric water levels in a 2-year time series showed that in shallow, inundated and well-developed Sphagnum-covered pits, those of all water levels fluctuated around the peat surface (Fig.14.9). The three ‘Muddy’ and six ‘Tree covered’ sites showed the largest range of ‘Basin’ water average depths – from 1 to 59 cm below the surface (SD 4.9–14.6 cm), although the
E. Lode et al. 40
Sph. covered
20
−20
Muddy
Tree covered
n=3 sites
n=2 sites
−40 −60
20
Cut-over peat field
Self-restored mire
Viru natural bog
Surf.
n=6 sites Basin
Surf.
Peat
Basin
Surf.
Peat
Basin
Peat
Basin
40
Surf.
n=3 sites
Peat
−80
−100
Hultamossen natural bog
0 −20 n=3 stations
−40 −60 −80
n=5 stations
n=2 stations
n=6 stations
max/min
Surf.
Peat
Basin
Surf.
Peat
Basin
Surf.
Peat
Basin
Surf.
−100 Peat
GW levels related to the surface, cm
Erioph. covered
0
Basin
GW levels related to the surface, cm
278
average
Fig. 14.9 Range of the recorded maximum-minimum (max/min in the graph) and the mean (average in graph) three-depth groundwater levels for: (a) self-recovering peat pits with different dominant characteristics, and (b) different reference sites, where ‘Basin’, ‘Peat’ and ‘Surf’ = the piezometric water levels at the depth of the mineral subsoil, 50–70 cm and 10–30 cm in the peat, respectively
average depths of the perched ‘Surface’ water in both cases were also close to the peat surfaces – from 2 to 10 cm below the surface (Lode 2001). No other studied sites had so large a range of the three-depth GW level ranges as the ‘Tree’ and ‘Mud’ covered sites.
14.4.5 Peat Pit Surface Cover and Hydrochemistry Surface water (‘Surf. w.’ in Table14.2) of all the self-recovering study sites showed similar hydrochemical characteristics – a low nutrient content and low pH around 4 (Fig.14.10). However, there was quite a clear hydrochemical stratification with depth towards a higher pH, i.e. from 3.9–4.0 to 4.5–6.8, and more nutrients in deeper layers (Table14.2).
Table 14.2 Average water chemistry of self-recovering hand-cut peat sites, where ‘Surf.w.’ = surface water sampling results, ‘Peat’ = sampling results at the 20–70 cm level in the peat, and ‘Basin’ = sampling results from the underlying mineral soils, ‘n’ = the number of analysed occasions. Values are averaged from the 1996 summer survey at the Swedish Läsarmossen, Björnmossen and Hultamossen sites, and from the 1999 summer survey at the self-recovering Viru bog site in Estonia Layer pH Ca mg l−1 Mg mg l−1 K mg l−1 NO3-N mg l−1 NH4-N mg l−1 Tot-N mg l−1 Tot-P mg l−1 Läsarmossen bog, n = 2 0.03 1.7 0.09 0.03 0.35 0.43 1.7 4.0 Surf.w. 0.07 2.7 0.05 0.01 0.35 0.39 2.6 3.9 Peat 0.03 4.7 2.08 0.01 0.34 0.23 2.1 5.5 Basin Björnmossen bog, n = 3 0.05 1.1 0.03 0.03 0.08 0.32 1.9 3.9 Surf.w. 0.06 2.1 0.41 0.02 0.40 0.45 3.1 3.9 Peat 0.08 3.6 1.62 0.02 0.62 0.98 7.3 5.9 Basin Hultamossen bog, n = 4 0.03 0.8 0.02 0.02 0.43 0.53 2.0 4.0 Surf.w. 0.06 1.6 0.37 0.01 0.18 3.0 5.9 4.9 Peat 0.05 2.4 1.11 0.00 1.06 22.2 30.4 6.8 Basin Self-recovering Viru bog site, n = 2 0.04 2.7 0.69 0.02 1.01 0.33 1.6 3.9 Surf.w. 0.04 4.1 1.95 0.02 1.03 0.19 1.2 4.0 Peat 0.08 4.8 3.35 0.03 1.01 0.27 1.3 4.5 Basin
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies 279
280
3.0 2.5 2.0 1.5 1.0 0.5 0.0
1a
1b
2a
2b
4.5 4.0 3.5 pH, Ca, K, mg/l
Tot-N, NH4-N, mg/l
3.5
100 90 80 70 60 50 40 30 20 10 0
Tot-N NH4-N DOC
DOC, mg/l
4.5 4.0
E. Lode et al.
Ca K pH
3.0 2.5 2.0 1.5 1.0 0.5 0.0 1a
1b
2a
2
Fig. 14.10 Hydrochemistry at the Läsarmossen pits, where ‘1a’ and ‘1b’ are pits with open water mainly, and ‘2a’ and ‘2b’ are pits with vegetation cover (see also Fig. 14.7)
Ammonia reached high concentrations (1.11–3.35 mg l−1) in deep layers as well as total nitrogen (2.4–4.8 mg l−1), being about three times higher in the mineral soil (‘Basin’) layers compared to surface waters. The self-recovering Viru bog site showed slight hydrochemical stratification with depth only for pH, NH4-N, Tot-N and Tot-P. In piezometric water from the Hultamossen ‘Basin’ layers, the average Ca and Mg concentrations were at high levels (30.4 and 22.2 mg l−1 respectively), an indication of more base-cation-rich mineral soils underlying the peat. Higher concentrations of these elements were also found in the ‘Peat’ water (20–70 cm below the pit ground surface), indicating a probable up-welling influence of the mineral-soil layer into the overlaying peat deposits. However, the content of Ca and Mg in the bog surface water was small. In general, the hydrochemical influence and interactions with vegetation would mainly concern the surface water chemistry. Differences in surface water chemistry, as observed in pits over the different study sites, were significantly smaller by comparison with the downward chemical stratigraphy seen at individual pit sites. At the Läsarmossen site, pits with rather well-established vegetation cover (2a and b pits in Fig. 14.7 and 14.10) showed lower pH, K, Al and SO4 concentrations (especially 2a pits in Fig. 14.7) compared to pits with less well-developed vegetation (1a and b pits in Fig. 14.7), while concentrations of DOC, NH4 and Tot-N were higher (especially in 2b pits in Fig. 14.7). One deviation from this was the Läsarmossen muddy P3 pit, where 0.45 mg l−1 K and 1.54 mg l−1 NH4–N concentrations were relatively high, with 0.11 mg l−1 Al and 0.10 mg l−1 SO4 being on a low level. This pattern was also partly valid for Björnmossen and Hultamossen pits with rather large mud-bottom areas, but their content of K was low (0.17 mg l−1) and SO4 comparably high (0.85 mg l−1). Tot-P content did not reveal any clear pattern, being high (0.075–0.092 mg l−1) at the Läsarmossen (2b pits in Fig. 14.7) and Björnmossen muddy pits and relatively high at Björnmossen (0.083 mg l−1) and Hultamossen (0.054 mg l−1) pits with well-developed vegetation. In plant-covered pits with mud-bottoms at the Läsarmossen site (2b pits in Fig. 14.7), the DOC content showed high values (90 mg l−1), but was comparably low (33–36 mg l−1) in Hultamossen Sphagnum re-covered and mud-bottom pits and in the Läsarmossen tree re-covered pit. C/N ratios were low (about 19–22) in pits with mud-bottoms (2b pits in Fig. 14.7) and higher (about 29–34) in pits with a good development of vegetation
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
281
(2a pits in Fig. 14.7) and in inundated surface water pits with poor vegetation cover (1b pits in Fig. 14.7). In these latter pits, a relatively high DOC content caused the fairly high C/N ratio. Water chemical composition in Läsarmossen pits with significantly fluctuating surface-water levels (Table 14.1) and extremely-poor plant cover showed comparably low pH (4.4), Al content (0.29 mg l−1) and C/N ratio (30), but the lowest NH4-N (0.14 mg l−1) and Tot-P (0.018 mg l−1) content.
14.5 Conclusions Investigations performed in four abandoned hand-cut peat fields showed comparably similar climatic and field environment prerequisites for self-recovering conditions, whereas differences in local site-specific inundation depths and water regulations were the main reasons for the development of different self-recovering pit environments (Fig.14.11). The pits, dominated by a re-established Sphagnum carpet characterised by a relatively large range of average inundation water depths (0.2–0.6 m above the soil surface), had comparably small water level fluctuations (SD = 1.1–2.6 cm). Pits dominated by Eriophorum vaginatum were on average less inundated (by up to 30 cm), but with more fluctuating water levels (SD = 3.7–4.8 cm).
Fig. 14.11 Examples of three self-establishing surface covers in Swedish Läsarmossen peat pits about 20 years after hand-cut peat extraction, where: photo upper left = peat-pond pit with about 90 cm inundation water depth; photo upper right = the self-recolonised tree-covered pit; photo lower centre = the self-recovering Sphagnum carpet pit with shallow 10–50 cm inundation depth below the carpet
282
E. Lode et al.
Even in cases with observed average water depths of around 78 cm (Table 14.1), E. vaginatum had developed at the more shallow inundated and elongated strips along the pits, formed by the soil surface’s convex cross-profile and remained there after hand-cutting. Study pits characterised by dwarf-shrubs, Pinus sylvestris, and Betula pubescens accompanied by E. vaginatum, as a rule, had an average water level around 4–20 cm below the peat soil surface, with comparably medium fluctuations in levels (SD = 2.3–6.9 cm). Thus, the quality and quantity of tree growth were obviously related to the degree of wetness at the site, i.e. the higher the water table, the fewer and lower the trees. Both the strongly fluctuating water levels (SD = 12.1–18.9 cm) in comparably deep inundated pits and the surface soaking conditions in non-inundated pits were the main reasons for non-establishment of plant cover. Due to the comparably small amplitude of surface-water-level fluctuations and comparably thick layer of self-recovering Sphagnum carpet, it could be assumed that these sites had reached the environmental conditions crucial for the start of continuing peat accumulation in the area. In this respect, from all study plots only three showed the same results. It has been stated that, overall, the variation of the chemical composition of water sources helps to control species distribution and vegetation composition in wetlands, and water quality can be of equal ecological importance to water quantity’ (Wheeler 1999). However, our study sites refer to poor, still ‘bog’-like conditions with quite acid and nutrient-poor hydrochemical environments. The hydrochemical conditions between pits showed fairly similar conditions and differences observed would be more likely due to influences in vegetation rather than from chemical conditions that deviated significantly at the start of self-restoration (Lode 2001). Thus, the study results from self-recovering peatlands showed a strong relation between the established surface cover and the peat field environment conditions remaining after hand-cut peat extraction. Favourable ‘advantages’ for the subsequent spontaneous self-recovery of these sites were: (a) area of peat harvested fields smaller then 100 ha; (b) broken topography in pit-covered peat fields, whereas the area of individual pit surfaces did not exceed 0.6 ha; (c) residual peat layers were thick enough (mainly over 2 m), non-compressed by machinery, and mainly in moist to wet conditions; (d) Sphagnum peat was the main residual surface cover both in- and outside the pits, maintaining mainly nutrient-poor and acid environments within pits, even in inundation conditions; (e) comparably low drainage depth into the mineral subsurface. Although these study sites were with successful new-biomass growth, there remains the great risk that some relatively small changes in water regimes, and probably also in water quality, could push the self-recovering processes in another direction rather than that of the progressive re-establishment of a peat-accumulating landscape. Under reasonably suitable hydrological conditions, plots with newlyformed biomass may develop in eco-hydrologically separate peatland patterns, but the extension of these developed plots over the restoring area is still questionable. Although in cut-over peatlands it might always seem that the effects of drainage can be reversed by re-wetting, in reality, entirely new soil profile conditions occur
14 Self-Recovery of Cut-over Bogs: Summary from Case Studies
283
after peat cutting, and these differ significantly from the original ones of the natural mire. Thus all restoration action should take this fully on board – and consider a new environment with a new landscape and site-specific properties. Acknowledgement The studies was made within the framework of several ‘Peatland restoration’ projects at the Department of Forest Soils of Swedish University of Agricultural Sciences (SLU) and were financed by The Royal Swedish Academy of Sciences, The Swedish Institute and The Swedish Peat Research Foundation. The database used for detailed peat deposit and plant descriptions was enlarged during an education collaboration project on ‘Mire Ecology: Restoration of Damaged Peatlands and Terminated Peat Cuttings’, between the Department of Forest Soils of SLU and Institute of Ecology at Tallinn University, Estonia. Landowners Hasselfors AB and later Svenskog AB provided the investigation with suitable field areas. Valuable information, ideas and field observations were provided by Mr. Sten-Ove Pettersson, who actually initiated the rewetting of the Läsarmossen bog. Chemical analyses were mainly carried out at the Department of Forest Soils, SLU by Anne-Marie Karlsson, Kjell Larsson, Ann-Christine Jansson and Anne Wiklander.
References Blankenburg J, Tonnis W (2004) Guidelines for wetland restoration of peat cutting areas. Geological Survey of Lower Saxony, Bremen, Germany Ilomets M, Animägi J, Kallas R (1995) Estonian Peatlands. A brief review of their development, state, conservation, peat resources and management. Ministry of Environment, Republic of Estonia, Tallinn Joosten JHJ (1992) Bog regeneration in the Netherlands: a review. In: Robertson RA (ed) Peatland ecosystems and man: an impact assessment. Department of Biological Sciences, University of Dundee, UK, pp 367–373 Joosten H, Clarke D (2002) Wise use of mires and peatlands – background and principles including a framework for decision-making. Saarijärven Offset Oy, International Mire Conservation Group and International Peat Society, Finland Lode E (2001) Natural mire hydrology in restoration of Peatland functions. Doctoral thesis, Acta Universitatis Agriculturae Sueciae. Silvestria 234. SLU Service/Repro, Uppsala Lode E, Ilomets M, Lundin L (1998) Three rewet peat cuttings. In: Sopo R (ed) The spirit of peatlands, 30 years of the International Peat Society, Proceedings of the International Peat Symposium. International Peat Society, Jyväskyla, Finland, pp 170–173 Paal J, Ilomets M, Fremstad E, Moen A, Børset E, Kuusemets V et al (1998) Estonian wetland inventory 1997. Estonian Ministry of the Environment, Tartu, Eesti Loodusfoto Schouten MGC (ed) (2002) Conservation and restoration of raised bogs. Geological, hydrological and ecological studies. Department of the Environment and Local Government, Staatsbosbeheer Stanek W, Worley IA (1983) A terminology of virgin peat and peatlands. In: Fuchsman CH, Spigarelli SA (eds) Proceedings of international symposium on peat utilization. Bemidji State University, October 10–13, 1983, Bemidji, Minnesota. Vasander H, Tuittila ES, Lode E, Lundin L, Ilomets M, Sallantaus T, Heikkilä R, Pitkänen M-L, Laine J (2003) Status and restoration of peatlands in northern Europe. Wetlands Ecol Manage 11:51–63 Wheeler BD (1999) Water and plants in freshwater wetlands. In: Baird AB, Wilby RL (eds) Ecohydrology, plants and water in terrestrial and aquatic environments. Routledge, London, pp 127–180 Wheeler BD, Shaw SC (1995) Restoration of damaged peatlands. HNSO, London
Chapter 15
Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands Hilde B.M. Tomassen, Alfons J.P. Smolders, Sake van der Schaaf, Leon P.M. Lamers, and Jan G.M. Roelofs
Abstract This chapter discusses and explains various peat bog restoration strategies relating to peat quality, water chemistry and hydrology based on case studies from the Netherlands. Inundation of bog remnants can lead to a rapid redevelopment of (floating) Sphagnum vegetation, usually when poorly humified Sphagnum peat is still present. After inundation, the peat either swells up to the newly created water table or becomes buoyant, in both cases creating a favourable substrate for Sphagnum mosses. Methane production rate and peat chemistry play an important role in the buoyancy of floating peat. The presence of (slightly) calcareous groundwater in the peat base may enhance the development of floating peat by stimulating decomposition processes. Alternatively, the growth of submerged Sphagnum species can also lead to the development of floating rafts. This depends on the availability of light and carbon dioxide in the water layer. Many bog remnants, however, only have strongly humified peat, which does not favour the redevelopment of Sphagnum carpets after deep inundation. On the other hand, most Sphagnum species appear to do very well on surface-soaked strongly humified peat, which is why shallow inundation is to be preferred in such cases. An important prerequisite for the successful restoration of bog remnants is the development of a hydrologically self-regulating acrotelm. Key species involved in this development are Sphagnum magellanicum, Sphagnum papillosum and Sphagnum rubellum. Since these species are usually very slow colonisers, introduction of key species can accelerate the development of a functional acrotelm. H.B.M. Tomassen (*) and A.J.P. Smolders B-WARE Research Centre, Radboud University Nijmegen, P.O. Box 6558, 6503 GB Nijmegen, The Netherlands e-mail:
[email protected] S. van der Schaaf Soil Physics, Ecohydrology and Groundwater Management, Environmental Sciences Group, Wageningen University, P.O. Box 47, 6700 AA Wageningen, The Netherlands L.P.M. Lamers and J.G.M. Roelofs Department of Aquatic Ecology and Environmental Biology, Radboud University Nijmegen, P.O. Box 9010, 6500 GL Nijmegen, The Netherlands M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_15, © Springer Science+Business Media B.V. 2010
285
286
H.B.M. Tomassen et al.
Keywords Carbon dioxide • Methane • Peat buoyancy • Peat quality • Sphagnum
15.1 Introduction Since the discovery of peat as an alternative to wood for fuel, large-scale peat extraction and subsequent transition of excavated peatlands into arable land have destroyed most of the Dutch bogs. At the beginning of the seventeenth century, areas totalling 180,000 ha were still covered by raised bogs (Neijenhuijs 1973). Around 1900, the area had been further reduced to approximately 90,000 ha and nowadays less than 3,600 ha are covered by bogs. These bog relics, however, have mostly lost the specific characteristics of raised bog systems due to peat cutting, drainage and burning. The total area of undisturbed, peat-forming raised bog vegetation in the Netherlands has been estimated at only 5 ha (Barkman 1992). The present-day scarcity of typical bog vegetation has made the restoration of cut-over bogs an important issue (Joosten 1995; Wheeler and Shaw 1995; Gorham and Rochefort 2003). In addition, bogs are important terrestrial carbon sinks or sources and have a potential influence on global carbon cycles (Gorham 1991). Restoration and conservation of the bog relics in the Netherlands started with the acquisition by the National Forest Service (Staatsbosbeheer) of a relatively intact part (Meerstalblok) of the former Bourtanger Moor in the north-eastern part of the country, in 1968 (Schouten et al. 1998). The aim of restoration management is to restore cut-over bogs into regenerating, self-sustaining ecosystems with the appearance and composition of ‘natural’ bogs (Wheeler and Shaw 1995). The restoration measures taken in bog remnants mostly entail rewetting of the desiccated cut-over bog by blocking drainage ditches and constructing bunds to retain precipitation (Wheeler and Shaw 1995; Price et al. 2003). However, the recovery of Sphagnumdominated vegetation has been found to be a complex issue, partly because little is known about the basic processes involved (Joosten 1995; Money 1995; Wheeler and Shaw 1995). A research project funded by the Dutch Ministry of Agriculture, Nature and Food Quality (LNV) was started in 1998, with the aim of obtaining more information on the processes involved in the restoration of peat bog vegetation in the Netherlands. The present chapter presents and discusses the results of fieldwork and laboratory experiments that were carried out to reveal the mechanisms involved in the development of Sphagnum mats in rewetted peat remnants. In four case studies presented, we focus on the different perspectives for peat bog restoration in shallowly inundated versus deeply inundated peat remnants, in relation to peat characteristics (decomposability and swelling ability of peat) and the presence of target species. The following two paragraphs give some basic information about bog restoration: blocking runoff by dams and some biogeochemical principles in bog restoration. Following that we discuss and explain various restoration strategies relating to peat quality, water chemistry and hydrology based on case studies from the Netherlands
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
287
(Bargerveen bog, Haaksbergerveen bog, Mariapeel bog and Fochteloërveen bog). Finally, we discuss the importance of the introduction of Sphagnum species and end this chapter with a concluding paragraph.
15.2 Blocking Runoff from Damaged or Cut-over Bogs by Dams If a bog has been partly cut away, the surface topography is affected. In temperate regions, this may result, in a few rare cases, in a flatter topography, especially if the central part of the bog is affected by drainage and cutting. However, unequal subsidence mostly causes steeper surface slopes in the remaining bog area, allowing water to discharge from the bog remnant at a faster rate than when the bog was in a more pristine state. Subsidence tends to be most severe at the face bank and decreases with distance. Partial cutting of peat, leaving a certain part of the deeper peat in place, may result in a stepwise spatial variation of the surface level in the remaining peatland. In bogs that have been cut over by modern methods such as vacuum harvesting, the surface topography usually remains rather flat. In bog restoration or regeneration, such changed surface slope conditions may pose difficulties in the design of adequate measures. Flow path length, flow pattern and local surface slope determine to a large extent the local ecotope conditions and hence conditions for growth or re-growth of the acrotelm (Van der Schaaf 2002; Van der Schaaf and Streefkerk 2002). If a bog has been partly cut away, not only surface slopes usually increase, but the average flow path length will decrease. Both have an adverse effect on local hydro-physical conditions that should support the development of a peat forming bog vegetation. Blocking drains may be a useful measure if the bog is drained internally (in fact such a measure, if carried out adequately, increases the effective flow path length on the bog surface), but may not be enough if a bog is severely damaged. In such cases, water conservation by building dams may be an effective measure. Such a measure increases the residence time of water in the bog, thus more or less simulating a longer flow path. However, the design of dams has to be made very carefully. Few and high dams may seem attractive, but cause large differences in water depth (Fig. 15.1). Permanent deep inundation (with water depths of several decimetres) is not desired, because in such a situation too little daylight will penetrate to a sufficient depth of the dark-coloured bog water to allow an abundant growth of Sphagnum, usually S. cuspidatum (see Section 15.4.1.4). Hummock forming species may not grow at all under such conditions. Under all conditions, discharge tubes of sufficient size should be installed to prevent water from running over the dam with the risk of collapse of the structure. Large dams look very unnatural affecting the aesthetic appearance of the bog. Hence, also from the viewpoint of landscaping, they should be avoided, if possible. Thus, dam design in Fig. 15.2 is preferred to that of Fig. 15.1.
288
H.B.M. Tomassen et al. Discharge tube
Dam
Cover to prevent dam erosion
Fig. 15.1 Large dams on a sloping peatland surface
Fig. 15.2 Small dams on a sloping peatland surface
In fact, the distance between dams should be determined based on the following considerations: 1. The water level upstream of a dam should not be deeper than a few decimetres at maximum inundation and should fall to the surface level in an average summer. 2. Based on the previous point, the horizontal distance between successive dams should reflect the surface slope: dam spacing can be large on gentle slopes and should be less on steeper slopes. In the nature reserves Bargerveen (Section 15.4.1) and Haaksbergerveen (Section 15.4.2), dam structures have been applied at a large scale. The first ones were installed in the Bargerveen in the 1970s. The material used was a strongly humified peat (so-called black peat or Schwarztorf), the older bog peat from below Weber’s Grenzhorizont (Weber 1900, 1907). This material is mechanically more stable than the overlying younger bog peat and its hydraulic conductivity is lower. There were two reasons to use this material: 1. It was feared that usage of mineral material, such as sand or loam, would introduce unwanted cations in the remaining peat, which might lead to another type of vegetation than bog vegetation. 2. Mineral dams on top of peat layers several meters deep would probably sink into the underlying peat. For small dams, the material proved successful. However, a large dam was built to surround the mostly uncut northern part of the Bargerveen, Meerstalblok. This dam was about 4 m high. It held for more than a decade, but partially collapsed in the autumn of 1998, when more than 100 mm of rain fell within 24 h. A similar experience was obtained in Raheenmore Bog, Ireland. The last dam, built in 1997 after experience with the other two, had led to a more secure building procedure and still stands.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
289
This experience shows that peat may not be a useful building material for large dams, but could be applied for small dam structures. However, there are two additional problems with this type of dam: 1. Because part of the material lies permanently above the water table, it is permanently aerated and thus oxidised. This causes a gradual decay and lowering of the dam. As yet, no exact information is available as to the speed of decay. 2. The peat for such dams has to come from somewhere, usually the nature reserve itself. Removal of strongly humified peat results in open water areas which are not very desirable in bog regeneration. Another point that became clear is that building dams on young peat (white peat or Weißtorf) is less effective than on strongly humified peat, because the younger peat has a considerably higher hydraulic conductivity and may cause leakage underneath a dam. In the Haaksbergerveen, only an insufficient layer of strongly humified peat was available over most of the area as the peat had been cut away over most of its depth and the remaining layer was, locally, only a few decimetres deep. There mostly dams of sand and loam were used. As for the vegetation development there are no signs that the sand, which has a very low content of minerals other than SiO2 and a minor content of Fe-compounds, has influenced the vegetation development, at least not to an extent that bog vegetation could not develop. The dams there are stable, even though in some places a permanent difference in water table of about 0.5 m between adjoining compartments is maintained. However, such dams can only be built where the remaining peat layer is shallow (probably <0.5 m deep), otherwise the peat has to be removed before the dam is built. The grain size composition of the sand has to be fine (loamy sand). Wooden dams have been applied in a few other bog regeneration projects in the Netherlands (including Fochteloërveen: Section 15.4.4). They consist of planks with a rail on one side and a groove in the other, which have been pushed side by side into the remaining peat. It has been effective so far, but the experience with this system is too short to make any assessment as to its durability.
15.3 Some Biogeochemical Principles in Bog Restoration As mentioned in the Introduction, the aim of restoration management is to restore cut-over bogs into regenerating, self-sustaining ecosystems with the appearance and composition of ‘natural’ bogs (Wheeler and Shaw 1995). In restoring the hydrologically self-regulating capacity of degraded bog remnants, the re-development of a functional acrotelm appears to be essential (Joosten 1995; Wheeler and Shaw 1995; Money and Wheeler 1999; Robert et al. 1999). The acrotelm is the uppermost layer of the bog peat, consisting of a fibric layer of organic matter which is produced by the vegetation (mostly Sphagnum mosses) and is progressively broken down. The spongy acrotelm has a high hydraulic conductivity and the ability to retain water in drier periods, thus having a strong self-regulating effect on the depth
290
H.B.M. Tomassen et al.
of the water table (Ingram 1978; Proctor 1995). The acrotelm lies on the permanently anaerobic catotelm, which contains most of the accumulated peat and consists of more or less strongly humified organic matter, with very low hydraulic conductivity. It is generally accepted that an enduring restoration of the peat bog vegetation can only be expected if a new, actively growing acrotelm re-develops (Joosten 1995; Wheeler and Shaw 1995; Money and Wheeler 1999). As Sphagnum mosses can only grow under wet conditions, rewetting of drained peat bogs is normally the first step in peat bog restoration. Blocking drainage ditches and bunding of peat bog remnants have been successfully applied to retain incoming precipitation (Wheeler and Shaw 1995). However, rewetting does not always result in the desired redevelopment of Sphagnum mats. Frequently, vast areas of the bog remnants are flooded to optimise water retention. Major parts of these inundated areas remain devoid of Sphagnum and where Sphagnum does develop, it is usually restricted to mats of Sphagnum cuspidatum in the more shallowly inundated parts (Joosten 1995; Money 1995; Wheeler and Shaw 1995; Smolders et al. 2003). In some cases, floating rafts develop over large parts of the inundated terrain. There is strong evidence that floating rafts present very favourable conditions for peat mosses and for peat accumulation, as they provide permanently wet conditions but are never flooded (Money 1995; Joosten 1995; Lamers et al. 1999; Smolders et al. 2002). The quality of the residual peat seems to play an important role in the perspectives for bog restoration. The degree of decomposition, for instance, differs greatly. Although the poorly humified material comprising the acrotelm can be distinguished from catotelm peat, the degree of decomposition of the catotelm peat can also differ considerably between different locations and depths. The degree of decomposition may determine the residual decomposition rate of the peat. Anaerobic decomposition processes produce carbon dioxide (CO2) and methane (CH4). Substrate-derived CO2 is an important carbon source for submerged growing Sphagnum cuspidatum (Paffen and Roelofs 1991; Riis and Sand-Jensen 1997; Smolders et al. 2003) and very probably also for emergent Sphagnum species (Smolders et al. 2001). CH4 seems to play an important role in the buoyancy of floating peat (Scott et al. 1999; Lamers et al. 1999; Smolders et al. 2002; Tomassen et al. 2004a). If floating rafts develop, they frequently remain dominated by Sphagnum cuspidatum or Sphagnum fallax. Although the development of a vegetation dominated by these species should be regarded as positive, if only because they may provide a good substrate for the target species Sphagnum magellanicum, S. papillosum and S. rubellum (Wheeler and Shaw 1995), it is essential that vegetation types dominated by these target species develop. These species are more productive and resistant to decay than the hollow species S. cuspidatum and S. fallax, and are therefore able to produce the right acrotelm characteristics required for the self-regulating mechanisms of a stable bog environment (Joosten 1995; Robert et al. 1999). The lack of colonisation of rewetted peat remnants by the typical hummock and lawn species has been ascribed to the high levels of atmospheric nitrogen deposition (Lütke Twenhöven 1992; Li and Vitt 1994). However, Sphagnum magellanicum and Sphagnum papillosum growing adjacent to Sphagnum cuspidatum mats have been observed to invade these mats, even in countries where atmospheric deposition levels are high (personal observations).
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
291
Fig. 15.3 Map of the Netherlands giving the location of the cut-bogs described as case studies in this chapter (Bargerveen, Haaksbergerveen, Fochteloërveen and Mariapeel)
In the next paragraph, the effects of rewetting measures taken in four cut-over bogs in the Netherlands (Fig. 15.3: Bargerveen bog, Haaksbergerveen bog, Mariapeel bog and Fochteloërveen bog) are described and explained in more detail. Rewetting measures have been taken by the management organisations (National Forest Service and Natuurmonumenten) and we were able to follow and interpret the effects based on existing knowledge, and additional measurements and experiments.
15.4 Case Studies 15.4.1 Bargerveen Bog 15.4.1.1 Introduction The Bargerveen bog (52°41¢N; 7°01¢E) is a remnant (2,100 ha) of the formerly extensiveBourtanger moor (300,000 ha) on the border between the Netherlands and
292
H.B.M. Tomassen et al.
Germany (Fig. 15.3). Once, the Bourtanger moor was the largest bog in Western Europe and covered a large area of the northern part of the Netherlands (160,000 ha). Due to large-scale peat extraction, the majority of the Bourtanger moor has been destroyed. Nowadays, the Dutch part of the bog is comprised of three areas: Meerstalblok (500 ha), Amsterdamse Veld (600 ha) and Schoonebeekerveld (900 ha). The Bargerveen bog is owned and managed by the National Forest Service. A considerable part of the Meerstalblok remained uncut and typical bog vegetation survived (70 ha). The characteristic vegetation includes 14 Sphagnum species (including Sphagnum pulchrum), Narthecium ossifragum, Andromeda polifolia, Vaccinium oxycoccos, Scirpus cespitosus, Eriophorum angustifolium, E. vaginatum, Drosera longifolia, D. rotundifolia and D. intermedia. The other part of the Meerstalblok has more or less been damaged by peat cutting. In the past, the Meerstalblok has been used for buckwheat fire culture. Large scale peat extraction has taken place in the Amsterdamse Veld and part of the Schoonebeekerveld, leaving a layer of 0.5–1 m strongly humified peat. Peat extraction has taken place until 1995. The other part of the Schoonebeekerveld area has been directly transformed into arable land. 15.4.1.2 Hydrological Measures In 1968, the first section (66 ha) of the Bargerveen bog has been acquired by the National Forest Service. Between 1970 and 1998, 40 km of peat dams were constructed to retain rain water and kilometres of drainage ditches and canals were blocked. In 1998, during a period with high rain fall, the peat dam collapsed with drastic effects for the bog and its surroundings. At the moment, a new rewetting project (EU LIFEproject) has started to further improve the hydrology of the bog (mainly Meerstalblok). The large peat dams have been replaced by high ‘sand’ dams (a mixture of sand, peat, loam and vegetable mould) as to further raise the water level. The retention of rain water has been moved from inside the bog to large water basins outside the bog. A high water basin permanently contains water and gives counter-pressure to the high water levels inside the area. The low water basins function as water storage. The main goal for the hydrological management is to limit water level fluctuations to 25 cm. 15.4.1.3 Effects of Hydrological Measures The construction of dams had positive effects on the hydrology of the Meerstalblok area. The increased water level stimulated the development of the Sphagnum-dominated vegetation that was still present. In the peat pits the remaining peat layer (bunkerde) became buoyant and floating rafts developed (see Section 15.4.2.4). These floating rafts were rapidly colonised by Sphagnum cuspidatum, followed by other ombrotrophic species. In some of the peat pits no floating rafts developed and the succession started with the development of submerged S. cuspidatum. Due to the high nitrogen deposition levels in the Netherlands (up to 4 g N m−2 year−1) and P availability, Molinia caerulea and Betula trees are able to dominate the vegetation at most sites (Fig. 15.4; Tomassen et al. 2003a, b). In the Section 15.4.1.5, the effects of high nitrogen deposition rates on bog vegetation are explained in more detail.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
293
Rewetting measures taken in the 1990s resulted in a large-scale inundation of strongly humified peat in the Amsterdamse Veld and Schoonebeekerveld area (Fig. 15.5). Development of submerged Sphagnum cuspidatum was hardly observed, even after 10 years. It was obvious that deep inundation of strongly humified peat was not the best way to restore the bog. The lack of submerged Sphagnum growth was the result of wave action, light limitation and carbon limitation. Growing conditions for submerged Sphagnum cuspidatum were identified in more detail and are described in the next paragraph. The large areas of open water attract numerous species of waterfowl.
Fig. 15.4 Dominance of Molinia caerulea and Betula trees in the Meerstalblok area, Bargerveen
Fig. 15.5 Overview of the Amsterdamse Veld area, part of the Bargerveen bog. Rewetting measures resulted in deeply inundated peat (over 0.5 m) with a lack of Sphagnum growth
294
H.B.M. Tomassen et al.
15.4.1.4 Growth of Submerged Sphagnum Species Under optimal growing conditions, submerged Sphagnum cuspidatum is able to fill up the entire water layer, thus forming a loose raft on which other species may ultimately establish (Wheeler and Shaw 1995; Money and Wheeler 1999). In large water bodies, agitation due to wave action hampers the development of floating Sphagnum cuspidatum rafts (Money 1995; Wheeler and Shaw 1995; Money and Wheeler 1999). This means that the creation of sheltered water bodies is to be preferred. In shallow inundated systems this can be provided by growth of emergent vascular plants such as Eriophorum vaginatum (Sliva 1998; Money and Wheeler 1999). The growth of Sphagnum species can, however, also be greatly influenced by water quality. In very clear lakes, for instance, bryophytes such as submerged Sphagnum can grow at depths of 9 m and more (Middleboe and Markager 1997; Riis and Sand-Jensen 1997). In most inundated peatlands, however, submerged Sphagnum growth is limited to waters whose depth does not exceed 0.5 m (Joosten 1995; Money 1995; Wheeler and Shaw 1995; Money and Wheeler 1999). Humic acids from the remaining peat substrate colour the water layer (dystrophic water) resulting in a limited penetration of light. Sphagnum cuspidatum plants start to float only when sufficient photosynthesis takes place and oxygen bubbles accumulate in the mosses (Paffen and Roelofs 1991). This means that at the onset of the growing season, Sphagnum cuspidatum has to be able to attain a sufficiently high photosynthetic rate in order to become buoyant. The light compensation point for submerged Sphagnum species may vary widely. A study of Sphagnum subsecundum growing at a depth of 9 m in lake Grane Langso (Denmark) and thus acclimatised to the low temperatures (8°C) prevailing at this depth, found light compensation points as low as 1.3% of mean surface irradiance (Riis and Sand-Jensen 1997). However, S. subsecundum which grew in the same lake at a depth of 0.7 m, and was thus acclimatised to higher temperatures (20°C), had a light compensation point of 5.5% of mean surface irradiance (Riis and Sand-Jensen 1997). Apparently, low temperatures in the hypolimnion of deep lakes may offer a metabolic advantage by reducing the rate of maintenance respiration (Riis and Sand-Jensen 1997). In bog pools or inundated bog remnants, however, temperatures will become high in summer and light compensation points will probably be close to 5% of surface irradiance. A good measure of the dissolved colouration, for instance due to humic acids, is presented by the extinction at 450 nm (E450) (Kirk 1994). Surface water E450 values between 0.100 and 0.300 are quite normal in inundated peatlands. At an E450 of 0.100, the depth to which 5% of the light penetrates will lie between 50 and 60 cm (Fig. 15.6). This means that even if the water is only slightly coloured with humic acids (E450 = 0.100) Sphagnum will not be able to grow at depths much exceeding 60 cm. At an E450 of 0.250, this depth is only 20 cm. Light limitation can thus very easily hamper the growth of submerged Sphagnum in inundated bog remnants, which might explain why in peatlands submerged Sphagnum is normally absent from waters whose depth exceeds 0.5 m (Money 1995; Wheeler and Shaw 1995; Lamers et al. 1999), in accordance with our observations in the Amsterdamse Veld area.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
295
Fig. 15.6 Relationships between the depths to which different percentages of light penetrate into the water layer and extinction at 450 nm (E450). Relations drawn in the Fig. are as follows: 2%: depth = 6.1525*(E450) − 1.0499 (R2 = 0.9988); 5%: depth = 4.1426*(E450) − 1.1080 (R2 = 0.9997); 10%: depth = 3.0232*(E450) − 1.1225 (R2 = 0.9989); 20%: depth = 2.0854*(E450) − 1.0715 (R2 = 0.9943); 30%: depth = 1.297*(E450) − 1.009 (R2 = 0.9778) (After Smolders et al. 2003)
Apart from light, CO2 availability is crucial to the growth of submerged Sphagnum species (Baker and Boatman 1990; Paffen and Roelofs 1991; Riis and Sand-Jensen 1997). CO2 in surface waters is derived from the sediment by seepage and, in organic substrates such as peaty sediments, from decomposition processes (Wetzel 2001). As a result, CO2 concentrations are highest just above the sediment layer (Casper et al. 2000). Therefore, shallower waters will not only present more favourable light conditions but also higher CO2 concentrations for submerged Sphagnum. In general water plants which depend on the water layer for their CO2 uptake can only develop if carbon dioxide fluxes from the sediment are sufficient. Paffen and Roelofs (1991) demonstrated that Sphagnum cuspidatum was only able to reach buoyancy when CO2 levels in the water were higher than 500 µmol l−1. In one of our experiments we grew Sphagnum cuspidatum at two different water depths (10 and 40 cm) in bog water from the Mariapeel bog (Section 15.4.3) with an E450 of 0.380. In this reserve, Sphagnum cuspidatum growth was only observed at locations where water depth did not exceed approximately 10 cm. We applied two different CO2 concentrations, 100 and 2,000 µmol l−1. The results clearly show
296
H.B.M. Tomassen et al.
that Sphagnum cuspidatum only achieved high net CO2 fixation and growth rates at a combination of high CO2 concentrations (2,000 µmol l−1) and high light levels (inundation depth of 10 cm) (Fig. 15.7). We conclude that the availability of both light (clear water and/or very shallow water) and CO2 have to be sufficient to enable submerged Sphagnum to reach high photosynthetic and growth rates. These conditions are frequently not met in inundated peatlands. Firstly, light attenuation is considerable in the dystrophic water. Secondly, if the substrate consists of acidic and or highly decomposed peat, CO2 fluxes from the peat to the water may be very low due to the low decomposition rates in such substrates (Lamers et al. 1999; Smolders et al. 2002; Tomassen et al. 2004a). Not only submerged growing Sphagnum mosses depend on substrate-derived CO2 for their growth. Emerged Sphagnum species also proved to use substrate-derived CO2 as an important carbon source. The effects of different CO2 concentrations in the acrotelm pore water on the development of Sphagnum magellanicum monoliths grown in wet conditions was determined in a laboratory experiment. Decomposition processes
Fig. 15.7 Mean net CO2 fixation rate (a) and biomass increase (b) of Sphagnum cuspidatum incubated at 100 and 2,000 µmol l−1 CO2 in combination with shallow (10 cm) and deep (40 cm) inundation in water with a coloration (E450) of 0.380. Values are means of four replicates. Vertical bars indicate standard deviation (After Smolders et al. 2003)
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
297
result in much higher CO2 concentrations in bog acrotelm than atmospheric concentrations. S. magellanicum hummocks developed only when CO2 concentrations in the acrotelm water were high (Fig. 15.8). Plants in the low CO2 treatment appeared to be carbon limited; they had increased nutrient concentrations and decreased ratios of carbon to other nutrients (Smolders et al. 2001). Chlorophyll concentrations were much higher in these plants, probably allowing them to harvest more CO2, but the allocation of carbon to structural tissues was lower, resulting in weak stems that were unable to remain upright. As a result of carbon-limitation, net biomass and height increases of the monoliths were much lower in the low CO2 treatment (Fig. 15.8). Results indicate that, at least in wet conditions, atmospheric CO2 alone is not sufficient to enable S. magellanicum to develop its normal vertical growth pattern. Therefore high peat water carbon dioxide concentrations are important for the development of Sphagnum-dominated vegetation. On locations with only strongly humified peat, the limited release of carbon dioxide from this peat hampers the rapid development of Sphagnum-dominated vegetation under wet conditions. 15.4.1.5 Effects of High Nitrogen Deposition Rates on Bog Vegetation In addition to the detrimental effects of drainage, Dutch bog relics have been exposed to high levels of atmospheric nitrogen (N) input since the middle of the twentieth century. Increased N inputs have greatly altered the vegetation composition
Fig. 15.8 Effect of peat water carbon dioxide concentrations (20, 750, 2,000 and 5,000 µmol l−1 CO2) on the growth of Sphagnum magellanicum monoliths. Net weight increase, height increase and increase of monolith surface area at t = 18 weeks, expressed as a percentage of the value at t = 0 weeks. Bars with the same letter do not differ significantly according to one-way ANOVA (After Smolders et al. 2001)
298
H.B.M. Tomassen et al.
in various ecosystems (Bobbink et al. 1998, 2003). The rapid invasion of Dutch bogs by Birch trees (Betula sp.) and Purple moor grass (Molinia caerulea) is likely to be the result of high N availability (Fig. 15.9). Molinia and Betula can only dominate bogs when nutrient availability in the rhizosphere is increased. The presence of a vital Sphagnum layer can immobilise large amounts of N (Aerts et al. 1992; Limpens et al. 2003a; Tomassen et al. 2003a) and thus act as an N buffer or filter (Lamers et al. 2000; Berendse et al. 2001). N uptake by Sphagnum mosses is influenced by environmental factors including moisture availability, shading by vascular plants and the availability of other nutrients (especially P) (Hayward and Clymo 1983; Aldous 2002; Limpens et al. 2004) and C (Smolders et al. 2001). As long as growing conditions are optimal for Sphagnum and N deposition rates are not too high, N concentrations in the rhizosphere will remain low or very low (Lamers et al. 2000; Limpens et al. 2003a; Tomassen et al. 2003a) and nitrophilous vascular species will be unable to become dominant. In Dutch bogs, however, nitrogen deposition loads are too high and growing conditions are frequently suboptimal in terms of water supply and vascular plant density (Limpens et al. 2004; Tomassen et al. 2004b). When such conditions are limiting the growth of Sphagnum mosses and/or N deposition rates are high, a surplus of N becomes available. At first, Sphagnum can immobilise the N overload by producing N-rich free amino acids (Nordin and Gunnarsson 2000; Tomassen et al. 2003a; Limpens and Berendse 2003a). Accumulation of free amino acids, including arginine, asparagine and glutamine, by Sphagnum is the first sign of N saturation, although N concentrations in the rhizosphere will still be relatively low. If the assimilation of N into free amino acids
Fig. 15.9 Above-ground biomass of Molinia caerulea after 3 years of exposure to various rates of experimental N addition (mean + 1 SE; n = 4). Different letters indicate significant differences (P £ 0.05) between N treatments (univariate GLM) (After Tomassen et al. 2004b)
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
299
cannot keep up with the N uptake, toxic NH4+ accumulates in the cells and leads to reduced Sphagnum growth (Limpens and Berendse 2003a). Amino acid accumulation results in higher capitulum N concentrations, increasing the sensitivity of Sphagnum to infections with its fungal parasite Lyophyllum palustre (Limpens et al. 2003b). Finally, the imbalance between N availability and N uptake by Sphagnum mosses results in increased availability of N in the rhizosphere. Vascular plants such as Betula and Molinia are able to profit from the extra N and may invade ombrotrophic bogs. Whether invasion by Molinia and Betula actually takes place depends on the availability of not only N but also other nutrients, especially P. N addition experiments revealed that Betula was not able to profit from the surplus of N, probably due to P limitation (Tomassen et al. 2003a; 2004b). Peat water oPO43− concentrations in Dutch bogs proved to be much higher than in Irish bogs (Fig. 15.10). The combined high availability of N and P, together with the often sub-optimal growing conditions for Sphagnum mosses, results in the invasion of Dutch bogs by Molinia and Betula. In Ireland we found that increased input of N and P by bird dropping led to local changes in the vegetation composition similar to those observed in the Netherlands (Tomassen et al. 2005). Invasion of ombrotrophic bogs by vascular plants like Molinia and Betula at high N deposition rates makes the vegetation structure more complex and increases the density, and high densities of vascular plants can hamper Sphagnum growth by shading (Hayward and Clymo 1983; Heijmans et al. 2001; Malmer et al. 2003). Limpens et al. (2004) showed that under favourable growing conditions, interception of more than approximately 50% of the light depresses Sphagnum growth. In addition, high densities of vascular plants such as Betula and Molinia also affect the interception of rainwater and the input from dry deposition. Increased total cover by Betula and
Fig. 15.10 Concentrations (mean ± 1 SE) of ammonium and phosphate at a depth of 0–10 cm in bogs in the Netherlands (NL; n = 12) and Ireland (IR; n = 7). * P £ 0.05; ** P £ 0.001 t-test (After Tomassen et al. 2004b)
300
H.B.M. Tomassen et al.
Fig. 15.11 Relationship between above-ground biomass of Betula pubescens plus Molinia caerulea (g m−2) and evapotranspiration (mm day−1) during the final growing season of the experiment. Different N treatments are indicated by different symbols (linear regression: R2 = 0.485; P < 0.001) (After Tomassen et al. 2003a)
Molinia has a stimulating effect on the evapotranspiration rate (Fig. 15.11). Bogs invaded by Molinia and Betula have an increased canopy density and the interception of precipitation results in a reduced input of precipitation onto the bog. In Dutch bogs already suffering from the effects of water table draw-down, increased interception stimulates further desiccation. It can be concluded that in areas with high N deposition rates, the changes in canopy density and structure of the bog vegetation caused by high densities of Betula and Molinia increase the interception of water and N by the vegetation. Thus, increased canopy density and the concomitant desiccation and N eutrophication can be expected to stimulate the growth of Betula and Molinia even more, and hamper the growth of Sphagnum species. This is a self-enhancing process, which might be interrupted by measures like Birch tree removal or mowing the vegetation.
15.4.2 Haaksbergerveen 15.4.2.1 Introduction The Haaksbergerveen bog (52°8¢N, 6°46¢E; 500 ha) is a cut-over bog situated in the eastern part of the Netherlands, close to the German border (Fig. 15.3). Since the beginning of the nineteenth century peat was extracted in the bog. Peat extraction was finished in the 1950s. Drainage ditches led to severe desiccation of the peat and invasion of Betula trees. In 1959, a peat fire resulted in the establishment of Betula trees on the burned peat (Ganzevles 1991). Before regeneration the heaths were characterised by an Ericion tetralicis community often dominated by Molinia caerulea, Betula
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
301
pubescens and B. pendula. The Haaksbergerveen bog houses a large population of the adder (Vipera berus), the only venomous snake in the Netherlands. 15.4.2.2 Hydrological Measures Since 1969, hydrological measures have been taken to improve the local hydrology of the bog. Restoration of the Haaksbergerveen bog involves flooding as a result of ditchblocking and the construction of sand bunds. As a result of subsidence the surface slope is relatively high (4 m). In 1974, a system of dams (‘sawah’-system) has been constructed as shown in Fig. 15.2. The success of these measures made it necessary to raise the dams in 1992. Since 2001, two straight roads crossing the bog have been closed and new, more ‘natural’ roads were constructed based on the relief present in the bog. 15.4.2.3 Effects of Hydrological Measures Flooding caused the death of Molinia and Betula, and within 10–15 years characteristic hummock-hollow vegetation dominated by several Sphagnum species (including S. magellanicum and S. papillosum) and characteristic ericaceous species (Andromeda polifolia, Vaccinium oxycoccos, Erica tetralix) had established successfully on peat remnants that had begun to float (Fig. 15.12). The influence of buffered groundwater probably stimulated floating raft development (Lamers et al. 1999). In the following paragraph the processes determining floating raft development are described in detail. However, Betula spp. and M. caerulea also rapidly colonised this floating bog, as observed in many other bog areas in the Netherlands
Fig. 15.12 After rewetting measures have been taken, floating rafts developed in the Haaksbergerveen bog
302
H.B.M. Tomassen et al.
(Schouwenaars et al. 2002) and in other Western European countries (Aaby 1994). This is a source of great concern for bog management (see Section 15.4.1.5). 15.4.2.4 Buoyancy of Residual Peat The buoyancy of floating peat mats is caused by CH4 bubbles accumulating in the peat (Scott et al. 1999; Lamers et al. 1999; Smolders et al. 2002), so methane production rates play an important role in the buoyancy of residual peat after inundation. Most of the floating rafts which developed in the Haaksbergerveen bog are permanently buoyant, but some only float in summer. We measured methane concentrations in one of these seasonally floating rafts and found that the floating raft was buoyant when CH4 concentrations exceeded 350–400 µmol l−1 (Fig. 15.13). During winter the CH4 concentrations were significantly lower and the raft was inundated. The optimum temperature for CH4 production in peat soils is between 25°C and 30°C (Williams and Crawford 1984; Dunfield et al. 1993; Bergman et al. 1998, 2000). However, temperature is not the only factor that accounts for summer buoyancy. The availability of easily degradable compounds (e.g. root exudates), which are important substrates for the methanogenic bacteria also peaks in summer. Together these two factors explain most of the seasonal variation in CH4 production rates (Bergman et al. 2000).
Fig. 15.13 Peat water methane concentrations (means ± 1 SE; n = 2) in a floating raft showing a seasonal buoyancy pattern in the Haaksbergerveen nature reserve, the Netherlands, between November 1998 and June 2000 at 0.2–0.5 m depth. Dark bars indicate buoyancy of the raft. t-test: P = 0.005 (After Tomassen et al. 2004a)
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
303
Most floating rafts are permanently buoyant, and their CH4 concentrations remain high throughout the year (Smolders et al. 2002). High CH4 concentrations depend not only on high production rates, but also on the peat’s capacity to retain the CH4 bubbles produced. Presumably the seasonal floating raft in the Haaksbergerveen did not have the appropriate structure to retain sufficient CH4 bubbles for buoyancy during periods with low CH4 production rates. Though the CH4 concentrations in the peat profiles increased with depth (Fig. 15.14), potential CH4 production rates decreased with depth, indicating that superficial, relatively young, poorly humified peat supports the greatest CH4 production rates. Other studies have also reported this decline in methanogenesis with depth (Williams and Crawford 1984; Yavitt et al. 1987, 2000). The surface layers have a high gas conductivity so that the CH4 produced is readily vented to the atmosphere or oxidised by methanotrophic bacteria, and relatively little is retained within the peat (Segers 1998). In the deeper layers CH4 is retained due to the low gas conductivity, and a substantial amount of CH4 is stored (Brown et al. 1989). The poor substrate quality of highly decomposed peat limits both CO2 and CH4 production rates, even though 95% of the peat consists of organic matter (Bridgham and Richardson 1992). Incubations with highly decomposed peat from the Amsterdamse Veld area showed that the low availability of appropriate organic substrate for methanogens caused the lack of methane production (Fig. 15.15). Addition of acetate as a substrate for methanogenic bacteria had a strong stimulating effect on the methane production rate. Carbon mineralisation rates are usually highest in recently formed relatively coarse, light organic fractions (Hassink 1995; Van den Pol-Van Dasselaar and Oenema 1999; Bozkurt et al. 2001). Since decomposition processes break down larger organic particles into smaller ones, an increase in the decomposition extent will result in the peat having a higher bulk density (Damman 1988; Wheeler and Shaw 1995; Van den Pol-Van Dasselaar and Oenema 1999; Bozkurt et al. 2001). Therefore, peat with a lower bulk density is
Fig. 15.14 Methane production rates and methane concentrations measured in the peat from nine bogs in the Netherlands and Ireland at various depths (0.1, 0.5 and 1–3 m). The relation between methane production rates and methane concentrations has been described by an exponential regression analysis (P = 0.000) (After Tomassen et al. 2004a)
304
H.B.M. Tomassen et al.
Fig. 15.15 Methane production in incubations of strongly humified peat from the Amsterdamse Veld (Bargerveen). After 30 days 500 µmol sodium acetate was added to the incubation. Means (n = 4) are given; bars indicate SD (After Smolders et al. 2002)
Fig. 15.16 Methane production (µmol l−1) plotted against bulk density (g DW l−1) for residual peat which has become buoyant (●, n = 13) and residual peat which has not become buoyant (□, n = 16) after inundation of the peat surface. All peat samples were taken from Dutch bog remnants which have been re-wetted during the last 5 years (After Smolders et al. 2003)
usually less decomposed and tends to have a higher potential decomposition rate than heavier peat (Fig. 15.16). Potential C production rates are positively related with concentrations of P, N and K in the peat (Fig. 15.17; Tomassen et al. 2004a). Coulson and Butterfield (1978) reported a strong correlation between the microbial decomposition of plant substrates and the N and P concentrations of the substrates. Given that nutrient availability can limit the
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
305
Fig. 15.17 C production rates (CH4 + CO2), measured by anaerobic incubations, and (a) P concentrations (P = 0.000) and (b) lignin:P ratios (P = 0.000) in the peat. Each dot represents one of the peat samples that had become buoyant (●; n = 13) or remained inundated (○; n = 17) after rewetting of the peat surface. The relations between the C production rates and P concentrations and lignin: P ratios have been described by a power regression analysis (After Tomassen et al. 2004a)
activity of decomposing organisms, it seems likely that decomposition processes and CH4 production rates will be hampered when C:N, C:P or C:K ratios are high, and when N, P or K concentrations are low (Swift et al. 1979; Updegraff et al. 1995; Beltman et al. 1996; Smolders et al. 2002). The potential decay rates of Sphagnum litter are controlled by N and P availability and are most strongly determined by P-related litter chemistry variables (Aerts et al. 2001; Tomassen et al. 2004a). Peat with high lignin concentrations have low C production rates indicating that lignin increases with humification and this may retard the activity of decomposing organisms (Swift et al. 1979; Bozkurt et al. 2001), resulting in a slower breakdown of organic litter (Yavitt et al. 1997; Aerts and Chapin 2000). Selective removal of
306
H.B.M. Tomassen et al.
the more easily metabolised carbon compounds by decomposer organisms results in larger proportions of resistant organic compounds, such as lignin, as decomposition proceeds (Bozkurt et al. 2001). Lignins from Sphagnum mosses are particularly rich in p-hydroxyphenols, which are the most stable phenolic compounds in surface peat (Yavitt et al. 2000). Increased pH is known to enhance CH4 production by stimulating the activity of methanogenic bacteria (Fig. 15.18; Williams and Crawford 1984; Dunfield et al. 1993; Segers 1998) and by increasing the hydrolysis of organic substrates (Kok and Van de Laar 1991). The enhanced hydrolysis, in turn, results in an increased availability of substrates for methanogenic bacteria such as acetate or H2 (Lamers et al. 1999; Smolders et al. 2002). Yet, Sphagnum peat often has pH below 4 and therefore even in poorly humified peat CH4 production rates may still be low. As Lamers et al. (1999) and Smolders et al. (2002) have demonstrated, buffered groundwater may increase pH of the peat water, thereby enhancing microbial decomposition of the peat and CH4 production. The influence of buffered groundwater in the peat base led to a rapid development of floating rafts in the Haaksbergerveen bog after rewetting measures had been taken (Lamers et al. 1999). In contrast to intact peat bogs, remaining peat layers in cut-over bog remnants may be relatively thin, and the influence of buffered groundwater in the peat base (or influence of underlying (buffered) fen peat or calcareous soil layers) may have a significant effect on decomposition processes in the residual peat by buffering the peat water pH (Wheeler and Shaw 1995; Lamers et al. 1999; Smolders et al. 2002). It is especially in remnants with a relatively thin peat layer that it is important to know whether the peat base is buffered by calcareous deposits and/or calcareous groundwater. Measurements in various peat remnants in the Netherlands, Estonia, and Ireland revealed that peat base buffering can be detected by analyzing the pH and bicarbonate and calcium concentrations of the peat base water. For those locations in which calcium bicarbonate buffering occurred, a linear relation was found
Fig. 15.18 The effects of incubation pH on the production of methane and carbon dioxide in peat from a cut-over bog remnant (Haaksbergerveen). Different characters in the bars indicate significant differences (one-way ANOVA; P < 0.05) (After Smolders et al. 2002)
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
307
Fig. 15.19 Bicarbonate concentration and pH plotted against calcium concentrations in peat water from the peat base of various bog remnants in the Netherlands, Ireland and Estonia. Locations without (●) and locations with calcium (bi)carbonate buffering (□) of the peat base are discriminated (After Smolders et al. 2003)
between the calcium concentrations and the pH and bicarbonate concentration of the peat base water (Fig. 15.19). Calcium concentrations exceeded 188 µmol l−1 and pH and bicarbonate concentration exceeded pH 5.0 and 100 µmol l−1 respectively. At the other locations, the peat base calcium concentrations were lower than 100 µmol l−1 and pH and bicarbonate concentration where lower than pH 5.0 and 100 µmol l−1, respectively. At these locations, no correlation existed between the calcium concentration and the pH or bicarbonate concentration (Fig. 15.19). In large parts of the Netherlands, however, groundwater is strongly enriched with sulphate (Lamers et al. 1998). Sulphate can hamper methanogenesis because sulphate-reducing and methanogenic bacteria compete for substrates (Bhattacharya et al. 1996; Lamers et al. 1999; Smolders et al. 2002). Sulphate-containing groundwater tends to inhibit methane production (Fig. 15.20) and increase the mobilisation of nutrients, and may therefore have a negative effect on floating raft development and re-establishment of ombrotrophic vegetation (Lamers et al. 1999; Smolders et al. 2002). Sulphate enriched (ground) water should therefore never be used to inundate bogs. If floating rafts are to develop after deep inundation of cut-over bogs, poorly humified peat must be present. Table 1 summarises some of the physical and chemical characteristics of peat that predispose peat to becoming buoyant after deep inundation. The ratio of pore water pH (squeezed from the peat) to peat bulk density appears to be a simple and reliable indicator of whether the peat is suitable for the formation of floating rafts. Peat that is suitable for floating raft formation has a pH above 4.0 and a bulk density below 75 g l−1, resulting in a pH:(bulk density) ratio above 0.05. This pH to bulk density ratio is easy for nature managers to measure. Among the other appropriate peat characteristics for determining buoyancy are the C:P and lignin:P ratio, and the size of the peat particles (Table 15.1).
308
H.B.M. Tomassen et al.
Fig. 15.20 The inhibiting effect of initial sulphate concentrations (µmol l−1) in peat water on the methane production of incubated peat (n = 4) from a cut-over bog remnant (Haaksbergerveen). Results of ANOVAs for repeated measures are as follows: date (P < 0.001); sulphate concentration (P < 0.001); interaction (P < 0.001) (After Smolders et al. 2002) Table 15.1 Physical and chemical prerequisites for peat able to form floating rafts after deep inundation of cut-over bogs. Data are based on the analysis of buoyant peat collected from 6 locations in the Netherlands (n = 13) (After Tomassen et al. 2004a) Chemical characteristics Buoyant peat pH ³4.0 Bulk density (g DW l−1 FW) £75 Fraction <1 mm £0.50 Fraction >5 mm ³0.40 P (µmol g−1 DW) ³10 C:P ratio (g g−1) £3,000 N:P ratio (g g−1) £75 CH4 production rate (µmol g−1 DW day−1) ³2 Total C production rate (µmol g−1 DW day−1) ³3 Hemicellulose (mg g−1 DW) ³220 Lignin (mg g−1 DW) £300 £1,000 Lignin:P ratio (g g−1) Lignin:N ratio (g g−1) £20 CH4 production rate (µmol l−1 FW day−1) ³150 pH:(bulk density) ratio ³0.05
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
309
In most Dutch cut-over bogs, however, the residual peat is often inadequate for floating raft formation, since it is mostly the strongly humified catotelm peat which is left after peat harvesting. This strongly humified peat does not become buoyant after inundation (Smolders et al. 2002; Tomassen et al. 2004a). In some areas, the surface layer of the peat including its vegetation (usually referred to by the German term bunkerde), has been returned after peat harvesting. After inundation, this poorly humified bunkerde became buoyant, providing a substrate for Sphagnum colonisation. In these locations, floating rafts subsequently developed, even though the residual peat was strongly humified. The introduction of weakly humified organic substrates from donor sites, for instance in the form of material produced by sod-cutting in wet heathlands, might be used in inundated compartments where no regeneration of floating rafts occurs. This measure is similar to returning the bunkerde. A feasibility study (Tomassen et al. 2003b) has recently been carried out successfully. Poorly humified substrates derived from sod-cutting in wet heathlands and from peat cutting activities in bogs all appeared to become buoyant if pore water pH was higher than 4.5 (Fig. 15.21). If the substrate was too acidic, incorporation of small amounts of lime was necessary to raise its pH and to stimulate CH4 production and so buoyancy of the substrate (Smolders et al. 2003; Tomassen et al. 2003b). The amount of lime that has to be added to obtain these values will depend on the acidity of the substrate, but in the substrates used in the experiment this did not exceed 4 g kg−1 FW (Fig. 15.21).
Fig. 15.21 Peat water pH and methane concentrations (n = 4) in peat from four locations and mixed with various amounts of lime in July 2001. The substrates within the circle were buoyant, while those outside the circle were completely or partly inundated. The numbers indicate the amount of lime added to the peat (0–8 g lime kg−1 fresh peat) (After Tomassen et al. 2003b)
310
H.B.M. Tomassen et al.
Higher pH values will not only lead to increased CH4 production but also to higher CO2 production rates. Additionally, the dissolution of added calcium carbonate will also lead to CO2 liberation. Increased benthic CO2 concentrations may be beneficial, as they are known to stimulate the growth of Sphagnum establishing on the newly formed floating rafts (Lamers et al. 1999; Smolders et al. 2001). In inundated peat remnants where the amount of poorly humified peat left is insufficient, the introduction of substrates might be a valuable method to stimulate floating raft formation. Of course, its feasibility will greatly depend on the availability and quality of donor substrate. The introduction of Sphagnum mosses will probably be necessary to re-vegetate the bare substrates once they become buoyant. Re-vegetation of the substrates is important because anaerobic decomposition of fresh and thus easily decomposable organic material has to provide sufficient CH4 to warrant buoyancy on the long term. Lime addition levels should not be higher than necessary, as there is some evidence that calcium toxicity may occur when calcium concentrations exceed 800 µmol l−1 (Money 1995; Sliva and Pfadenhauer 1999). Furthermore, the addition of large amounts of lime might lead to eutrophication due to excessive decomposition rates of the peat (Lamers et al. 1999; Smolders et al. 2002). Based on the above reported effect of several physical and chemical characteristics of the peat on C production rates, we conclude that it is possible to determine if peat has the appropriate composition for the development of floating rafts. If the remaining peat layers in a cut-over bog are strongly decomposed, deep inundation (>0.5 m) is not advisable unless poorly humified peat can be introduced.
15.4.3 Mariapeel 15.4.3.1 Introduction The Mariapeel bog (1,199 ha; 51°25¢N; 5°55¢E) is located in the South-eastern part of the Netherlands (Fig. 15.3). Small scale peat extraction started already in the Middle Ages. Large scale peat harvesting occurred until the mid of the twentieth century. As a result of the extensive system of canals and ditches (dug into the sandy subsoil) the area was severely desiccated. In addition, extensive animal husbandry in the area resulted in the highest atmospheric nitrogen loads of the Netherlands (up to 10 g N m−2 year−1). The inlet of water from the river Meuse in the past has increased the nutrient influx in the Mariapeel. As a result of small scale peat extraction small peat pits arose. Peat farmers could only harvest peat during one day from a particularly location since the next day the pit was filled with water. In these small pits (in Dutch Boerenkuilen) bog regeneration took place by the development of a dense layer of submerse Sphagnum cuspidatum. In such a S. cuspidatum raft hummock forming bog mosses can establish after a while. Nowadays these pits harbour characteristic bog species such as Sphagnum magellanicum, S. papillosum, Vaccinium oxycoccos and Andromeda polifolia.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
311
15.4.3.2 Hydrological Measures Since 1993, rewetting measures were taken to preserve the bog. Rewetting started with hydrological measures taken outside the bog and in the border area by the Water board. Groundwater is preserved by the construction of hydrological buffer zones in the agricultural area surrounding the Mariapeel bog. In 1997–1998, six dams (5.7 km length) were constructed on the border between bog and agricultural land. These dams, constructed from sand, were connected to the mineral subsoil (loam). As from 1997, no water from the river Meuse was let in to compensate for water shortage. Optimal preservation of rain water was obtained by constructing compartments (1998–1999) with limited water level differences and overflows on fixed levels (according to the model of Wheeler and Shaw 1995; Fig. 15.2). The compartments were constructed by using the natural relief, and roads and tracks present in the area. 15.4.3.3 Effects of Hydrological Measures The rewetting measures resulted in higher groundwater and surface water levels, and a reduction of water level fluctuations. In contrast to the extremely dry year of 1976, in which the Mariapeel bog was completely dried out, during the dry summer of 2003 all of the compartments retained water (reduction of the water level by 30–50 cm). The water level increase caused the die back of Betula trees and Molinia caerulea (Fig. 15.22), and an increase in Sphagnum cover. In the period 1997–2004, the area covered by Sphagnum (mainly S. cuspidatum) showed a
Fig. 15.22 Hydrological measures taken in 1999 resulted in increased water table levels, and concomitant die back of Betula trees and Molinia caerulea
312
H.B.M. Tomassen et al.
Fig. 15.23 Small floating raft in an inundated part of the Mariapeel bog. After inundation of large areas only small patches of peat became buoyant. These patches consisted of young, poorly humified, Sphagnum peat that was still present in former depressions. In the major part of the area, however, only strongly humified peat was present and no floating raft formation was observed. The stagnation of the development of the precious small peat pits was stopped by the rewetting
tenfold increase (Bossenbroek et al. 2005). Within 5–6 years Sphagnum layers up to 0.5 m thickness developed. In former ditches that were already filled up with new peat, inundation led to the development of floating peat (Fig. 15.23). These rafts were rapidly colonised by Sphagnum cuspidatum, but due to the high nutrient availability Molinia caerulea and Juncus effusus were also able to establish (Fig. 15.26). A dense vascular plant layer hampers the growth of bog mosses by shading (see also Section 15.4.1.5). Hummock forming bog mosses such as S. papillosum, S. magellanicum restarted their growth and Vaccinium oxycoccos, Andromeda polifolia, Rhynchospora alba and Drosera rotundifolia expanded rapidly. In part of the Mariapeel bog (Horsterdriehoek and Kanaalbos) rewetting led to the mobilisation of nutrients from the former desiccated peat layer (internal eutrophication). Within a year a dense floating vegetation dominated by Spirodela polyrhiza (95% cover) and Azolla filiculoides (5–10% cover) developed (Fig. 15.24). During the last century the sulphur deposition was extremely high. After rewetting these oxidised sulphur compounds were reduced to sulphides which bound strongly to iron-phosphate complexes in the peat. As a result phosphate is mobilised (60 µmol
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
313
Fig. 15.24 Rewetting of the Horsterdriehoek area (Mariapeel) led to a strong mobilisation of nutrients, resulting in dominance of Spirodela polyrhiza and Azolla filiculoides
l−1 o-PO43− and 500 µmol l−1 NH4+ in the top layer of the peat). To combat this negative development, each March part of the water layer was drained away to lower the nutrient availability (Bossenbroek et al. 2005). The explosive development of Spirodela and Azolla has been drastically reduced by this method. 15.4.3.4 Bog Restoration at High Nutrient Availability The high availability of nutrients (N and P) in the Mariapeel bog (Fig. 15.25) can have negative effects on the vegetation development. Successful long-term restoration, for instance, requires the colonisation of the newly formed rafts by Sphagnum, since high methane production rates depend on the continuous availability of poorly humified peat. New floating rafts are normally colonised first by Sphagnum cuspidatum or S. fallax. Other Sphagnum species and vascular plants may establish later. Due to differences in decomposability between the Sphagnum species that are characteristic of hollows (S. cuspidatum and S. fallax) and species forming hummocks (e.g., S. magellanicum and S. papillosum), peat accumulation proceeds much more slowly when only species of hollows are present (Johnson and Damman 1991; Limpens and Berendse 2003b). In addition to the species-dependent decay of Sphagnum, nutrient concentrations of the Sphagnum litter influence decomposability (Coulson and Butterfield 1978; Aerts et al. 2001; Limpens and Berendse 2003b; Tomassen et al. 2004a). High N deposition rates affect Sphagnum litter N concentrations (Tomassen et al. 2003a)
314
H.B.M. Tomassen et al.
Fig. 15.25 Concentrations (means ± 1 SE) of phosphate and ammonium at depths of 0 - 10 cm at various Dutch bogs, not including the Mariapeel bog (NL; n = 12) and at the Mariapeel bog (MP; n = 8). * P £ 0.01 (t-test) (H.B.M. Tomassen and A.J.P. Smolders, 2002)
and thereby N release (Limpens and Berendse 2003b). Stimulated mobilisation of N results in increased availability of N for nitrophilous vascular plants, including Molinia caerulea. High availability of both N and P in the Mariapeel bog led to a rapid colonisation of newly developed rafts by Molinia and Juncus effusus (Fig. 15.26). This highly productive and dense vegetation may hamper Sphagnum growth by shading (Hayward and Clymo 1983; Heijmans et al. 2001; Limpens et al. 2004) and thereby constrain peat accumulation. In addition, this dense vegetation increases the weight of the floating raft and hampers the accumulation of CH4 because large amounts of the CH4 produced are vented into the atmosphere via the aerenchyma of these vascular plants (Lloyd et al. 1998). Ultimately, these unfavourable processes may lead to the sinking of the raft (personal observations). Eutrophication is therefore expected to affect the perspectives for floating bog formation in many locations.
15.4.4 Fochteloërveen 15.4.4.1 Introduction The Fochteloërveen bog (52°59¢N; 6°24¢E) is located in the North-eastern part of the Netherlands (Fig. 15.3) and is a remnant of the former extended ‘Smildigervenen’ bog area. Large scale peat extraction has taken place between 1,600 and 1,900 and resulted in two bog remnants: ‘Witterveld’ and ‘Fochteloërveen’ (2,700 ha). In 1938, the central part (200 ha) of the Fochteloërveen bog was saved when it was acquired by two nature conservation organisations. The central part of the bog is
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
315
Fig. 15.26 Rewetting measures taken at the Mariapeel nature reserve stimulated the development of floating rafts colonised by Sphagnum cuspidatum (a). Within a few years, however, Molinia caerulea and Juncus effusus invaded the raft (b) due to high nutrient availability (both N and P)
still characterised by a relatively thick peat layer (1.5–2.0 m) but the acrotelm has been destroyed and an extensive system of drainage ditches has been constructed for the cultivation of buckwheat. 15.4.4.2 Hydrological Measures Since the 1960s hydrological measures have been taken to conserve water in the area. Among the measures that were taken in the past are blocking of drainage ditches (1965 and 1975) and the building of 15 km of dams (1984/1985). The dams consist of peat with a core of plastic foil that was connected to the impermeable ‘gliede’ layer (aerobically decomposed oligotrophic peat). Despite these rewetting measures, the Fochteloërveen bog was still severely desiccated which is characterised by the limited
316
H.B.M. Tomassen et al.
Fig. 15.27 Overview of the central part of the Fochteloërveen bog
growth of bog mosses and the dominance of Molinia caerulea in the area (Fig. 15.27). Hummock forming Sphagnum species are still present in the area. Recently, a large scale restoration plan was developed for the Fochteloërveen bog (De Vries and Van der Werff 1997). The main objective of the project was the rewetting of the central part of the bog. The compartmentalisation of the area by the construction of bunds in the 1980s has resulted in much wetter conditions, but the water level fluctuations were still very high. Within the compartments the expansion of characteristic species, such as Eriophorum vaginatum, Andromeda polifolia, Sphagnum capillifolium and S. rubellum, has taken place. The area open water also increased and on shallow inundated peat a vegetation of S. cuspidatum and Eriophorum angustifolium developed. By improving the compartmentalisation of the bog area the water level will increase and water level fluctuations will be subsidised. For this, 24.5 km of new bunds (area of 1,000 ha), constructed of pine wooden boards (length of 1.5–2.5) funded in the sandy subsoil and covered by peat (Van ‘t Hullenaar 1997), were constructed between 1999 and 2001. The peat originated from the inside of the dam. The bunds are equipped with 46 overflows to regulate the water level in the different compartments. 15.4.4.3 Effects of Hydrological Measures As a result of large scale construction of dams, the water level increased. In the central part of the bog, the water table increased approximately 30 cm and part of the still present hummocks were partly inundated. Due to the fact that we were concerned about the development of the hummock forming bog mosses we started to follow the vegetation development in eight permanent plots. The construction of the dams was finished in 1999 and we started the experiment in February 2000. At first, rewetting resulted in an increase if the area open water (Fig. 15.28 and 15.29).
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
317
Fig. 15.28 Vegetation development in the Fochteloërveen bog after rewetting. (a) First winter after rewetting (November 2001) and (b) third year after rewetting (December 2003). Molinia and open water has been replaced by Sphagnum cuspidatum, Eriophorum angustifolium and E. vaginatum
After 2 years, the open water was filled with Sphagnum cuspidatum (Fig. 15.28). Heather species (Calluna vulgaris and Erica tetralix) decreased and the cover of E. vaginatum and E. angustifolium increased strongly. Cover of Sphagnum papillosum increased and Sphagnum rubellum established (Fig. 15.29). As a result of rewetting of poorly humified peat, the peat swells up and gives optimal growing conditions for Sphagnum species. A large scale vegetation survey in the Fochteloërveen bog also revealed that in most areas the cover by Molinia caerulea has decreased and the cover of characteristic bog species has increased (Altenburg et al. 2005). Laboratory experiments with desiccated poorly humified peat sods (originating from the Tuspeel bog, Mariapeel bog, Bargerveen bog (all the Netherlands) and Clara bog (Ireland) revealed similar results after rewetting. The expansion of Eriophorum after rewetting was correlated with the nutrient availability of the sods.
318
H.B.M. Tomassen et al.
Fig. 15.29 Estimated cover of plant species after rewetting of the central part of the Fochteloërveen bog (n = 8)
The presence of Eriophorum can have positive effects on the development of Sphagnum (Wheeler and Shaw 1995; Buttler et al. 1998). During dry periods Eriophorum can prevent Sphagnum from drying out and during high water levels Sphagnum can use the Eriophorum as support (Fig. 15.30). If the cover of Eriophorum becomes too high there is a risk of growth reduction by shading.
15.4.4.4 Shallow Inundation of Peat Field observations, e.g. Fochteloërveen, have revealed that surface-soaked or very shallowly inundated (<30 cm) peat substrates present very good conditions for peat moss growth (Meade 1992; Money 1995; Wheeler and Shaw 1995; Sliva 1998; Sliva and Pfadenhauer 1999). Under waterlogged (surface-soaked) conditions, the species are shielded from drought. We conducted a laboratory experiment which showed that under these conditions (including the current Dutch background deposition levels of nitrogen and phosphorus), the introduced peat mosses grew very well on both poorly humified (white) and strongly humified (black) peat, and the hummock and lawn forming species did just as well as the hollow species
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
319
Fig. 15.30 Sphagnum fallax growing in a tussock of Eriophorum vaginatum in a recently developed bog remnant (Tuspeel, The Netherlands)
(Fig. 15.31). In the treatment where during the first 6 months the water table was kept 10 cm below the surface (‘dry’ conditions), all introduced Sphagnum species showed poor growth during this period, independent of the substrate type. However, after rewetting the growth recovered remarkably and after another year final biomasses produced did not differ significantly from the treatments that had been permanently waterlogged. In inundated conditions, the species showed very poor growth on the strongly humified peat, although none of the introduced Sphagnum species actually died completely (not even after 6 months of inundation). All species formed extremely small capitula, with thin long branches, typical symptoms of CO2 limitation (Rice and Schuepp 1995). In fact, the appearance of the mosses was far from normal and hardly similar to the normal growth form. Indeed, measurements of the CO2 concentrations showed that these concentrations in the water layer remained very low (<20 µmol l−1). Plants with this elongated growth form may be more prone to desiccation during drier events than individuals with a regular growth form (Rochefort et al. 2002). After restoring surface soaked conditions growths recovered, although at the end of the experiment the total biomass production was much lower on the strongly humified peat which had been flooded for the first half year of the experiment than in the other treatments. The inundated poorly humified peat swelled up to the newly established water level, resulting in a very good development of especially Sphagnum papillosum. It can be concluded that waterlogging provides good conditions for the establishment of peat mosses and that hummock species are doing equally well as hollow species. Temporal inundation or desiccation are well supported by the peat mosses, although particularly prolonged flooding on strongly humified peat may retard the growth of the peat mosses on the longer term, especially if CO2 concentrations in the water layer are low. Creating waterlogged conditions over large surface areas is not easy to achieve in the field over the entire hydrological season, as water levels will tend to drop too
320
H.B.M. Tomassen et al.
Fig. 15.31 Total biomass production of the different Sphagnum species at the end of the shallow inundation experiment. Species were introduced on strongly humified peat or poorly humified peat in the laboratory at three different water tables: 10 cm below peat surface (dry), surface soaked and 10 cm above peat surface (inundated). After 6 months all treatments were surface soaked for one more year. Different characters above the bars indicate significant differences between the treatments for the total biomass produced (one-way ANOVA with Tukey’s posteriori test; n = 4) (After Smolders et al. 2003)
far in summer due to evaporation (Wheeler and Shaw 1995). Hydrological compartmentalisationof the terrain, however, could create rainwater storage reservoirs in a part of the terrain. These compartments could supply water to the substrate-soaked target compartments in dry periods in order to maintain waterlogged (surface-soaked) conditions. This could be a valuable method especially in large peat remnants. In contrast to deeply inundated locations, Eriophorum species, Molinia caerulea and Juncus effusus are able to grow on shallowly inundated peat surfaces. The tussocks of these species present an ideal matrix for peat mosses in shallowly inundated surfaces (Wheeler and Shaw 1995; Sliva 1998). Therefore in the field situation shallow inundation of strongly humified peat (<30 cm) has proved to be very efficient (Sliva 1998; Sliva et al. 1997; Sliva and Pfadenhauer 1999). Once a vegetation of pioneer
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
321
species is established Sphagnum can survive temporarily high water levels because it uses the vascular plants as a ‘climbing frame’ (Sliva et al. 1997). Furthermore, these tussocks will create a favourable humid microclimate where peat mosses can survive dryer periods (Grosvernier et al. 1995; Sliva et al. 1997; Buttler et al. 1998). The presence of tussocks under surface-soaked conditions or shallow inundation will thus protect the mosses from the adverse effects of water table fluctuations in the field. To achieve such a nurse crop on bare peat surfaces, the phasing of restoration seems to be crucial. After pre-soaking, water should be drained off in order to allow the development of a suitable nurse crop (Sliva 1998). Sowing diaspores and/or introducing diaspore-rich mulch can be applied to speed up such a development (Sliva et al. 1997; Sliva and Pfadenhauer 1999). After establishment of the nurse crop, water tables should be raised sufficiently to enable Sphagnum species to grow. For desiccated bogs in which the acrotelm or poorly humified peat is still present, hydrological regulation may be assisted because these peat types possess the ability to swell and shrink according to the availability of water (Wheeler and Shaw 1995). Inundation of acrotelm or poorly humified peat thus results in a swelling of the peat surface and ultimately to wetter but not flooded conditions.
15.5 Introduction of Target Species In most restored peat bogs, Sphagnum development is largely restricted to carpets of Sphagnum cuspidatum or Sphagnum fallax (Money 1995). Sphagnum magellanicum, S. papillosum and S. rubellum, which are in fact the essential hummock and lawn forming bog species, normally do not appear in the vegetation. The reestablishment of these species in bogs under restoration is generally a slow process (Joosten 1995; Money 1995). In the Netherlands, these species are often absent or very poorly represented in the restored remnants. The reason is that in many cases no viable diaspore bank is left because the upper peat layers have been removed and only the deeper catotelm peat (or part of it) remains. Although some studies have found that dispersal limitation did not play an important role in Sphagnum species (Poschlod 1995; Rydin and Barber 2001), sporulation of especially S. magellanicum, S. papillosum and S. rubellum is rarely observed in Europe, unlike S. cuspidatum and S. fallax, which sporulate frequently (Cronberg 1991). It is also possible that the abiotic conditions are not suitable for the development of hummock and lawn forming peat mosses from spores (Cronberg 1991). The lack of hummock and lawn species in recently restored bog remnants is probably due to colonisation problems and not necessarily by the lack of suitable growth conditions. To test the idea that colonisation determines the presence of the hummock and lawn forming species we introduced small sods of S. magellanicum and S. papillosum in a well developed dense S. cuspidatum carpet which had recently developed in a drain at the edge of Clara bog (Ireland). Both species did very well in the S. cuspidatum vegetation (Fig. 15.32), and we conclude that the absence of these species in the plot is clearly not due to the lack of suitable abiotic conditions or to
322
H.B.M. Tomassen et al.
Fig. 15.32 Increase in surface covered by Sphagnum magellanicum and S. papillosum introduced at Clara Bog (County Offaly, Ireland), on a bare peat surface (a) and in a Sphagnum cuspidatum vegetation (b). The surface areas in 1999 (a) and 1998 (b) represent the introduced surface areas. All values are means of three replicates. Vertical bars represent standard deviation (After Smolders et al. 2003)
inter-specific competition at this particular location. Similar results were obtained for the introduction of S. magellanicum in a Dutch S. fallax carpet (data not shown). In a similar experiment, Sphagnum magellanicum and Sphagnum papillosum were introduced on a completely bare peat substrate, also at Clara bog. Again, both species appeared to expand strongly after their introduction (Fig. 15.32). The results are in agreement with those obtained by Rydin (1993), who found that S. fuscum was able to establish very well in hollows dominated by S. balticum and S. tenellum.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
323
The niche width of the Sphagnum species generally appears to be very wide and differences in colonising ability appear to be more important for their distribution than differences in competitive ability (Watson 1980; Rydin 1993; Campeau and Rochefort 1996; Soro et al. 1999; Rydin and Barber 2001). Once established, the hummock and lawn forming species create their own habitat, which is more hostile to the hollow species (Rydin 1993). If the acrotelm is still present in a reserve, rewetting is usually sufficient to restore Sphagnum growth. In most cases, the hummock and lawn species are still present in desiccated acrotelm peat, so one would not expect problems for these species to recover and disseminate after rewetting (Poschlod 1995; Campeau and Rochefort 1996; Sundberg and Rydin 2000). In many Dutch bog remnants, the hummock and lawn forming species are absent or very scarce and spontaneous establishment of these species will usually take at least decades and more likely even centuries (Joosten 1995). These species may be desirable from a botanical point of view, but above all they are crucial for the development of an acrotelm with the self-regulating properties which are essential for an active bog forming system (Wheeler and Shaw 1995; Joosten 1995). In our opinion, the introduction of the ‘key species’ (i.e. Sphagnum magellanicum, Sphagnum papillosum and Sphagnum rubellum) should be considered when these are absent or poorly distributed within a remnant, in order to speed up the re-establishment of acrotelm-forming Sphagnum carpets (Money and Wheeler 1999). The North American approach to restore of cut-over bogs also involves a diaspore introduction as these peatlands do not rehabilitate well spontaneously (Rochefort et al. 2003; Campeau et al. 2004). Since the introduction of species from nearby reserves or, if possible, from the same reserve is to be preferred for botanical or ethical reasons, farming Sphagnum might be necessary to acquire sufficient inoculum to successfully reintroduce the target species on a large scale (see also Money 1995).
15.6 Conclusions Based on the above-described case studies and mechanisms we have developed a scheme outlining the perspectives for peat bog restoration (Fig. 15.33). In most cut-over bog remnants strongly humified catotelm peat (black peat) is the main peat substrate. In some parts, the surface layer of vegetation and peat which is removed prior to peat cutting, known in German as bunkerde, has been returned to the cuttings afterwards and overlies the strongly humified peat. This bunkerde consists of poorly humified material and thus might have properties that favour bog regeneration. It may easily become buoyant due to relatively high methane production rates while it may also provide a source of diaspores for recolonisation (Money and Wheeler 1999). Inundation of large areas of cut-over bog remnants has been applied on a large scale and has locally resulted in substrates becoming buoyant. This has occurred especially at locations where bunkerde was returned or in patches where Sphagnum was still growing in peat pits prior to flooding, so poorly humified Sphagnum peat
324
H.B.M. Tomassen et al.
Fig. 15.33 Mechanisms involved in the restoration of bog remnants (After Smolders et al. 2003)
was still present. The influence of buffered groundwater may be important as it stimulates methane production via the buffering of acids. The growth of submerged Sphagnum species such as S. cuspidatum may also result in the development of floating rafts. However, water depth (coloration) and CO2 fluxes from the residual peat layer seem to determine the growth of submerged Sphagnum. In the deeply inundated parts (>0.5 m), Sphagnum growth is normally very poor due to light limitation. The growth of submerged Sphagnum can be stimulated by shallow inundation (<0.3 m), especially if the water is coloured by humic substances (Money 1995). If no floating rafts develop and waterlogged conditions or shallow inundation cannot be achieved, the introduction of poorly humified substrate, for instance derived from sod-cutting in wet heathlands or from peat-cutting activities in other areas, may be a good option. This measure is comparable to the restitution of bunkerde in the past. If the substrate is too acidic, mixing with lime may be necessary to raise the pH of the substrate and so stimulate CH4 production rates. It can be concluded that deep flooding of bog remnants containing only strongly humified peat is usually not very successful, while shallow inundation or surfacesoaking seems to be a very promising technique and provides good conditions for the establishment of hummock and lawn species. However, compartmentalisation of the terrain will be necessary to maintain sufficiently high water levels in summer.
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
325
Furthermore it is highly beneficial to have a good ‘nurse crop’ of pioneer species (such as Eriophorum vaginatum), which can serve as a matrix for Sphagnum and break up the open water. The poor colonisation rate by hummock and lawn species (S. magellanicum, S. papillosum and S. rubellum) seems to be the main reason why these species remain absent after restoration, at least for decades, in Sphagnum-dominated vegetation. Even if floating rafts develop, the introduction of hummock and lawn forming species may be important, as normally only S. cuspidatum and S. fallax tend to colonise the bare peat substrates. Since much effort is being devoted to maintaining favourable hydrological conditions in bog remnants under restoration, it is not very desirable to wait for more than 50 years to see whether the target species will reestablish by natural processes. Introduction of these target species will enable or facilitate the formation of an acrotelm layer with self-regulating hydrological conditions, which is a prerequisite for a functioning, carbon-fixing bog system. If any bog restoration project wants to have a reasonable chance of success, the rapid development of such an acrotelm should be one of its main goals. The introduction of ‘key species’ will often be an essential part of such a project. The high atmospheric nitrogen deposition levels (2 – 4 g N m−2 year−1) in the Netherlands are more difficult to cope with. Fortunately, N deposition levels are decreasing as a result of effective measures against the emission of N compounds taken in recent decades. A reduction in deposition can be effective in the short term. The best prospects for ombrotrophic bogs undoubtedly lie in further reduction of the emission of N compounds, as N deposition levels still greatly exceed the critical N level of 0.5–1 g m−2 year−1. The production of Sphagnum and the role of this species as an ecosystem engineer (Malmer et al. 2003) form the key to the continued existence of bog vegetation despite the excessive N deposition levels. Under optimal growing conditions (water availability, available P and C, low vascular plant cover) the Sphagnum layer should be able to cope with a total N deposition of approximately 2.5 g N m−2 year−1, keeping the availability of N to vascular plants low. Additional management measures aimed at optimising overall growth conditions for Sphagnum can help to avoid an important part of the unfavourable effects of N. Acknowledgements We thank ‘Staatsbosbeheer’ and ‘Natuurmonumenten’ for giving permission to perform the experiments in their bogs. Mark van Mullekom, Jeroen Geurts, Hein Pijnappel, Jurgen Memelink, Frank Spikmans, Ellen van Halteren and Marjo van Herk provided practical assistance. Phlip Bossenbroek gave information about restoration measures in the Mariapeel bog. Part of this research was funded by the Dutch Ministry of Agriculture, Nature and Food Quality.
References Aaby B (1994) Monitoring Danish raised bogs. In: Grünig A (ed) Mires and man. Mire conservation in a densely populated country – the Swiss experience. Swiss Federal Institute for Forest, Snow and Landscape Research, Birmensdorf, Switzerland, pp 284–300 Aerts R, Chapin FS III (2000) The mineral nutrition of wild plants revisited: a re-evaluation of processes and patterns. Adv Ecol Res 30:1–67
326
H.B.M. Tomassen et al.
Aerts R, Wallén B, Malmer N (1992) Growth limiting nutrients in Sphagnum-dominated bogs subject to low and high atmospheric nitrogen supply. J Ecol 80:131–140 Aerts R, Wallén B, Malmer N, De Caluwe H (2001) Nutritional constraints on Sphagnum-growth and potential decay in northern peatlands. J Ecol 89:292–299 Aldous AR (2002) Nitrogen retention by Sphagnum mosses: responses to atmospheric nitrogen deposition and drought. Can J Bot 80:721–731 Altenburg W, Groeneweg M, Van der Veen K (2005) Hoogveenvegetaties in het Fochteloërveen. De Levende Natuur 106:102–106 Baker RGE, Boatman DJ (1990) Some effects of nitrogen, phosphorus, potassium and carbon dioxide concentration on the morphology and vegetative reproduction of Sphagnum cuspidatum Ehrh. New Phytol 116:605–611 Barkman JJ (1992) Plant communities and synecology of bogs and heath pools in the Netherlands. In: Verhoeven JTA (ed) Fens and bogs in the Netherlands: vegetation, history, nutrient dynamics and conservation. Kluwer, Dordrecht, The Netherlands, pp 173–235 Beltman B, Kooijman AM, Rouwenhorst G, Van Kerkhoven M (1996) Nutrient availability and plant growth limitation in blanket mires in Ireland. Biology and Environment. Proc R Irish Acad 96B:77–87 Berendse F, Van Breemen N, Rydin H, Buttler A, Heijmans M, Hoosbeek MR, Lee JA, Mitchell E, Saarinen T, Vasander H, Wallén B (2001) Raised atmospheric CO2 levels and increased N deposition cause shift in plant species composition and production in Sphagnum bogs. Glob Change Biol 7:591–598 Bergman I, Svensson BH, Nilsson M (1998) Regulation of methane production in a Swedish acid mire by pH, temperature and substrate. Soil Biol Biochem 30:729–741 Bergman I, Klarqvist M, Nilsson M (2000) Seasonal variation in rates of methane production from peat of various botanical origins: effects of temperature and substrate quality. FEMS Microbiol Ecol 33:181–189 Bhattacharya SK, Uberoi V, Dronamraju MM (1996) Interaction between acetate fed sulfate reducers and methanogens. Water Res 30:2239–2246 Bobbink R, Hornung M, Roelofs JGM (1998) The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural European vegetation. J Ecol 86:717–738 Bobbink R, Ashmore M, Braun S, Flückiger W, Van den Wyngaert IJJ (2003) Empirical nitrogen critical loads for natural and semi-natural ecosystems. In: Achermann B, Bobbink R (eds) Empirical critical loads for nitrogen. Proceedings expert workshop. Swiss Agency for the Environment, Forests and Landscape, Berne, Switzerland, pp 43–170 Bossenbroek Ph, De Glopper A, Verdonschot F (2005) Veenmos in de Peel. Tussentijdse evaluatie van het verrnattingsproject Mariapeel-Deurnese peel. Natuurhistorisch Maandblad 94:222–226 Bozkurt S, Lucisano M, Moreno L, Neretnieks I (2001) Peat as a potential analogue for the longterm evolution in landfills. Earth Sci Rev 53:95–147 Bridgham SD, Richardson CJ (1992) Mechanisms controlling soil respiration (CO2 and CH4) in southern peatlands. Soil Biol Biochem 24:1089–1099 Brown A, Mathur SP, Kushner DJ (1989) An ombrotrophic bog as a methane reservoir. Global Biogeochem Cy 3:205–213 Buttler A, Grosvernier P, Matthey Y (1998) Development of Sphagnum fallax diaspores on bare peat with implications for the restoration of cut-over bogs. J Appl Ecol 35:800–810 Campeau S, Rochefort L (1996) Sphagnum regeneration on bare peat surfaces: field and greenhouse experiments. J Appl Ecol 33:599–608 Campeau S, Rochefort L, Price JS (2004) On the use of shallow basins to restore cutover peatlands: plant establishment. Restor Ecol 12:471–482 Casper P, Maberly SC, Hall GH, Finlay BJ (2000) Fluxes of methane and carbon dioxide from a small productive lake to the atmosphere. Biogeochemistry 49:1–19 Coulson JC, Butterfield J (1978) An investigation of the biotic factors determining the rates of plant decomposition on blanket bog. J Ecol 66:631–650 Cronberg N (1991) Reproductive biology of Sphagnum. Lindbergia 17:69–82
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
327
Damman WH (1988) Regulation of nitrogen removal and retention in Sphagnum bogs and other peatlands. Oikos 51:291–305 De Vries HJ, Van der Werff L (1997) Integraal restauratieplan Fochteloërveen. Report Arcadis Heidemij Advies, The Netherlands Dunfield P, Knowles R, Dumont R, Moore TR (1993) Methane production and consumption in temperate and subarctic peat soils: response to temperature and pH. Soil Biol Biochem 25:321–326 Ganzevles PHJ (1991) Haaksbergerveen. Beheersplan voor de Periode 1991–2001. Report Buro Hemmen Gorham E (1991) Northern peatlands: role in the carbon cycle and probable responses to climatic warming. Ecol Appl 1:182–195 Gorham E, Rochefort L (2003) Peatland restoration: a brief assessment with special reference to Sphagnum bogs. Wetlands Ecol Manage 11:109–119 Grosvernier P, Matthey Y, Buttler A (1995) Microclimate and physical properties of peat: new clues to the understanding of bog restoration processes. In: Wheeler BD, Shaw SC, Fojt WJ, Robertson RA (eds) Restoration of temperate wetlands. Wiley, Chichester, pp 435–450 Hassink J (1995) Density fractions of soil macroorganic matter and microbial biomass as predictors of C and N mineralization. Soil Biol Biochem 8:1099–1108 Hayward PM, Clymo RS (1983) The growth of Sphagnum: experiments on, and simulation of, some effects of light flux and water-table depth. J Ecol 71:845–863 Heijmans MMPD, Berendse F, Arp WJ, Masselink AK, Klees H, De Visser W, Van Breemen N (2001) Effects of elevated carbon dioxide and increased nitrogen deposition on bog vegetation in the Netherlands. J Ecol 89:268–279 Van ‘t Hullenaar JW (1997) Hydrologisch inrichtingsplan voor hoogveengeneratie in het Fochteloërveen. Report Hullenaar Ecohydrologisch adviesbureau, Zwolle, The Netherlands Ingram HAP (1978) Soil layers in mires: function and terminology. J Soil Sci 29:224–227 Johnson LC, Damman AWH (1991) Species-controlled Sphagnum decay on a South Swedish raised bog. Oikos 61:234–242 Joosten JHJ (1995) Time to regenerate: long-term perspectives of raised bog regeneration with special emphasis on palaeoecological studies. In: Wheeler BD, Shaw SC, Fojt WJ, Robertson RA (eds) Restoration of temperate wetlands. Wiley, Chichester, pp 379–404 Kirk JTO (1994) Light and photosynthesis in aquatic ecosystems. Cambridge University Press, Cambridge Kok CJ, Van de Laar BJ (1991) Influence of pH and buffering capacity on the decomposition of Nymphaea alba L. detritus in laboratory experiments: a possible explanation for the inhibition of decomposition at low alkalinity. Verhandlungen der Internationale Verein für Theoretische und Angewandte Limnologie 24:2689–2692 Lamers LPM, Tomassen HBM, Roelofs JGM (1998) Sulfate-induced eutrophication and phytotoxicity in freshwater wetlands. Environ Sci Technol 32:199–205 Lamers LPM, Farhoush C, Van Groenendael JM, Roelofs JGM (1999) Calcareous groundwater raises bogs; the concept of ombrotrophy revisited. J Ecol 87:639–648 Lamers LPM, Bobbink R, Roelofs JGM (2000) Natural nitrogen filter fails in polluted raised bogs. Global Change Biol 6:583–586 Li Y, Vitt DH (1994) The dynamics of moss establishment: temporal responses to nutrient gradients. The Bryologist 97:357–364 Limpens J, Berendse F (2003a) Growth reduction of Sphagnum magellanicum subjected to high nitrogen deposition: the role of amino acid nitrogen concentration. Oecologia 135:339–345 Limpens J, Berendse F (2003b) How litter quality affects mass loss and N loss from decomposing Sphagnum. Oikos 103:537–547 Limpens J, Berendse F, Klees H (2003a) N deposition affects N availability in interstitial water, growth of Sphagnum and invasion of vascular plants in bog vegetation. New Phytol 157:339–347 Limpens J, Raymakers JTAG, Baar J, Berendse F, Zijlstra JD (2003b) The interactions between epiphytic algae, a parasitic fungus and Sphagnum as affected by N and P. Oikos 103:59–68
328
H.B.M. Tomassen et al.
Limpens J, Berendse F, Klees H (2004) How P availability affects the impact of N deposition on Sphagnum and vascular plants in bogs. Ecosystems 7:793–804 Lloyd D, Thomas KL, Benstead J, Davies KL, Lloyd SH, Arah JRM, Stephens KD (1998) Methanogenesis and CO2 exchange in an ombrotrophic peat bog. Atmosph Environ 32:3223–3238 Lütke Twenhöven F (1992) Competition between two Sphagnum species under different deposition levels. J Bryol 17:71–80 Malmer N, Albinsson C, Svensson M, Wallén B (2003) Interferences between Sphagnum and vascular plants: effects on plant community structure and peat formation. Oikos 100:469–482 Meade R (1992) Some early changes following the rewetting of a vegetated cutover peatland surface at Danes Moss, Cheshire, UK, and their relevance to conservation management. Biol Conserv 61:31–40 Middleboe AL, Markager S (1997) Depth limits and minimum light requirements of freshwater macrophytes. Freshw Biol 37:553–568 Money RP (1995) Re-establishment of a Sphagnum-dominated flora on cut-over lowland raised bogs. In: Wheeler BD, Shaw SC, Fojt WJ, Robertson RA (eds) Restoration of temperate wetlands. Wiley, Chichester, pp 405–422 Money RP, Wheeler BD (1999) Some critical questions concerning the restorability of damaged raised bogs. Appl Veg Sci 2:107–116 Neijenhuijs F (1973) Raised bogs in the Netherlands: a disappearing type of landscape? Natuur Landschap 27:98–126 Nordin A, Gunnarsson U (2000) Amino acid accumulation and growth of Sphagnum under different levels of N deposition. Ecoscience 7:474–480 Paffen BPG, Roelofs JGM (1991) Impact of carbon dioxide and ammonium on the growth of submerged Sphagnum cuspidatum. Aquat Bot 40:61–71 Poschlod P (1995) Diaspore rain and diaspore bank in raised bogs and implications for the restoration of peat-mined sites. In: Wheeler BD, Shaw SC, Fojt WJ, Robertson RA (eds) Restoration of temperate wetlands. Wiley, Chichester, pp 471–494 Price JS, Heathwaite AL, Baird AJ (2003) Hydrological processes in abandoned and restored peatlands: an overview of management approaches. Wetlands Ecol Manage 11:65–83 Proctor MCF (1995) The ombrogenous bog environment. In: Wheeler BD, Shaw SC, Fojt WJ, Robertson RA (eds) Restoration of temperate wetlands. Wiley, Chichester, pp 287–303 Rice SK, Schuepp PH (1995) On the ecological and evolutionary significance of branch and leaf morphology in aquatic Sphagnum (Sphagnaceae). Am J Bot 82:833–846 Riis T, Sand-Jensen K (1997) Growth reconstruction and photosynthesis of aquatic mosses: influence of light, temperature and carbon dioxide at depth. J Ecol 85:359–372 Robert CE, Rochefort L, Garneau M (1999) Natural revegetation of two block-cut mined peatlands in eastern Canada. Can J Bot 77:447–459 Rochefort L, Campeau S, Bugnon JL (2002) Does prolonged flooding prevent or enhance regeneration and growth of Sphagnum? Aquat Bot 74:327–341 Rochefort L, Quinty F, Campeau S, Johnson K, Malterer T (2003) North American appoach to the restoration of Sphagnum dominated peatlands. Wetlands Ecol Manage 11:3–20 Rydin H (1993) Interspecific competition among Sphagnum mosses on a raised bog. Oikos 66:413–423 Rydin H, Barber KE (2001) Long-term and fine-scale coexistence of closely related species. Folia Geobotanica 36:53–61 Schouten MGC, Schouwenaars JM, Esselink H, Lamers LPM, Van der Molen PC (1998). Bog restoration in the Netherlands – dreams and reality. In: Bobbink R, Roelofs JGM, Tomassen HBM (eds) Nature management and the conservation of biodiversity in the Netherlands. Nijmegen, the Netherlands: Katholieke Universiteit Nijmegen, in Dutch, pp 93–112 Schouwenaars JM, Esselink H, Lamers LPM, Van der Molen PC (2002) Restoration of peat bogs in the Netherlands. Current knowledge and future research. Report 2002/084. Ministry of Agriculture, Nature Management and Fisheries, Wageningen, The Netherlands
15 Restoration of Raised Bogs: Mechanisms and Case Studies from the Netherlands
329
Scott KJ, Kelly CA, Rudd JWM (1999) The importance of floating peat to methane fluxes from flooded peatlands. Biogeochemistry 47:187–202 Segers R (1998) Methane production and methane consumption; a review of processes underlying wetland methane fluxes. Biogeochemistry 41:23–51 Sliva J (1998) Regeneration of milled peat bog: a large scale approach in Kollerfilze (Bavaria, Southern Germany). In: T. Malterer T, Johnson K, Stuart J (eds) Peatland restoration and reclamation. Proceedings of the 1998 international peat symposium, Duluth, Minnesota. International Peat Society, Duluth, MN, pp 82–87 Sliva J, Pfadenhauer J (1999) Restoration of cut-over raised bogs in southern Germany – a comparison of methods. Appl Veg Sci 2:137–148 Sliva J, Maas D, Pfadenhauer J (1997) Rehabilitation of milled fields. In: Parkyn L, Stoneman RE, Ingram HAP (eds) Conserving peatlands. CAB International, Wallingford, pp 295–314 Smolders AJP, Tomassen HBM, Pijnappel H, Lamers LPM, Roelofs JGM (2001) Substratederived CO2 is important in the development of Sphagnum spp. New Phytol 152:325–332 Smolders AJP, Tomassen HBM, Lamers LPM, Lomans BP, Roelofs JGM (2002) Peat bog formation by floating raft formation: the effects of groundwater and peat quality. J Appl Ecol 39:391–401 Smolders AJP, Tomassen HBM, Van Mullekom M, Lamers LPM, Roelofs JGM (2003) Mechanisms involved in the re-establishment of Sphagnum-dominated vegetation in rewetted bog remnants. Wetlands Ecol Manage 11:403–418 Soro A, Sundberg S, Rydin H (1999) Species diversity, niche metrics and species associations in harvested and undisturbed bogs. J Veg Sci 10:549–560 Sundberg S, Rydin H (2000) Experimental evidence for a persistent spore bank in Sphagnum. New Phytol 148:105–116 Swift MJ, Heal OW, Anderson JM (1979) Decomposition in terrestrial ecosystems. University of California Press, Berkely, CA Tomassen HBM, Smolders AJP, Lamers LPM, Roelofs JGM (2003a) Stimulated growth of Betula pubescens and Molinia caerulea on ombrotrophic bogs: role of high levels of atmospheric nitrogen deposition. J Ecol 91:357–370 Tomassen HBM, Smolders AJP, Van Herk JM, Lamers LPM, Roelofs JGM (2003b) Restoration of cut-over bogs by floating raft formation: an experimental feasibility study. Appl Veg Sci 6:141–152 Tomassen HBM, Smolders AJP, Lamers LPM, Roelofs JGM (2004a) Development of floating rafts after the rewetting of cut-over bogs: the importance of peat quality. Biogeochemistry 71:69–87 Tomassen HBM, Smolders AJP, Limpens J, Lamers LPM, Roelofs JGM (2004b) Expansion of invasive species on ombrotrophic bogs: desiccation or high N deposition? J Appl Ecol 41:139–150 Tomassen HBM, Smolders AJP, Lamers LPM, Roelofs JGM (2005) How bird droppings can affect the vegetation composition of ombrotrophic bogs. Can J Bot 83:1046–1056 Updegraff K, Pastor J, Bridgham SD, Johnston CA (1995) Environmental and substrate controls over carbon and nitrogen mineralization in northern wetlands. Ecol Appl 5:151–163 Van den Pol-van Dasselaar A, Oenema O (1999) Methane production and carbon mineralisation of size and density fractions of peat soils. Soil Biol Biochem 31:877–886 Van der Schaaf S (2002) Using surface topography to assess potential and actual ecological conditions in Irish Midland raised bogs. Annals of Warsaw Agricultural University. Land Reclam 33:49–56 Van der Schaaf S, Streefkerk JG (2002) Relationships between biotic and abiotic conditions. In: Schouten MGC (ed) Conservation and restoration of raised bogs. Geological, hydrological and ecological studies. Dúchas, The Heritage Service of the Environment and Local Government, Geological Survey of Ireland, Dublin, Ireland, Staatsbosbeheer, pp 186–209 Watson MA (1980) Shifts in patterns of microhabitat occupation by six closely related species of mosses along a complex altitudinal gradient. Oecologia 47:46–55
330
H.B.M. Tomassen et al.
Weber CA (1900) Ueber die Moore mit besonderer berücksichtigung der zwischen Unterweser und Unterelbe liegenden. Jahresbericht der Männer vom Morgenstern 3:3–23 Weber CA (1907) Aufbau und Vegetation der Moore Norddeutschlands. Beiblatt zu Englers Bot Jahrb 90:19–34 Wetzel RG (2001) Limnology: lake and river ecosystems, 3rd edn. Academic, San Diego, CA Wheeler BD, Shaw SC (1995) Restoration of damaged peatlands. Department of the Environment, HMSO, London Williams RT, Crawford RL (1984) Methane production in Minnesota peatlands. Appl Environ Microbiol 47:266–271 Yavitt JB, Lang GE, Wieder RK (1987) Control of carbon mineralization to CH4 and CO2 in anaerobic, Sphagnum-derived peat from Big Run Bog, West Virginia. Biogeochemistry 4:141–157 Yavitt JB, Williams CJ, Wieder RK (1997) Production of methane and carbon dioxide in peatland ecosystems across North America: effects of temperature, aeration, and organic chemistry of peat. Geomicrobiol J 14:299–316 Yavitt JB, Williams CJ, Wieder RK (2000) Controls on microbial production of methane and carbon dioxide in three Sphagnum-dominated peatland ecosystems as revealed by a reciprocal field peat transplant experiment. Geomicrobiol J 17:61–88
Chapter 16
Restoration of Drained Mires in the Šumava National Park, Czech Republic Ivana Bufková, František Stíbal, and Eva Mikulášková
Abstract About 70% of mires in the Šumava Mountains have been influenced by past drainage for variously forestry, agriculture and peat extraction. Since 1999, a comprehensive “Mire Restoration Programme” has been implemented in the Šumava National Park, primarily focused on the disturbed hydrology. The main restoration technique used has been the blocking of drainage ditches. Selected drained and intact mires were also monitored with the aim: (i) to characterise the degradation changes induced by the disturbed hydrology; and (ii) to evaluate the success of restoration. Water-table fluctuations, hydrochemistry of groundwater, surface-water outflow, amount of precipitation and the vegetation on permanent plots have been monitored. Pre-restoration monitoring data had shown differences between drained and intact sites. On drained bog sites, water tables were maintained at lower levels and exhibited higher fluctuations compared to intact sites. Expansion of the more competitive grasses and trees towards the bog expanse was recorded on drained bogs as well as expansion of dwarf shrub vegetation and reduction of Trichophorum lawns and hollow vegetation. Preliminary results from the initial post-restoration phase have shown a rise in the water table, stabilisation of the hydrological regime and some changes in hydrochemistry at restored sites. Keywords Drainage • Hydrochemistry • Mire restoration • Vegetation
16.1 Introduction In the densely-populated area of central Europe, mires have been used for many purposes. These include, for instance, agriculture, traditional fuel, forestry or horticulture (Joosten and Clarke 2002; Bragg and Lindsay 2003). In response to land I. Bufková (*), F. Stíbal, and E. Mikulášková Administration of the Šumava National Park, Sušická 399, 341 92, Kašperské Hory, Czech Republic e-mail:
[email protected] M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_16, © Springer Science+Business Media B.V. 2010
331
332
I. Bufková et al.
use and human demands, mires have been subject to artificial drainage over many centuries (Holden et al. 2004). The large-scale drainage of mires for agriculture (30% of original area), forestry (15%) and peat harvesting (6%) has reduced the area of natural mires to less than half their original extension (Joosten 1997). Much like in the rest of Europe, mires in the Šumava Mountains have been influenced by various human activities in the past (Schreiber 1924), and amongst these human impacts, artificial drainage represents one of the most important threats. Mires in Šumava have been drained for forestry, agriculture and peat-extraction purposes since the nineteenth century (Schreiber 1924). Almost 70% of the peatlands have been found to have been influenced at least once by drainage; however, the intensity of disturbance varies greatly across the area. Shallow surface ditches from the turn of the nineteenth and twentieth century are widespread but do not represent the most severe impact. In contrast, the relatively scarce but very intensive drainage that was implemented from the 1960s to the 1980s constitutes one of the most prevalent and serious mire conservation problems. Numerous mires, and especially ombrotrophic raised bogs, have been traditionally preserved in the Šumava Mountains as strictly protected areas since the first half of the twentieth century. Non-intervention has been the preferred management at these sites despite some having been drained in the past. Current degradation changes on disturbed mires, although resulting from past human impacts, have created the need for a more active approach to ensure the conservation of mires in the long term. As a result, the long-term “Mire Restoration Programme in the Šumava Mountains” was implemented in 1999 – a programme primarily focused on mires with a disturbed hydrology, including some restoration projects dealing with industrially-cut peatbogs or mires disturbed by road construction. The main programme goals are: (i) the restoration of a natural or near-natural mire hydrology; (ii) enhancement of natural mire biodiversity and peat-forming vegetation; and (iii) involvement of the public in local mire conservation. Within the implemented Mire Restoration Programme, mires are viewed generally in relation to the surrounding landscape; this means that not only the individual mires but the entire hydrological units representing their small subcatchments are restored. The Programme thus reflects the much recommended approach of carrying out restoration projects on a broader scale – thus also encompassing buffer zones outside the immediate area of peat (Labadz et al. 2002). The headwaters, as well as upstream catchments, were also held as general priorities. At present, the total area that has undergone restoration is about 500 ha and includes more than 40 km of blocked drainage ditches. Since 2005, the Mire Restoration Programme has been coupled with a monitoring programme aimed at studying the degradation changes caused by drainage and evaluating the success of the restoration efforts. This chapter aims to characterise the main symptoms of degradation found in drained mires and show some of the first responses of mires to the restoration measures applied. The restoration methods and techniques used in the area are also presented.
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
333
16.2 Study Area 16.2.1 Site Location and Environmental Condition The Šumava National Park (Šumava NP, 700 km2) lies about 200 km southwest of Prague in the southwestern limits of the Czech Republic (Fig. 16.1). The Šumava Mountains (‘Šumava Mts.’ or often referred to in English as the ‘Bohemian Forest’) is an area that forms one of the most important mire regions in central Europe. More than 6,000 ha of peatlands occur between the altitudes of 700–1,200 m asl. Both ombrotrophic peatbogs and minerotrophic mires have developed here, especially on the central plateau and flat river valleys (Svobodová et al. 2002; Bufková et al. 2006). Ombrotrophic peatbogs include both patterned raised bogs located at higher altitudes and valley raised bogs covered by dwarf shrubs or bog pine forest with Pinus rotundata found in river floodplains. Transitional sedge mires and extensive waterlogged spruce forests are also frequent (Soukupová 1996). All the sites studied are situated on the central plateau Modravské slatě at an altitude of about 1,100 m asl. The geological bedrock is formed mainly of paragneisses, with some granite in places. The mean annual temperature is 3.2°C and annual precipitation 1,330 mm (Svobodová et al. 2002).
16.2.2 Description of Study Sites Four mire complexes, each representing a small subcatchment, were studied in the area – two drained mires and two intact ones. Some basic characteristics of all four sites are summarised in Table 16.1.
Fig. 16.1 Location of the Šumava National Park (study area)
334 Table 16.1 Basic characteristics of monitored sites Total Recent bog Site Code Disturbances area (ha) area (ha) Blatenská slať B Intact 10.7 7 mire Šárecká slať mire SR Intact 18.3 13 Schachtenfilz SCH Drained 4.9 1.2 mire Nad Rybárnou R Drained 6.1 0.3 mire
I. Bufková et al.
Altitude 1,250
Time of restoration –
1,020 1,140
– September– October 2008 July–August 2008
1,020
16.2.2.1 Blatenská slať Mire (B) Blatenská slať mire is an example of a relatively undisturbed bog habitat. It is located on the central plateau of the Šumava Mts. at an altitude of about 1,250 m. The site includes an ombrotrophic raised bog of about 7 ha surrounded by waterlogged spruce forest. The raised bog is still active with a large open bog expanse. It is a good example of a patterned sloping bog with a well-developed surface hummock-hollow microtopography and several bog pools. The peat depth is about 7 m and the vegetation cover is of a very natural character. The open bog expanse is covered by a mosaic of lawns with Trichophorum caespitosum, hollows with Carex limosa, and hummocks with dwarf shrub vegetation of Vaccinium uliginosum. The marginal rand is formed of Pinus × pseudopumilio krummholz. The waterlogged spruce forest in the bog’s surroundings is currently in a ‘break-up’ phase due to the expansion of bark beetle. The whole mire complex slopes in a west to east direction. 16.2.2.2 Šárecká slať Mire (SR) Šárecká slať mire also represents an intact bog habitat. It is located on the central plateau at an altitude of about 1,020 m. The ombrotrophic raised bog has an area of about 13 ha surrounded by waterlogged spruce forest. The still-active mire is characterised by an open bog expanse with a complex hummock-hollow microtopography and several bog pools. Peat depth is about 6–7 m. The natural vegetation cover includes a mosaic of lawns with Trichophorum caespitosum, hollows with Carex limosa and hummock vegetation with Vaccinium uliginosum. The marginal rand of Pinus x pseudopumilio krummholz is well developed. The waterlogged spruce forest surrounding the bog is still uninfluenced by the bark-beetle expansion. The mire complex slopes in a W–E direction. 16.2.2.3 Schachtenfilz Mire (SCH) Schachtenfilz mire is a local example of a drained mire. It is located in the western part of the central plateau at an altitude of about 1,140 m. This small ombrotrophic raised bog (only 1.2 ha in area) is covered predominantly by dwarf shrubs of
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
335
Vaccinium uliginosum forming a complex mosaic with the lawns of Trichophorum caespitosum. The hollow vegetation is missing. Dwarf spruce trees are scattered over the treeless bog vegetation. Peat depth is about 5 m. Around the bog, a rand of waterlogged spruce forest is well developed. The whole mire complex slopes in a NW–SE direction. This site was intensively drained in the past. The original drainage was probably already made in the nineteenth century but it was then intensified in the 1970s and 1980s using explosives. The peatbog was cut by open ditches arranged in several parallel lines and connected to a peripheral drainage network from the surrounding waterlogged forest. The depth of the ditches varies from 50 to 170 cm with the deepest channels crossing the raised bog. The ditches are mostly spaced at a distance of about 20–30 m.
16.2.2.4 Nad Rybárnou Mire (R) This mire is another example of a drained mire complex located in the northern part of the central plateau at an altitude of about 1,020 m. The body of the original ombrotrophic peatbog (0.3 ha) is almost completely covered by tree and shrub vegetation (Betula pubescens, Pinus x pseudopumilio), probably due to the intensive drainage. Only small fragments of open dwarf shrub vegetation dominated by Vaccinium uliginosum remain to be fragmentarily preserved. The waterlogged spruce forests and spruce mires surrounding the present peatbog seem to be partly developed on the original bog dome. This mire complex was also drained in the past with similar origins of drainage to that of the previous mire (SCH). The ditches are of considerable size (up to 3 m wide and 2 m deep) and artificially drain both the mire complex and a spring area situated above the mire.
16.3 Methods 16.3.1 Monitoring Programme The four mire complexes with different levels of disturbance, including control (sites B, SR), drained (SCH) and heavily-drained (R) sites, have been studied since 2005 (see Table 16.1). The mire habitats under the programme of monitoring are both ombrotrophic raised bogs and waterlogged or mire spruce forests. Monitoring of the drained sites started 3 years prior to restoration and continued in the postrestoration phase. Permanent plots (71), with associated water wells (boreholes), were established to characterise the microtopography, vegetation and drainage patterns of the different mire sites. Microtopographical and vegetation types (microsites) studied included
336
I. Bufková et al.
hollows (with vegetation of the All. Leuco-Scheuchzerion), lawns (All. Oxycocco-Ericion dominated by Trichophorum caespitosum), hummocks (All. Sphagnion medii dominated by Vaccinium uliginosum), waterlogged spruce forest (Ass. Bazzanio trilobatae – Piceetum) and spruce mire (Ass. Sphagno-Piceetum). The height of the water table was measured manually in all installed plastic boreholes (71) at nearly fortnight intervals during the frost-free season (April – November) in the years 2005–2009. Automatic gauging (at 1 h intervals) of the water table by piezometers was than used in a selected 49 boreholes. Water samples from the boreholes, ditches, runoff (outflow) profiles and stream were taken monthly for detailed hydrochemical analysis, including content of main cations and anions (SO4, NO3, NH4, PO4, Ca, Mg, Al, total Fe), pH, conductivity and DOC (dissolved organic carbon). Conductivity and pH values were corrected according to Sjörs (1950). Runoff profiles were monitored only at drained sites. The stream profile sampled for hydrochemistry was situated in a mountain stream (Roklansky potok) that naturally drains the whole catchment area (ca. 45 km2) encompassing the monitored sites. The runoff from drained subcatchments as well as amounts of precipitation were measured continually. Vegetation was sampled in 1 × 1 m permanent plots around each borehole. The percentage cover values for all vascular plants and bryophytes present on the permanent plots were estimated visually.
16.3.2 Restoration Measures The two drained mire-complexes studied were restored in 2008 within the framework of the Mire Restoration Programme. The main restoration technique used was the blocking of drainage ditches by a series of small wooden dams. The aim of the damming was to raise the water table to a more natural (pre-drainage) level, reduce its fluctuation, and retain enough water in the mires for improved natural functioning but especially for the most critical dry periods. The environmental changes induced by the blocking of ditches were expected to stop the degradation processes and conversely to enhance the functioning mire ecology and peat-forming vegetation. The restoration was undertaken from the end of July to August 2008 (site R) and from the end of September to October 2008 (site SCH). It was based on the target water table (TWT) concept which is used generally in all mire restoration projects implemented in the Šumava NP. TWT in this approach means the water table to be reached in mire habitats by restoration and corresponds with a natural water table characteristic for undisturbed mires of a distinct type. As a good general rule, it is important to construct dams frequently enough to make it possible for the water level to reach its original level (TWT) over areas included between ditches (Sallantaus et al. 2003). This rule is especially important in steeply-inclined areas, as are the sloping mountain mires on the central plateau of the Šumava Mts. The TWT according to this concept represents a key parameter for judging the number of dams and their distribution along the ditch. For the purpose of project
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
337
design, the TWT could be expressed as the minimum tolerable water level immediately downstream of the dam (under spillway). The correct spacing of dams along the ditch was than judged from the TWT values and surface gradient (Fig. 16.2). TWT for individual ditch segments was determined by the type of mire (mire vegetation) being crossed by the ditch. TWTs for distinct mire types reported from the Šumava Mts. are shown in Table 16.2. These values were set up based on results of local mire monitoring (Bufková I., 2005) and data reported from comparable mires (Dierssen and Dierssen 2001; Neuhäusl 1972; Neuhäusl 1975; Rybníček 1974; Rybníček et al. 1984). Various wooden dams sometimes combined with geotextiles were used to block the ditches. The use of peat dams, commonly used in northern countries, was
Fig. 16.2 Scheme of dam insertion along the ditch. TWT (target water table) and surface gradient (adjacent angle) as key variables for determination of correct spacing of dams Table 16.2 Target water table for distinct mire types restored in the Šumava Mts Mire type TWT (cm under surface) 5–10 Ombrotrophic raised bog – open bog expanse 20–30 – shrubby bog margin 10–15 – lag Sphagnum spruce mire 15–20 Waterlogged spruce forests (different types): 20–40 5–15 Treeless minerotrophic mires: 15–30 – Sphagno recurvi-Caricion canescentis – Caricion fuscae
338
I. Bufková et al.
impossible here due to insufficient amounts of peat at the restored sites (due to the dwarf character of central European mountain mires with areas mostly only up to 30 ha). After damming, the terrestrialisation of newly-developed water bodies between dams was supported by the placement of various natural materials (fascines, branches, stems, bunches of Sphagnum) into the water segments, namely in the deep ditches. The spontaneous terrestrialisation of shallow ditches, especially under good light conditions, usually proceeds very well without any support. Because of the high vulnerability of restored habitats, all work was done manually without the use of heavy machines. The experience gained from the restoration of numerous European mires (Jehl 1994; Rodwell 1988; Stoneman and Brooks 1997; Bragg 2003) was used in setting up the TWT concept and in deciding on the restoration techniques to be applied in the Mire Restoration Programme.
16.4 Results 16.4.1 Hydrology The mean water table from the manual measurements for distinct mire types or microsites (microtopographical-vegetation units) during the drained, pre-restoration condition (2005–2007), is shown in Fig. 16.3. The automatically-gauged height of the water table and its fluctuation, together with amounts of precipitation, during the pre-restoration phase (2005–2007) are shown in Fig. 16.4. Three years of prerestoration monitoring had shown that the water table (WT) was generally maintained at a lower position than it would be for a natural system and exhibited higher fluctuations in direct relation to the amount of precipitation received on the drained sites. Furthermore, differences were found between distinct microsites of mire habitats. The water table was maintained at a lower height beneath the ‘drier’ dwarfshrub sites (Vaccinium-uliginosum-dominated) that prevailed on the drained bogs. The water table beneath these sites ranged mostly between 10 and 30 cm below the mire surface, with regular drops down to a level of 40 or even 50 cm below the surface during dry periods. On intact sites, the water table beneath comparable (hummock) microsites was maintained mostly 10 cm higher and exhibited much lower fluctuation. As can be seen in Fig. 16.4, the water table on intact sites also reflected the lack of water input from precipitation at certain times, but exhibited much better buffering of small water deficits (both the depth and rate of ‘dips’ in level are different). The highest water table and most stable water regime were found in the hollows with Carex limosa on intact bogs. However, these microsites had practically disappeared in the drained bogs due to the degradation changes in surface structure and vegetation. Fragments of Trichophorum lawns in drained bogs generally exhibited a more stable water regime than hummock microsites, with a mean water table similar to
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
339
10 0 −10 −20 Water table (cm under surfae)
−30 −40 −50 −60 −70 −80 −90 −100 −110
sph forest drain
wet forest drain
vacc drain
trich drain
vacc int
trich int
−130
hollow int
−120
Fig. 16.3 Mean water table (manual measurement) for distinct microsites of studied intact and drained sites (B, SCH, R) during the years 2005–2007 (pre-restoration condition). Explanations: int – intact mire (B), drain – drained mire (SCH, R), hollow – hollows with Carex limosa, trich – Trichophorum caespitosum lawns, vacc – Vaccinium dominated dwarf shrubs, wet forest – waterlogged spruce forest, sph forest – spruce mire
Fig. 16.4 Automatically-gauged height of water table and precipitation on intact site (B) and drained site (SCH) during years 2005–2007 (pre-restoration phase). Explanations: hollow int – hollows with Carex limosa (site B), trich int – Trichophorum caespitosum lawns (site B), vacc – Vaccinium dominated dwarf shrubs (site SCH)
340
I. Bufková et al.
Wet forest drain-rest
Wet forest drain
vacc drain H-rest
vacc drain H
vacc drain-rest
vacc drain
Schachtenfilz
trich drain-rest
10 0 −10 −20 −30 −40 −50 −60 −70 −80 −90 −100 −110 −120 −130
trich drain
Water table (cm under surface)
that of the intact bogs (Fig. 16.3). On the other hand, lawn microsites were much reduced and related only to small, hydrologically-more stable, core bog sites (site SCH), or were even completely missing (site R) on the drained sites. The lowest water table with the highest amplitude of fluctuation was recorded from the drained waterlogged spruce forests surrounding the studied mires (Fig. 16.3). A comparison of mean water tables measured manually during the pre-restoration (2005–2007) and post-restoration phases (2009) on both the restored mires (SCH, R) is shown in Figs. 16.5 and 16.6. Restoration measures seemed to be more effective in the Schachtenfilz mire (Fig. 16.5), with a higher effect in terms of raising the water table and stabilisation beneath the waterlogged spruce forest and Vaccinium dwarf shrub vegetation on the bog. The mean water table beneath the relatively-stable lawns on the drained bogs remained at almost the same level, but the amplitude of its fluctuation became lower. In contrast, the response of the heavily-drained and highly-forested Nad Rybárnou mire was different (Fig. 16.6). Increases in the water table beneath the waterlogged spruce forest were less evident during the first year after restoration.
Fig. 16.5 Mean water table (manual measurement) for distinct microsites on drained site SCH during the pre-restoration condition (2005–2007), and post-restoration phase (2009). Explanations: drain – moderately drained microsite, drain H – heavily drained microsite, drain-rest – moderately drained microsite after restoration, drain H-rest – heavily drained microsite after restoration, trich – Trichophorum caespitosum lawns, vacc – Vaccinium-dominated dwarf shrubs, wet forest – waterlogged spruce forest
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
341
Sph forest drain-rest
Sph forest drain
Wet forest drain-rest
wet forest drain
vacc drain H-rest
vacc drain H
Water Table (cm under surface)
Nad Rybarnou 10 0 −10 −20 −30 −40 −50 −60 −70 −80 −90 −100 −110 −120 −130
Fig. 16.6 Mean water table (manual measurement) for distinct microsites on drained site R during pre-restoration condition (2005–2007) and post-restoration phase (2009). For explanation of abbreviations (see Fig. 16.5)
The water table beneath hummock Vaccinium vegetation in bog fragments, as well as beneath spruce mire forest, even exhibited deeper ‘dips’ in level, though less occasionally, in comparison with the pre-restoration phase. The automaticallygauged height of the water table and its fluctuations, together with amount of precipitation, during restoration (2008) and the first year after restoration (2009) at the site Schachtenfilz (SCH) are shown in Fig. 16.7.
16.4.2 Vegetation–Environment Relationships As restored mires are in their very early stages of post-restoration succession (only 1 year after restoration), only a comparison of both disturbed and intact sites based on pre-restoration data is presented in order to demonstrate the impacts of drainage on vegetation. The indirect gradient analysis (DCA) of vegetation data from the pre-restoration phase (until 2008), with passively inset environmental variables, is shown in Fig. 16.8. DCA apparently arranged bog species from permanent plots along the first ordination axis, which can be interpreted as a moisture gradient and explained 29.6% of variability in the vegetation data. The second ordination axis
342
I. Bufková et al.
Fig. 16.7 Automatically-gauged height of the water table and precipitation during the year of restoration (2008) and post-restoration phase (2009) in drained Schachtenfilz mire (SCH). For explanation of abbreviations (see Fig. 16.3)
explained 31.2% of variability in the vegetation data and appeared to reflect gradient of nutrients. Species from the open and wettest parts of undisturbed peatbogs like Scheuchzeria palustris, Carex limosa, Trichophorum caespitosum or Sphagnum cuspidatum and Warnstorfia fluitans are seen in the left part of the graph. Higher mean WT and minimum WT are their typical features together with a higher pH of groundwater. Vegetation dominated by these species is reduced or even missing in drained bogs. In contrast, species of drier dwarf-shrub vegetation are placed rather to the right in the graph. They are related to sites characterised by a lower pH and higher WT fluctuations, higher conductivity, DOC and concentration of almost all cations and anions in the groundwater. Species typical for disturbed sites (at some places even tree species appeared), are concentrated rather in the upper right of the graph. They include, for example, Vaccinium myrtillus, Polytrichum strictum or Melampyrum pratense. Higher concentrations of Al, DOC, PO4 or base cations (Ca, Mg) could characterise these sites. Mosses tolerant to drier conditions like Dicranum scoparium or Pleurozium schreberi are placed to the lower right. These species frequently occur in disturbed hummock vegetation dominated by dwarf Vaccinium shrubs that are generally expanding in drained bogs, somewhat also at the expense of Trichophorum lawns and hollow vegetation.
16.4.3 Changes in Hydrochemistry Changes in chemical composition of both groundwater (GW) and surface water at drained sites during the period between 2007 (pre-restoration) and 2009 (post- restoration) are shown in Figs. 16.9–16.12. The absence of data for runoff surface water from the SCH site in 2008 was due to the lack of water in the channel profile in that year.
343
5
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
PiceAbE2 SphaMage SphaFala SphaFlex
SphaCusp
GymnlnfI
SchePalu CareLimo
WansFlui
pH Wtmin Wtmean OxycPalu TrichCae
PolyStri VaccMyrt Ca Mg AI MelaPrat SphaGirg DOC PO4 Kond PinuPsE2 CI SphaRuss PicAbiJu IntBog NH4 PtilCili DrBog Fe WTflukt Erio Vagl NO3 Wtmax VaccUling DicrScop AndrPoli EmpeNigr
SO4
SphaCapi PleuSchr
−2
MyliAnom
−2
8
Fig. 16.8 Indirect gradient analysis (DCA) of species recorded in permanent plots around each borehole in monitored bog sites. Explanations: WT mean-mean water table, Wtmin-minimum water table, Wtmax-maximum water table, Wtflukt-fluctuations of water table, Kond-conductivity, IntBog-Intact Bog, DrBog-drained bog, AndrPoli-Andromeda polifolia, CareLimo-Carex limosa, DicrScop-Dicranum scoparium, EmpeNigr-Empetrum nigrum, ErioVagi-Eriophorum vaginatum, GymnInfl-Gymnocolea inflata, MelaPrat-Melampyrum pratense, MyliAnom-Mylia anomala, OxycPalu-Oxycoccus palustris, PicAbiJu-Picea abies juvenile, PiceAbE2-Picea abies in shrub layer, PinuPsE2-Pinus x pseudopumilio in shrub layer, PleurSchr-Pleurozium schreberi, PolyStriPolytrichum strictum, PtilCili-Ptilidium ciliare, SchePalu-Scheuchzeria palustris, SphaCapiSphagnum capillifolium, SphaCusp-Sphagnum cuspidatum, SphaFala-Sphagnum fallax, SphaFlex-Sphagnum flexuosum, SphaGirg-Sphagnum girgensohnii, SphaMage-Sphagnum magellanicum, TrichCae-Trichophorum caespitosum, VaccMyrt-Vaccinium myrtillus, VaccUligVaccinium uliginosum, WansFlui-Warnstorfia fluitans
Both conductivity and pH of the bog GW and ditch surface water were lower in 2007 than those in the next 2 years (Fig. 16.9). Higher conductivity and lower pH of GW were found beneath Vaccinium shrubs in the highly-disturbed bog R compared to the less-disturbed bog SCH during all 3 years. Conductivity of both bog GW and ditch water seemed to increase with the start of restoration work in 2008 in both the SCH and R sites. In 2009, conductivity was lower again, especially in the bog ditches. The most significant change in pH values was recorded for runoff water from the heavily-disturbed site R after restoration where pH of surface water fell from a level of about 5.5 to values about 4.5 (see Fig. 16.9f). Seasonal patterns in DOC concentration were quite similar at all sites for all three frost-free seasons monitored (Fig. 16.10). It usually showed a lower spring concentration followed by a general rise in summer and decline in late autumn
344
I. Bufková et al.
Fig. 16.9 Conductivity and pH of groundwater and surface water at drained sites (SCH, R) during the years 2007–2009. Explanations: S-Schachtenfilz site, R-Nad Rybárnou site
(Fig. 16.10). In 2007, DOC was generally higher in bog GW and ditch surface water compared to the two following years, a difference that was especially well pronounced at site R. DOC in bog GW was generally higher at the heavily-disturbed site (R) than in the less-disturbed SCH. Sulphate concentrations were generally much lower in bog GW and bog ditches in comparison to both forest ditches and runoff water. Almost no interannual differences were found in the variation of sulphates in the bog habitats (GW and ditches). On the other hand, sulphates appeared to decrease in the forest ditch after restoration. Sulphates also decreased after restoration in the runoff water from the heavily-disturbed and forested site R but DOC was slightly increased here. Opposite trends in the concentrations of both components were observed in the runoff water from the SCH site (Fig. 16.10 e–f).
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
345
Fig. 16.10 Sulphate and dissolved organic carbon (DOC) concentrations in groundwater and surface water at drained sites (SCH, R) during the years 2007–2009. Explanations: see Fig. 16.9
Phosphate concentrations were mostly very low with low seasonal variation (Fig. 16.11), especially in bog habitats, where values were usually less than 0.1 mg.ml−1. However, a well-pronounced rise in phosphate concentrations was found in the surface water of forest ditches after restoration. Only a slight increase in phosphate concentrations can be seen in runoff water after restoration, especially from site R. Concentrations of ammonium ions were found to be highest in the bog GW of the less-disturbed SCH site and slightly higher in forest ditches. No changes in ammonium concentrations have been observed after restoration at both sites. Aluminium concentrations were generally very low both in GW and ditch water in bog habitats with the exception of one extreme increase in the autumn of 2008
346
I. Bufková et al.
Fig. 16.11 Phosphate and ammonium concentrations in groundwater and surface water at drained sites (SCH, R) during the years 2007–2009. Explanations: see Fig. 16.9
(Fig. 16.12). Higher Al values were recorded from both forest ditch and runoff water. After restoration, Al concentrations clearly increased in runoff water from both sites (SCH and R). The highest Fe concentrations before restoration were in bog GW of the heavily-disturbed site R. A pronounced increase in Fe concentration was observed both in forest ditches and runoff water from both sites (R and SCH) after restoration. For comparative purposes, hydrochemistry including conductivity, pH and concentrations of DOC, sulphates, phosphates, ammonium ions, Al and Fe in the ‘main stream’ during the years 2007–2009 are shown in Fig. 16.13. The main stream naturally drains the whole catchment area within which the studied sites are part.
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
347
Fig. 16.12 Aluminium and total Fe concentrations in groundwater and surface water at drained sites (SCH, R) during the years 2007–2009. Explanations: see Fig. 16.9
16.5 Discussion and Conclusions 16.5.1 Restoration Measures In recent years, the restoration of disturbed mires has received considerable attention(Wheeler et al. 1995; Malterer et al. 1998) having experienced significant advances in the 1990s (Lode 2001; Price et al. 2003). Great efforts have been
348
I. Bufková et al.
Fig. 16.13 Conductivity, pH and concentrations of dissolved organic carbon (DOC) and main anions and cations in stream water (Roklanský potok stream) during the years 2007–2009
mainly focused on the restoration of cut-over peatlands (Lode 1999; Rochefort and Lode 2006); even the peat industry has become concerned with the problem and supported current research (Hood 2000). The restoration of mires drained for forestry or agriculture has also become increasingly important, mostly during the last two decades. Many projects aimed to restore natural development and biodiversity in drained mires have been undertaken in northern Europe (e.g. Lindsay 1995; Stoneman and Brooks 1997; Sallantaus et al. 2003; Kuuluvainen et al. 2002; Vasander et al. 2003; Ruseckas and Grigaliünas 2008), as well as central Europe (e.g. Pfadenhauer and Grootjans 1999; Joosten 2000), where most of the mires have been lost or degraded due to long-term human impacts. The majority of restoration projects in drained peatlands, including those implemented in the Šumava National Park, reflect the ‘biodiversity strategy’ sensu Joosten (2000) which primarily consists of realising a reasonable approximation to a peatland’s original condition. Either way, the restoration process in mires should ensure the return of some order of functionality of the ecosystem necessary for its self-perpetuity (Lode 2001). The main goals of restoration projects in mires, as stated for example by the U.S. National Research Council (NRC 1992), are: “to emulate a natural, functioning, self-regulating system that is integrated with the ecological landscape in which it occurs”. This will encompass a return to the ecosystem of its structure, function, trophic organisation, and biodiversity characteristic for its type (Rochefort 2000). Restoration projects implemented in the Šumava National Park are a reflection of
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
349
these efforts. Restoration measures are considered here only as time-limited activities focused on the re-establishment of natural or ‘new natural’ conditions (sensu Vasander et al. 2003), and in particular, including a more or less normal hydrology. Subsequent autogenic plant succession and the self-regulating development of a distinct mire type are then expected after restoration. The long-term monitoring programme that started in 2005, was designed to demonstrate if this presumption would be correct. As the key element in restoring peatlands is water, rewetting is obviously the fundamental prerequisite for restoration (Gorham and Rochefort 2003). The first step in most restoration projects is therefore to raise the water table and ideally stabilise it close to the peat surface (Laine et al. 2006). Dam construction or the filling-in of drains are common techniques used to achieve this. Around the world, dams of various type and construction (e.g. plastic, woody or peaty dams) have been used to restore hydrological functions in drained mires (Rodwell 1988; Stoneman and Brooks 1997; Labadz et al. 2002). Compared to other areas, the restoration of the relatively slightly-sloping mires in the Šumava National Park was quite specific. It was impossible to use peat dams due to the lack of sufficient peat available for the restoration of valuable sites, the area of which usually do not exceed 30 ha. Plastic dams could not be utilised either because of the frequent presence of woody layers in the peat profile. As a result, only wooden dams have been used to block drainage ditches in this instance. In addition to the piled-board dams hammered vertically in the large bog channels (Fig. 16.14), horizontal board dams combined with inert geotextiles (Fig. 16.15) were also used. Results from the Šumava Mts. showed that, despite the insertion of horizontal-wooden dams being generally viewed as less successful (Rodwell 1988), it can work if carefully constructed and combined with geotextiles and a partial infilling of blocked ditch segments. Horizontal dams were usually installed at sites where other damming methods had already failed: for example, in wet forests or other sites with only shallow layers of peat. The use of a cascade of dams is generally recommended for a sloping surface (Rodwell 1988; Stoneman and Brooks 1997; Sallantaus et al. 2003). In the Šumava Mts., where frequent damming was necessary, the target water table concept was used to judge the spacing of dams (see Section 16.3). Because of the steep slopes, the distance between dams was often less than 10 m. All restoration measures, including the transport of materials, had to be made manually due to the little or no access to restored sites and their high vulnerability. However, the combination of professional craftmen (dam insertion) and volunteers (e.g. ditch infilling, transport of materials) worked very effectively and the final costs per dam usually ranged between 40 and 90 EUR only.
16.5.2 Monitoring Although many mires have been restored both in Europe and America and experiences are quite promising, many open questions still remain (Sallantaus
350
I. Bufková et al.
Fig. 16.14 Cascade of piled-board dams installed according to target water concept in a large bog ditch within the Schachtenfilz (SCH) site; (a) insertion phase, (b) 3 weeks after damming (just before partial infilling of the ditch by natural materials). TWT for that ditch segment was 10 cm
et al. 2003; Laine et al. 2006). Practical restoration projects should therefore be closely linked with monitoring and research whenever possible (Bakker et al. 2000; Labadz et al. 2002; Gorham and Rochefort 2003; Sallantaus et al. 2003). The coupling of monitoring and restoration within the programme in the Šumava NP tries to reflect those needs. The first results from the early post-restoration phase suggest that damming has had a positive effect on the hydrology in the moderately-disturbed mire (SCH). The mean water table increased and fluctuations were reduced, especially in the dwarf-shrub bog ‘compartments’ and wet forests (Fig. 16.5). Elevated water tables and a
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
351
Fig. 16.15 An example of horizontal woody dams constructed from two board layers with geotextile in between. (a) insertion phase, (b) 1 year after damming, (c) partial infilling of water bodies between dams by natural materials to enhance terrestrialisation processes
s tabilised hydrology after the damming of ditches have been recorded from many restored sites (for example by, Jauhiainen et al. 2002; Vasander et al. 2003; Holden et al. 2004; Iqbal et al. 2005; Worrall et al. 2007). However, the hydrological response of other restored mires has been found to be more complicated (Price 1997; Ruseckas and Grigaliünas 2008). In our study, the restoration of hydrological functions was found to be less effective in the heavily-disturbed and largely- forested mire-complex (R). The reason for this different response to damming still remains unknown. In some mires, high peat subsidence after drainage has been reported, complicating the rewetting of drained sites (Sallantaus et al. 2003).
352
I. Bufková et al.
Fluctuations of the water table might also be expected to increase due to the increased bulk density of the surface peat after drainage (Minkkinen and Laine 1998) and the re-establishment of original water conditions could be somewhat difficult. In addition to that, the possible negative influence of subsurface soil pipes on rewetted mires has been discussed (Holden et al. 2004). Subsequent long-term extended monitoring will probably help to identify the reason for a less-successful restoration in the case of the heavily-disturbed site R. Regarding hydrochemical changes, distinctly higher phosphate concentrations in the forest ditches and slightly increased phosphate concentrations in surface runoff water were found in both restored sites. Increased phosphorus leaching from restored sites has been recorded by many authors (e.g. Kalbitz et al. 1999; Sallantaus et al. 2003; Vasander et al. 2003; Rupp et al. 2004; Tiemeyer et al. 2005). However, this phenomenon seems to be relatively limited in time as a decrease in phosphorus concentrations a few years after restoration has been demonstrated by several studies (Vasander et al. 2003). In the Šumava Mts., there has been recorded an increase of phosphates in the groundwater of wet forests (Bufková, unpublished data) first year after the restoration, although their concentrations in bog groundwater and bog ditches remained unchanged (Fig. 16.11). These results suggest that only restored wet forests, and not ombrotrophic bogs, contribute to phosphate leaching from restored sites in this study area. In the heavily-disturbed site R, sulphate concentrations were found to be lower in the forest ditches and runoff water after restoration, whereas DOC values increased in runoff water from the same site. This corresponds with other studies showing that drain-blocking is not a very effective technique for reducing DOC in the short term (Worrall et al. 2007). Among other things, it could be related with the reduction of sulphate suppression in the soil solution after rewetting, as has been demonstrated by Evans et al. (2005) or Clark et al. (2005). Nevertheless, DOC was found significantly reduced around 3 or 4 years after drain blocking, as reported by Wallage et al. (2006). Therefore, the short-term response of a mire to drain-blocking may not be the same as its long-term response (Worrall et al. 2007). As a result, only long-term monitoring will be necessary for a full understanding of the ecological processes and changes promoted by restoration. Acknowledgements The study was supported by a grant of the Czech Ministry of Environment no. SP2d1/113/07. We thank Kamila Lencová for computation of multidimensional analysis, and the Šumava National Park and Protected Landscape Area Administration for their support.
References Bakker JP, Grootjans AP, Hermy M, Poschlod P (2000) How to define targets for ecological restoration? Appl Veg Sci 3:3–6 Bragg O (ed) (2003) Sharing expertise for the conservation of peatlands in central and eastern Europe. University of Dundee, UK, pp 1–200 Bragg O, Lindsay R (eds) (2003) Strategy and action plan for mire and peatland conservation in Central Europe. Wetlands International Publication 18, Wageningen, 93 pp
16 Restoration of Drained Mires in the Šumava National Park, Czech Republic
353
Bufková I, Prach K, Bastl M (2006) Linking vegetation pattern to hydrology and hydrochemistry in a montane river floodplain, the Šumava National Park, Central Europe. Wetlands Ecol and Manage 14:317–327 Clark JM, Chapman PJ, Adamson JK, Lane SN (2005) Influence of drought-induced acidification on the mobility of dissolved organic carbon in peat soils. Global Change Biology 11: 791–809 Dierssen K, Dierssen B (2001) Moore (Ökosysteme Mitteleuropas aus geobotanischer Sicht). Ulmer, Stuttgart (Hohenheim), 230 pp Evans CD, Montieth DT, Cooper DM (2005) Long-term increases in surface water dissolved organic carbon: observations, possible causes and environmental impacts. Environmental Pollution 137: 55-71 Gorham E, Rochefort L (2003) Peatland restoration: a brief assessment with special reference to Sphagnum bogs. Wetlands Ecol Manage 11:109–119 Holden J, Chapman PJ, Labadz JC (2004) Artificial drainage of peatlands: hydrological and hydrochemical process and wetland restoration. Prog Phys Geogr 28(1):95–123 Hood G (2000) Canadian peat producers adopt new policy. Peatlands Int 1:1–2 Iqbal R, Hotes S, Tachibana H (2005) Water quality restoration after damming and its relevance to vegetation succession in a degraded mire. J Environ Syst Eng JSCE 790/VII – 35:59–69 Jauhiainen S, Laiho R, Vasander H (2002) Ecohydrological and vegetational changes in a restored bog and fen. Ann Bot Fennici 39:185–199 Jehl H (1994) Ein Moor im (Gesinnungs-) Wandel der Zeit. Wiedervernässung im Groben Filz bei Riedlhütte (Nationalpark Bayerischer Wald). Research Report. Nationalparkverwaltung Bayerischer Wald, Grafenau, 48 pp Joosten H (1997) European mires: a preliminary status report. Int Mire Conserv Group Members Newsl 3:10–13 Joosten H (2000) Peatland conservation in central and southern Europe. In: Rochefort L, Daigle JY (eds) Sustaining our peatlands. Proceedings of the 11th International Peat Congress, vol 2, Quebec City, Canada, 6–12 August 2000, pp 1044–1049 Joosten H, Clarke D (2002) Wise use of mires and peatlands. Background and principles including a framework for decision-making. International Mire Conservation Group and International Peat Society, Saarijärvi, Finland, 303 pp Kalbitz K, Rupp H, Meibner R, Braumann F (1999) Folgewirkungen der Renaturierung eines Niedermoores auf die Stickstoff-, Phosphor- und Kohlenstoffgehalte im Boden- und Grundwasser. Zeitsch Kulturtech Landetwick 40:22–28 Kuuluvainen T, Aapala K, Ahlroth P, Kuusinen M, Lindholm T, Sallantaus T, Siitonen J, Tukia H (2002) Principles of Ecological Restoration of Boreal Forested Ecosystems: Finland as an Example. Silva Fennica 36(1):409–422 Labadz JC, Butcher DP, Sinnott D (2002) Wetlands and still waters. In: Perrow MR, Davy AJ. Handbook of Ecological Restoration, Vol 1, Principles of Restoration. Cambridge University Press, 106–132 Laine J, Laiho R, Minkkinen K, Vasander H (2006) Forestry and boreal peatlands. In: Wieder RK, Vitt DH (eds) Boreal Peatland Ecosystems. Springer-Verlag, Berlin/Heidelberg/New York, 435 pp Lindsay R (1995) Bogs: the ecology, classification and conservation of ombrotrophic mires. Scottish Natural Heritage, Battleby, 119 pp Lode E (1999) Wetland restoration: a survey of options for restoring peatlands. Studia Forestalia Suecica, Swedish University of Agricultural Sciences, Uppsala, Sweden, vol 205, 30 pp Lode E (2001) Natural mire hydrology in restoration of peatland functions. Doctoral thesis, Acta Universitatis Agrivuturae Sueciae, Silvestria 234, Uppsala Malterer T, Johnson K, Steward J (eds) (1998) Peatland Restoration and Reclamation. Proceedings of the International Peat Symposium, Duluth, Minnesota, USA Minkkinen K, Laine J (1998) Long-term effect of forest drainage on the peat carbon stores of pine mires in Finland. Can J For Res 28:1267–1275 National Research Council (1992) Restoration of aquatic ecosystems. National Academy Press, Washington, DC
354
I. Bufková et al.
Neuhäusl R (1972) Subkontinentale Hochmoore und ihre Vegetation. Studie ČSAV, Praha 13:1–144 Neuhäusl R (1975) Hochmoore am Teich Velké Dářko. Vegetace ČSSR. A9, Academia, Praha, 412 pp Pfadenhauer J, Grootjans A (1999) Wetland restoration in central Europe: aims and methods. Appl Veg Sci 2:95–106 Price J (1997) Soil moisture, water tension, and water table relationships in a managed cutover bog. J Hydrol 202:21–32 Price JS, Heathwaite AL, Baird AJ (2003) Hydrological processes in abandoned and restored peatlands: an overview of management approaches. Wetlands Ecol Manage 11:65–83 Rochefort L (2000) Sphagnum – a keystone genus in habitat restoration. Bryologist 103:503–508 Rochefort L, Lode E (2006) Restoration of degraded Boreal peatlands, 380–423. In: Wieder RK, Vitt DH, Boreal peatland ecosystems. Ecol Stud 188:435 pp Rodwell TA (1988) The peatland management handbook. Research and Survey in Nature Conservation, No. 14 (Peterborough), Nature Conservancy Council, 110 pp Rupp H, Meissner R, Leinweber P (2004) Effects of intensive land use and re-wetting on diffuse phosphorus pollution in fen areas – results from a case study in the Drömling catchment. Germany. J Plant Nutr Soil Sci 167:408–416 Ruseckas J, Grigaliünas V (2008) Effect of drain-blocking and meteorological factors on groundwater table fluctuations in Kamanos mire. J Environ Eng Landscape Manage 16(4):168–177 Rybníček K (1974) Die Vegetation der Moore im Südlichen Teil der Böhmisch-Mährischen Höhe. Vegetace ČSSR A6, Academia, Praha, 243 pp Rybníček K, Balátová-Tuláčková E, Neuhäusl R (1984) Přehled rostlinných společenstev rašelinišť a mokřadních luk Československa. Academia, Praha, 123 pp (in Czech) Sallantaus T, Kondelin H, Heikkilä R (2003) Hydrological problems associated with mire restoration. In: Haikkilä R, Lindholm T (eds) Biodiversity and conservation of boreal nature. Finn Environ 485:256–261 Schreiber H (1924) Moore des Böhmerwaldes und des deutschen Südböhmen. IV. Sebastianberg, 119 pp Sjörs H (1950) On the relation between vegetation and electrolytes in north Swedish mire waters. Oikos 2(2):241–258 Soukupová L (1996) Developmental diversity of peatlands in Bohemian Forest. Silva Gabreta 1:99–107 Stoneman R, Brooks S (eds) (1997) Conserving bogs. The management handbook. Edinburgh, 286 pp Svobodová H, Soukupová L, Reille M (2002) Diversified development of mountain mires, Bohemian Forest, Central Europe, in the last 13 000 years. Quater Int 91:123–135 Tiemeyer B, Lennartz B, Schlichting A, Vegelin K (2005) Risk assessment of the phophorus export from a re-wetted peatland. Phys Chem Earth 30:550–560 Vasander H, Tuittila E-S, Lode E, Lundin L, Illomets M, Sallantaus T, Heikkilä R, Pitkänen M-L, Laine J (2003) Status and restoration of peatlands in northern Europe. Wetlands Ecol Manage 11:51–63 Wallage ZE, Holden J, Mc Donald AT (2006) Drain blocking: an effective treatment for reducing dissolved organic carbon loss and water discolouration in a drained peatland. Sci Total Environ 367:811–821 Wheeler GD, Shaw SC, Fojt WJ, Robertson RA (1995) Restoration of temperate wetlands. Wiley, Chichester Worrall F, Armstrong A, Holden J (2007) Short-term impact of peat drain-blocking on water colour, dissolved organic carbon concentration, and water table depth. J Hydrol 337:315–325
Chapter 17
Local and Global Impacts of Mire Drainage: An Impetus for Hydrology Restoration: Yelnia Mire, Belarus Alexander Kozulin, Sergey Zuyonok, and Viacheslav Rakovich
Abstract The quantity of annually-transferred carbon from biogenic to geological recycling depends on the conditions of the water regime, plants’ mineral nutrition and the duration of the biologically-active temperature period. In Belarus, one hectare of virgin peatland transfers from the atmosphere through biogenic recycling to its geological sink about 150–500 kg of carbon per year, some 550–1,800 kg CO2. One hectare of drained but unused peatland emits 21–23 t (fen mire) or 14–16 t (raised bog) of CO2 annually until the residual peat layer is gone. When mires are drained, they stop performing their natural biospherical functions immediately, resulting in a destabilisation of biospherical processes and cycles. Mire ecosystems remain degraded, with their flora and fauna being destroyed. One of the methods to solve this problem is secondary swamping of drained peatlands in order to restore their natural functionality. A major difficulty of the Yelnia restoration project was to propose a suitable, cheap and reliable dam design that would involve people rather than machines – heavy devices unable to make their way through the mire. Equally important was to calculate the hydrological parameters to avoid flooding the forests surrounding the mire. Having analysed the applicability of different types of water-retaining constructions for the local conditions, it was considered optimal to build pile-fence wooden dams. On canals accessible with heavy machinery, it was proposed to build stone-filled water-retaining constructions. Keywords Peatlands • Drained mires • Mire restoration • Carbon dioxide A. Kozulin (*) Institute of Zoology of the National Academy of Sciences of Belarus, Akademicheskaya Street 27, 220072 Minsk, Belarus e-mail:
[email protected] S. Zuyonok National Union for Bird Conservation of Belarus, Makaenka Street 8-313, 220023 Minsk, Belarus V. Rakovich Institute for Nature Management of the National Academy of Sciences of Belarus, Scoryny Street 10, 220114 Minsk, Belarus M. Eiseltová (ed.), Restoration of Lakes, Streams, Floodplains, and Bogs in Europe: Principles and Case Studies, Wetlands: Ecology, Conservation and Management 3, DOI 10.1007/978-90-481-9265-6_17, © Springer Science+Business Media B.V. 2010
355
356
A. Kozulin et al.
17.1 Introduction In the early 1960s, peatlands covered 29,390 km2 or 14.2% of Belarus. Large-scale amelioration carried out over the past 40 years has resulted in some 54% of peatlands being drained. Today, around 13,450 km2 of peatlands (covering 6.4% of the country) remain intact (Zemlya Belarusi 2002). Drainage of such a huge area of wetlands has led to a number of negative impacts – soil erosion, degradation of natural ecosystems, decrease of biological diversity, and extensive fires which periodically cover huge areas. The majority of mires were drained without sufficient scientific grounds and it has become more apparent in recent years that these drained areas are being used inefficiently. In many cases the drainage of mire areas is uneconomic since agricultural production on such areas is very low and the growth of wood in sylviculture is slow. Realising the problem, the government undertook an inventory of disturbed mires with the aim of ensuring their use in a suitable manner. The inventory showed that the total area of inefficiently-used drained mires was 453,500 ha (78,000 ha in agriculture, 209,000 ha of exhausted peat fields, 166,000 ha of forest meliorated areas) and would, therefore, be suitable for restoration. Abandoned from active use, such sites face a constant threat of fire as their drainage systems keep functioning. The risk of fires on these drained areas is especially high in dry years when the top layer of peat dries out. Inevitably, drainage of mires also contributes to global climate change as a result of organic matter mineralisation and the release of vast amounts of carbon dioxide to the atmosphere. Mires which in their intact state function as an important sink for CO2 become its source when drained. One of the examples of inefficient forest melioration which damages mire ecosystems is the forest melioration of Yelnia mire – situated in one of the largest mire complexes of Belarusian Poozerie.
17.2 Pilot Project on Hydrology Restoration: Yelnia Mire 17.2.1 Description of the Site The Yelnia mire (55°34’N 27°55’E) in Vitebsk Oblast, with an area of 23,200 ha, belongs to Belarus’ largest complex of bogs and transitional mires, with numerous lakes attaching a string of uniqueness to the monotonous mire landscape. Small islands covered by deciduous and spruce forests are scattered all over the complex. Most of the mire has been overgrown by low pine stands. However, relatively large open spaces with numerous small lakes and open water are also quite common. Vegetation of the wetland is typical for bogs and is represented by pine-shrub-Sphagnum and shrubSphagnum communities. The site has been declared a National Hydrological Reserve in 1968, an Important Bird Area (IBA) in 1998, and a Ramsar site in 2002.
17 Local and Global Impacts of Mire Drainage
357
The Yelnia mire is located on the watershed of two river basins – the Disna and West Dvina. Its central mound is some 7 m above its outer periphery; the maximum depth of peat is 8.3 m, with an average depth of 3.8 m. The mire is fed by precipitation, underground and surface water and drains into three rivers. More than 100 lakes are located within the IBA. Most of the lakes are linked to each other by rivers and anabranching channels. Artificial drainage of part of this mire complex, and vast adjoining areas, has disrupted the hydrological regime of the mire. The construction of numerous canals and ditches (see Fig. 17.1), as well as the canalisation of rivers, led to the lowering of the groundwater table which is one of the causes of the large severe fires that occur on the site almost every year. The difficult access and the specific nature of the landscape (about 60% of forests are swampy and of low-production) has limited local people’s utilisation of the mire. Forestry is carried out mainly along the periphery of the site and on mineral ‘islands’. Except for the commercial and domestic collection of mushrooms and berries, the site has almost no agricultural use. Amateur fishing occurs on the lakes.
Fig. 17.1 A land-use map of Yelnia bog and close surroundings with drainage canals marked
358
A. Kozulin et al.
17.2.2 The Mire’s Biological Value A total of 98 bird species have been recorded on the site, including 23 National Red Data Book species. Yelnia hosts scattered colonies of waders and numerous gull colonies, the latter found mainly on lakes. It is also a breeding ground for a number of typical bog species rare in Belarus, such as black-throated diver (Gavia arctica), willow grouse (Lagopus lagopus), golden plover (Pluvialis apricaria), whimbrel (Numenius phaeopus), jack snipe (Lymnocryptes minimus) and greenshank (Tringa nebularia). Merlin (Falco columbarius) and short-toed eagle (Circaetus gallicus) breed here. Population estimates for the most important bird species are listed in Table 17.1. Yelnia is also an important stop-over ground for migrating geese and cranes in spring and autumn. During these times the following species are commonly very numerous: bean goose (Anser fabalis), white-fronted goose (Anser albifrons), wigeon (Anas penelope), garganey (Anas querquedula). Less common are greylag goose (Anser anser), lesser white-fronted goose (Anser erythropus), pintail (Anas acuta). Of the plants growing on the mire and mineral islands, eleven are listed in the National Red Data Book, including dwarf birch (Betula nana) and cloudberry (Rubus chamaemorus). The site also has 7 amphibian, 5 reptile, and 31 mammal species. Most mammals occur along the periphery of the mire and visit the mire in search for food. Notable is the high abundance of adder (Vipera berus).
17.2.3 Drainage of the Mire and Its Consequences Considerable changes in the mire’s hydrological situation inevitably followed the changes described above, that were mainly effectuated between 1955 and 1965. The network of canals and ditches (Fig. 17.1) accelerated the discharge of water from the mire and significantly lowered the ground water table already before summer; the rivers and lakes were drying out. As a result, woody shrubs encroached into open areas and the species composition of the mire changed. The lowered water table and frequent fires that followed created suitable conditions for associations dominated by 2–3 m tall Pinetum-Betuletum (Fig. 17.2) and Polytrichum strictum forming a continuous carpet (sometimes with Sphagnum mosses). At the peripheral parts of the mire, the disappearance of water-loving birds such as blackthroated diver (Gavia arctica), willow grouse (Lagopus lagopus), golden plover (Pluvialis apricaria), jack snipe (Lymnocryptes minimus), whimbrel (Numenius phaeopus) and greenshank (Tringa nebularia) was observed. Significant changes to the mire ecosystem caused frequent fires to occur at the surface of the mire almost every year and resulted in abundant tree falls (Fig. 17.3). During the 1990s, particularly strong fires took place in 1993, 1994 and 1999; the area affected by the fires being from several to hundreds of hectares. Substantial finances were spent fighting them. The strongest fire, in 1999, made
17 Local and Global Impacts of Mire Drainage
359
it clear that what was needed was to understand the causes of the fires and prevent them, rather than trying to meet the year-on-year increase in expenditure fighting the effects of the fires. Table 17.1 Population estimates of most important bird species of Yelnia Reserve (Biryukov, et al. 1993), ‘–’ means no estimate available Population (pairs) English name Scientific name Status Min Max Black-throated diver Breeding 10 15 Gavia arctica Little greebe Tachibaptus ruficollis Breeding 0 2 Teal Anas crecca Breeding – – Mallard Anas platyrhynchos Breeding 100 150 Tufted duck Aythya fuligula Breeding 10 15 Goldeneye Bucephala clangula Breeding – – Snake eagle Circaetus gallicus Breeding 2 3 Hen harrier Circus cyaneus Breeding – – Golden eagle Aquila chrysaetos Breeding 1 1 Merlin Falco columbarius Breeding 15 20 Hobby Falco subbuteo Breeding – – Willow grouse Lagopus lagopus Breeding 90 120 Black grouse Tetrao tetrix Breeding 150 200 Capercaillie Tetrao urogallus Breeding – – Crane Grus grus Breeding 60 60 Crane Grus grus Breeding 2,000 3,000 Golden plover Pluvialis apricaria Breeding 80 100 Lapwing Vanellus vanellus Breeding – – Jack snipe Lymnocryptes minimus Breeding – – Common snipe Gallinago gallinago Breeding – – Great snipe Gallinago media Breeding 20 – Black tailed godwit Limosa limosa Breeding – – Whimbrel Numenius phaeopus Breeding 20 30 Curlew Numenius arquata Breeding 150 170 Redshank Tringa totanus Breeding – – Greenshank Tringa nebularia Breeding 15 20 Green sandpiper Tringa ochropus Breeding – – Wood sandpiper Tringa glareola Breeding – – Black-headed gull Larus ridibundus Breeding – – Common gull Larus canus Breeding 50 70 Herring gull Larus argentatus Breeding 40 50 Short-eared owl Asio flammeus Breeding 40 50 Meadow pipit Anthus pratensis Breeding – – Yellow wagtail Motacilla f lava Breeding – – White wagtail Motacilla alba Breeding – – Willow warbler Phylloscopus trochilus Breeding – – Crested tit Parus cristatus Breeding – – Great grey shrike Lanius excubitor Breeding 15 20 Hooded crow Corvus corone Breeding – –
360
A. Kozulin et al.
Fig. 17.2 Part of a drainage canal downstream of a dam soon after its construction (water is not yet flowing over the lip of the dam)
Fig. 17.3 Consequences of a fire: trees falling as a result of the burning of the top soil layer along with tree roots
17 Local and Global Impacts of Mire Drainage
361
17.2.4 Restoration In 1999, at a regular meeting of representatives from the Environment Ministry and two NGO BirdLife partners (Akhova Ptushak Belarusi [APB] and the Royal Society for the Protection of Birds [RSPB], UK), the issue of saving the Yelnia IBA was raised. It was deemed necessary to start a project to understand the causes of the fires, and to develop and implement a set of measures aimed at an optimisation of the ecological situation. RSPB took the decision to finance such a pilot project and develop guidelines to save the mire. Experts from different fields were brought together by APB to implement the pilot project. Based on the results of their ecological expertise and data analysis of water-carrying capacity of rivers and ditches draining the mire, those rivers and ditches having the biggest impact on the hydrological regime of Yelnia mire were determined. To optimise the water table over the entire mire, it was recommended to make water-retaining constructions on 21 of them (for small natural streams and canals of small water-carrying capacity, dam construction was not planned). After intensive fundraising, finances were secured with the Belarus Ministry of Natural Resources and Environmental Protection and the Ministry of Foreign Affairs of the Netherlands (DGIS) within the framework of the Global Peatland Initiative (implemented by Wetlands International in cooperation with Dutch IUCN National Committee, International Mire Conservation Group and International Peat Society). The engineering project of mire rehabilitation was developed by specialists from RUP “Belgiprovodkhoz”. Construction work was carried out by local people hired by a local construction company (Yazno town). Disna forestry (land user) provided overall supervision, human resources, and committed itself to maintenance of the constructed facilities beyond the project. A major difficulty of the project was to propose suitable, cheap and reliable dams to be constructed without the use of machinery as the use of machinery in a fragile ecosystem such as the mire was not desirable. Equally important was to calculate the hydrological parameters (such as discharge of water, parameters of the canal, height of dams) to avoid flooding the forests surrounding the mire. Having analysed the applicability of different types of water-retaining constructions for the local conditions, it was considered optimal to build pile-fence wooden dams. Earlier in the nineteenth to twentieth centuries such water regulating devices were widely used for land ‘improvement’ in Belarus. By their design, such dams fully correspond to the range of water levels to be maintained in the ditches and canals. The pressure of flowing water is withstood by the wooden construction, while shift resistance is ensured by securing the wooden piles in their foundations with soil, gravel, etc. Pine was used for the dam construction – pine being the most widelydistributed tree species with wood stable in changing humidity. On canals accessible to heavy machinery, it was proposed to build stone-filled water-retaining constructions. To take into account the considerable down slope of canals, the construction sites for both wooden and stone dams were located as close as possible to the edges of the mire.
362
A. Kozulin et al.
Fig. 17.4 A dam on one of the canals in Yelnia bog
Altogether, 17 drainage ditches were effectively ‘closed’ by dams between December 2001 and June 2002, contributing – almost instantly – to an increase in the groundwater table (Fig. 17.4). The immediate effects were visible already in June 2002. A monitoring field trip showed that the groundwater level on the peripheral parts of the bog was up to 0.3 m above the surface, whereas previously, at the same time of year, the water table used to drop as low as 0.5 m below the surface. Unlike previous years (with virtually annual destructive fires sweeping across the area), the 2002 fire affected only the above-ground vegetation, without going deep into the peat layer. To monitor the success, eight hydrological monitoring plots were established to check water levels in the area.
17.3 Situation After the Restoration One year after the construction of the dams, the water level in the area had gone up, flooding the edges of the mire. Some of the paths that local people used were also flooded, and this certainly was not appreciated. So, already by September 2002, 4 out of 17 dams had been destroyed by locals. To resolve the arising conflict, a special action of awareness training for people from villages surrounding Yelnia IBA was implemented by APB specialists. Posters and leaflets advertising the natural values of the mire and the urgent need for dam construction in order to save it were distributed among local people and placed in public places (shops, hospitals, schools). Seminars and public meetings were held with local citizens and representatives of local authorities with the participation of journalists from major newspapers. The action was widely covered in the central and regional media.
17 Local and Global Impacts of Mire Drainage
363
A volunteer warden of Yelnia IBA now looks after the condition of the dams and does small-scale maintenance. According to information from this IBA warden, a high flood in spring 2004 damaged the top parts of several dams, but these have now already been repaired. The warden has also informed of another factor that appeared in 2004. Beavers that inhabit canals and ditches have started to use the dams as a base for their own constructions, increasing the water level even more. Sometimes they even damage the artificial dams, although in this case they immediately build their own dams nearby. It is expected that the mire vegetation on areas with a disturbed hydrological regime will now slowly restore, since full restoration of vegetation, including dwarf birch Betula nana and cloudberry Rubus chamaemorus, can occur even in seriously affected areas (Anonymous 2000). Restoration of the mire’s hydrological regime and vegetation structure will in turn help conserve the most important breeding sites of rare bird species (golden plover, black tailed godwit, and whimbrel) and the value of the mire for migrating cranes and geese will increase too. It is difficult to estimate how much time is needed for a full restoration of the mire – correcting errors is always more difficult than making them – but the first step towards improving the situation has been made just in time, when it was still possible to save the situation. An assessment of the ecological situation of 2005–2006 led to the conclusion that additional activities are needed to restore an optimal hydrological regime for the mire; in particular, construction of cascade dams on the canals within the mire will be necessary.
17.4 Wider Impacts of the Yelnia Restoration Project When in 2002, peat and forest fires in Belarus turned into a national disaster, national authorities, especially those dealing with emergency planning and nature protection, turned to the project team for support in a wide-scale replication of the project in other fire-prone areas of the country – where the hydrological regime was similarly disturbed by humans in the past. Building on the outcomes of the project, the Ministry of Natural Resources and Environmental Protection, the Ministry of Emergencies, and the Forestry Committee of the Council of Ministers of Belarus jointly decided on the elaboration of a National Programme to Prevent Fires and Establish Sustainable Use of Degraded Peatlands, through rehabilitation of the hydrological regime. Adoption of the programme will confirm the Government’s commitment to restore what was unwisely done to Belarusian wetlands in the Soviet past. The National Programme, which will build on government funding, is expected to result in a wide-scale rehabilitation of degraded peatlands – reversing the trend of biodiversity loss, greenhouse gas emissions, land degradation and fires. UNDP in Belarus, working through the GEF and in cooperation with bilateral donors and international organisations, such as RSPB (UK), is about to start a recently-approved GEF
364
A. Kozulin et al.
preparatory grant for a project that would further strengthen the Government’s commitments described above. The commitments of these national authorities, NGOs, and international organisations are expected to result in: • A detailed survey of all degraded peatlands to enable decision-making as to their future use and rehabilitation strategy, focusing on the most fire-prone areas. • A rehabilitation of degraded peatlands abandoned for economic reasons, through re-wetting. • A GEF-UNDP project on the restoration of degraded peatlands started in 2006. Presently engineering projects for restoration of the hydrological regime of 4 mires are being elaborated within the project. In total 17 degraded peatlands with a total area of around 42 000 ha are planned to be restored within the framework of the project. As a preparation to the inter-ministerial meeting, in December 2002, the Forestry Committee of the Council of Ministers of Belarus held a meeting for all forestries of Belarus, in which the means for preventing peat and forest fires were discussed and the Yelnia IBA project presented as a model of optimisation of hydrological conditions in fire-prone areas. As can be seen, the project, although small, has managed to raise a substantial commitment from national and international stakeholders to resolve the issue of degraded peatlands: this will not only help prevent fires, but also improve the biosphere functions of mires, preserve globally-significant biodiversity, avoid greenhouse gas mitigation and land degradation.
17.4.1 Contribution of Natural and Drained Peatlands to CO2 Balance in the Atmosphere Natural peatlands transfer carbon dioxide from the biogenic to the geological cycle, thereby taking this greenhouse gas from the atmosphere and burying it in accumulating layers of peat. In Belarus, 1 ha of a natural peatland can act as a carbon sink, converting CO2 into organic peat material at a rate of 550–1,800 kg of CO2 per year (on average, 710 kg of CO2 for a fen mire and about 1,450 kg of CO2 for a raised bog; Bambalov and Rakovih 2003). The drainage of peatlands inevitably leads to a reversal of this process – drained peatlands are no longer net accumulators of peat but sites of peat mineralisation and considerable emitters of CO2. For drained peatlands, the following measurements of CO2 released to the atmosphere per year were made: 20.9 ± 3.4 t ha−1 for arable agriculture; 12.8 ± 2.3 t ha−1 for grain cultivation; 7.5 ± 1.3 t ha−1 for perennial grasslands; and 4.26 ± 1.1 t ha−1 for meadows (Rakovih and Bambalov 1996). One hectare of a drained but unused peatland emits 21–23 t (fen mire) or 14–16 t (raised bog) of CO2 annually until the residual peat layer is gone (Rakovih and Bambalov 1996). Overall CO2 sequestration by the intact ecosystem of Yelnia bog averages out at 26,000 t per year (17,984 ha of natural peatland ×1,450 kg = 26,076.8 t per year).
17 Local and Global Impacts of Mire Drainage
365
One hectare of the drained and abandoned Yelnia peatbog – depending on the type of peat – annually emits 14,000–15,500 kg of CO2 (on average 15,000 kg.year−1). Overall CO2 emission by the completely drained and abandoned section of the Yelnia peatbog (around 2,000 ha) therefore averages some 30,000 t of CO2. Thus ignoring some 3,000 ha of partially-affected mire, annual emission of CO2 from 2,000 ha of drained mire exceeds the CO2 sequestration of some 18,000 ha of intact mire (Rakovih and Bambalov 1996). These estimates do not include the CO2 released during summer fires.
17.4.2 Methane Production in Peatlands Methane is formed in peatlands during incomplete or anaerobic decomposition of organic matter. In the aerated part of the acrotelm, CH4 is further oxidised to CO2 (see, for example, Crill et al. 1994). Methane is the second most important greenhouse gas after CO2 and is expected to contribute 18% of total expected global warming over the next 50 years (Joosten and Clarke 2002). According to some scientists (e.g. Odum 1986; Carrels et al. 1975; Scholes et al. 2000), however, CH4 performs a positive function by sustaining the atmosphere’s ozone layer, which blocks deadly UV radiation. Impacted by the sun’s rays, CH4 disintegrates – yielding hydrogen radicals. These radicals interact with oxygen to produce water vapour which is a major component of the radiation balance, whereas hydroxyl ions (OH) form a major component in ozone photochemistry – another basic component of the radiation balance of the atmosphere. Methane production is a most important function performed by peatlands and shallow seas (e.g. Odum 1986; Carrels et al. 1975). In Belarus, no specific assessments of CH4 emissions by peatlands have yet been undertaken. Relevant data published abroad were therefore used to estimate CH4 emissions (Makhov et al. 1994; Glagolev 1999; Glukhova et al. 1999). The high variability in baseline data for CH4 emissions (from 0.05 to 16.61 g m−2 per season) is indicative of an insufficient level of investigation into the subject (Moore et al. 1994). Peatlands may produce a one-off ‘burst’ of emissions of CH4: up to 65% of the overall emission from the entire vegetation period. Therefore, correct annual assessments will require further research into the long-term dynamics of methane emission.
References Anonymous (2000) Restoration of hydrological regime and prevention of fires in hydrological zakaznik “Yelnya”, an IBA and potential Ramsar site. Final project report to RSPB Bambalov NN, Rakovih VA (2003) Quantitative estimation of contribution of natural and drained peatlands into emission and sequestration of greenhouse gasses. Environmental and economic aspects of forest drainage, vol 58. Compilation of scientific papers, Gomel, pp 91–96 Biryukov V, Kozlov VP, Kuzmenko VY (1993) Role of Elnya raised bog (Vitebsk region, Belorussia) as natural reservation of waterfowl and wetland birds. Ring 15(1–2):348–350
366
A. Kozulin et al.
Carrels RM, Mackenzie FT, Hunt C (1975) Chemical cycles and the global environments; assessing human influences, William Kaufmann, Los Altos, CA, 206 p Crill, P. M., Martikainen, P. J., Nykänen, H. and Silvola, J. (1994), Temperature and N fertilization effects on methane oxidation in a drained peatland soil. Soil Biology and Biochemistry 26:1331–1339 Glagolev MV (1999) Mathematic modeling of methane emission by peatlands. Peatlands and wetland forests in sustainable nature use. Conference proceedings. Moscow, pp 175–177 Glukhova TV, Kovalev AG, Smagina MV, Vompersky SE (1999) Estimation of several biotic components of the CO2 cycle of peatlands and forests. Peatlands and wetland forests in sustainable nature use. Conference proceedings. Moscow, pp 182–185 Joosten H., Clarke D. 2002. Wise use of mires and peatlands – background and principles including a framework for decision-making. Saarijarvi, Finland. 304 pp Makhov, G.A., Bazhin, N.M. and Efremova, T.T. 1994. Methane emission from peatlands in between rivers Ob’ and Tom’. Chemistry in the interest of stable development 2:619–622 Moore TR, Heyes A & Roulet NT (1994) Methane emissions from wetlands, southern Hudson. Bay lowland. J. Geophys. Res. 99: 1455–1467 Odum E (1986) Ecology (Translation from English). Mir, Moscow Rakovih VA, Bambalov NN (1996) Functions of depleted peatlands in biosphere, Prirodopo lzovanie. IPUNRE, Minsk, N1, pp 158–163 Scholes MC, Matrai PA, Smith KA, Andreae MC, Guenther A (2000) Biosphere-atmosphere interactions. http://medias.obs-mip.fr:8000/igac/html/book/index.html Zemlya Belarusi 2001: Reference Aid / Under the editorship of G.I.Kuznetsov., G.V.Dudko. (2002).UP “Belnitszem”, Minsk. 120 pp
Index
A Abramis brama, 111 Acrotelm, 243–262, 287, 289, 290, 296, 297, 315, 321, 323, 325, 365 Afforestation, 193 Aggradation process, 190 Alisma, 38 Alkalinity, 6, 81, 87, 89, 96, 116 Allochthonous, 62, 162, 164, 165 Alum. See Aluminium sulphate Aluminium sulphate bioaccumulation, 87 effect on pH, 86, 88, 89 toxicity of, 87 Amino acids, 298, 299 Ammonia, release of, 42 Anabaena, 115 Ancylus, 163 Anodonta, 109 Aphanizomenon, 42, 115 Aphanizomenon flos-aquae, 115 Austria Alte Donau, 83 Danube National Park, 218–219 Danube River, 195–197 Rothwald virgin forest, 4 Autochthonus, 162, 164 B Base cations, loss of, 12, 15 Bed load, 13, 149, 194, 195, 227 Belarus, Yelnia mire, 355–365 Betula, 29, 271, 273, 275, 276, 282, 292, 293, 298–301, 311, 335, 358, 363 Betula nana, 358, 363 Betula pubescens, 271, 273, 275, 276, 282, 300, 335
Bicarbonate, 306, 307 Bio-corridor, 187, 199, 204 Biofilm, 153, 158, 161–162, 167–168 Biological indication, 67 Biological retention, 168 Biomanipulation, 47, 51 Bird droppings, 299 Birds management for, 34 of mires, 358, 361, 363 Blue-green algae, 39, 42, 43, 47, 48, 91, 107 Bog cut-over, 265–283, 286–289, 291, 300, 306–310, 323 patterned raised, 333 raised, 30, 243–262, 285–325, 332–335, 337, 364 valley raised, 333 Bog pool, 294, 334 Bottom-up control, 49 Brachyptera risi, 159 Buffered groundwater, 301, 306, 324 Buffer-strip, 176, 226, 228–233, 239, 240 Building-blocks, 227, 229, 233, 235, 238 Bulgaria, Danube islands, 204, 207, 218 Bulk density, 272, 273, 303, 304, 307, 308, 352 Bund, 286, 301, 316 Buoyancy, 138, 290, 295, 302–310 C Calcareous groundwater, 306 Calcium, 12, 79, 81, 83, 86, 89, 130, 143, 306, 307, 310 Callitriche, 158, 161 Calluna vulgaris, 249, 273, 275, 276, 317 Canopy density, 300
367
368 Carbon, 31, 39, 41, 42, 79, 156, 163, 169, 286, 290, 293, 295–297, 303, 306, 325, 336, 345, 348, 356, 364 Carbon dioxide release of, 177, 297, 356, 364 shortage of, 39, 41 Carex, 29, 131, 134, 135, 138, 141, 194, 248, 272, 334, 338, 339, 342, 343 Carex acuta, 131 Carex fen peat, 272 Carex limosa, 334, 338, 339, 342, 343 Catotelm, 290, 309, 321, 323 Ceratophyllum, 124, 128, 129, 214 Channel stability, 227 Chara, 39, 124 Chlorophyll a concentration, 47, 48, 54, 97, 98, 115 Cicuta, 38 Cladocera, 31, 42, 51–53, 59, 64, 66, 67, 111 Cladophora, 42 Climate change, 21, 23, 198, 356 Climate stabilisation, 187 Coarse detritus, accumulation of, 29, 73, 134 Colonisation, 126, 162, 269, 290, 309, 313, 314, 321, 325 Compartmentalisation, 316, 320, 324 Conductivity, 5–7, 11, 19, 214, 245, 251, 255, 260, 288–290, 303, 336, 342–344, 346, 348 Copepoda, 51, 53, 55, 57–59, 64, 67 Cryoturbation, 26 Ctenopharyngodon idella, 124 Cyanobacteria, water blooms, 50, 60, 114 Cyprinus carpio, 62, 119 Czech Republic Hubenov reservoir, 56 Římov reservoir, 55–59 Slapy reservoir, 52, 53, 55, 57–59 Šumava NP, 333, 336, 350 Vajgar Lake (fish pond), 46, 113–121 D Dam, 114, 119, 153, 170, 171, 187, 190, 197, 213–215, 245, 262, 287–289, 292, 316, 337, 349, 360–362 Damping, 20–22 Daylightening, 228, 237, 240 Debris, 5, 7, 8, 20, 51, 153, 157, 162, 165–167, 227, 230, 234 Decomposition, 5, 8, 12, 16, 41, 78, 109, 128, 160, 162, 164, 177, 231, 271, 290, 295, 296, 303–306, 310, 365 Deep inundation, 287, 293, 307, 308, 310
Index Deglaciation, 26 Denitrification, 72, 78–81, 168, 169, 230, 237–239 Denmark Brede Å, 174, 177 Gels Å, 174 Gjessø Bæk, 149 Spørring Å, 150 Varde Å, 175 Desertification, 198 Desiccation, 300, 319 Desulphurication, 79–81, 98 Diaspores, 31, 186, 212, 321, 323 Dissipative-ecological-unit, 8–11 Dissipative structures, 3, 13 Dissolved organic carbon, 336, 345, 348 Dissolved organic matter, 163–164 Drainage. See also Hydrology contribution to climate change, 356 effects of, 227, 234, 243–262, 282, 292, 297, 359, 362 inefficency of, 356 Drainage ditches, blocking of, 286, 290, 315, 336, 352 Drained mires inefficient use, 356 risk of fires, 356 source of carbon dioxide, 356, 364 Dynamic equilibrium, 49, 178, 179 E Ecosystem engineer, 325 Ecosystem services, 232, 234, 237, 239, 240 Elitoral, 38–39 Elodea, 38, 98, 124, 158, 161 Elodea canadensis, 98, 158 Emergent macrophytes, 29, 30, 125–139 Energy dissipation, 8, 13, 15, 20 Energy transfer, 46, 61 Ephemeral therophytes, 193 Epilimnion, 93 Equisetum fluviatile, 107–108 Erica tetralix, 301, 317 Eriophorum angustifolium, 273, 275, 292, 316, 317 Eriophorum vaginatum, 270, 273, 276, 281, 316, 319, 325, 343 Erosion, 13, 21, 26, 27, 65, 94, 149, 155, 166, 176, 214, 227, 234, 236, 240, 356 Erosive zone, 33 Esox lucius, 111 Estonia, Viru bog, 267, 269, 272, 277, 279
Index Eulittoral, 38 European strategy on flooding, 173 Eutrophic, 27, 40, 41, 77–83, 85–99, 105, 114, 121, 214, 216 Eutrophication, 2, 39, 41, 77, 78, 124, 126, 169, 198, 214, 238, 300, 310, 312, 314 Evapotranspiration, maximising of, 16 F Feedback control, 5 Fish feeding activity, 46, 49 feeding pressure, 42 kills, 42, 74, 107 ladder, 171 management, 41, 46, 47, 62, 64, 67, 111, 114, 235 migration, 60, 170, 173, 197, 203 predation, 46, 50–53, 66 stock, 39, 41, 42, 46, 49–52, 54–66, 114, 119, 197 Floating peat, 290, 302, 312 Flooding, 131, 146, 148, 151–153, 165, 166, 169, 171–173, 177–178, 186, 188, 190, 193, 194, 197, 199– 202, 204, 211–213, 216, 219, 227, 301, 319, 323, 324, 361, 362 Flooding frequency, 171, 186, 188, 212 Flood peak, 194, 200 Floodplain connection, 190, 200 functions, 145–180, 186, 187, 193, 197, 199–200, 207, 213, 219, 220, 234 loss, 190, 193, 194, 197–199, 214 restoration, 2, 146, 170, 185–221, 240 vegetation, 147, 153, 156, 161, 164, 175, 186, 194, 199, 200, 211, 226, 238 Flood retention, 186, 190, 200, 201, 203, 207 Flow regime, 170–174, 178, 179, 236 Food chain, 46, 61, 98, 161–162, 164 Food web, management of, 47, 74 Form factor, 13 G Gammarus pulex, 163 Geomorphological processes, 147 Germany Gross-Glienicker Lake, 83, 90, 93 Kühkopf-Knoblochsaue Nature Protection Reserve, 208–213 Rastatt Nature Reserve, 190, 213–218
369 Rhine River, 201, 203, 213, 214, 216 Stör River catchment, 11–15, 20, 23 Tegeler Lake, 93 Glaciation, 26, 90, 248 Gravel extraction, 190, 213–215 Grazing, 21, 37, 129, 130, 219 Green Danube Programme, 204, 205 Groenlandia densa, 214 Groundwater recharge, 79, 99, 186, 228, 232, 239 supply, 74, 79, 95, 142, 187, 199, 214 Growth limitation, 21, 293, 294, 297, 319, 324 Gyttja, 26, 30, 104, 108, 137, 143 H Habitat, 9, 11, 39, 51, 66, 114, 131, 141, 146, 147, 151–160, 165, 170, 171, 173–176, 178, 179, 186–188, 190, 193–195, 199, 203, 211–214, 216, 218–220, 227, 230, 231, 236, 239, 323, 334–336, 338, 344, 345 Hardwood, 193, 209, 211, 216, 218 Headwaters, 148, 149, 151, 153, 155, 160, 161, 163–165, 167, 168, 172, 178, 227, 228, 231, 239, 332 Hippophae rhamnoides, 193 Hippuris vulgaris, 214, 216 Hollow, 248, 249, 254, 255, 258, 260, 277, 290, 301, 313, 318, 319, 322, 323, 334–336, 338, 339, 342 Holocene, 26, 27, 33 Horseshoe, 233–234, 236 Hubbard brook, 165, 172 Humic acids, 294 Humification, 255, 262, 305 Hummock, 248, 249, 254, 255, 258, 260, 277, 287, 290, 297, 301, 310, 312, 313, 316, 318, 319, 321, 323–325, 334, 336, 338, 341, 342 Hydraulic conductivity, 245, 251, 255, 260, 288–290 Hydraulic gradient, 250 Hydraulic radius, 13 Hydrochemistry, 270, 278–281, 336, 342–347 Hydrogen sulphide, 79–81, 88, 92 Hydrograph, 13–16, 18, 208, 209, 227–229 Hydrological regime, 15, 186, 190, 198, 199, 204, 207, 218–220, 357, 361, 363, 364 Hydrology aquifer, 91, 251 management, 150, 292
370 Hydropsyche, 162, 163 Hypolimnetic aeration, 94, 98 Hypolimnion, 92–94, 294 Hyporheic zone, 151, 226–228, 236, 238 I Ice-pressing, 33, 34 Intact mires, 339, 365 Integrated Rhine Programme, 200–201, 203, 213, 220 Interception, 228, 233, 234, 299, 300 Internal nutrient supply, 27 Inundation, 65, 169, 190, 195, 200, 249, 275–277, 281, 282, 287, 288, 293, 296, 302, 304, 307–310, 312, 318–321, 323, 324 Invertebrates, 49, 61, 68, 147, 151, 153, 156–159, 161–165, 167, 170, 173, 177, 211, 212, 220, 227, 235, 236, 299–231 Ireland Clara Bog, 243, 244, 248, 317, 321, 322 Irish Midlands, 243, 248 Midlandian, 248 Raheenmore Bog, 243, 244, 248 Iron chloride, 81, 83, 86, 88–90, 94, 96, 97 hydroxy-oxide, 94–96 sulphide, 79–81, 88, 92, 98, 312 treatment, 78, 80, 81, 92, 94–98 Isoëtes, 38, 40 J Juncus effusus, 312, 314, 315, 320 K Key species, 175, 267, 323, 325 K-strategy, 3 L Lagg, 29, 30, 243, 266 Lake ageing See also Wetland ageing, Landscape ageing 25, 27, 29, 102, 129, 133–134, 141, 143 Lake eutrophication, 2, 77, 78, 124, 126, 198 Lake evolution, 18, 25–34, 142 Lake marl, 26, 131, 137, 143, 248 Lake oligotrophication, 98 Landscape ageing, 17
Index Landscape overheating, 23 Landscape sustainability, 3, 15–19 Lemna, 38, 129 Light availability, 164, 295, 296 limitation, 293, 294, 324 Lignin, 304–308 Litter, 5, 80, 158, 161, 162, 164–167, 172, 228, 229, 231, 262, 266, 305, 313 Littoral, 27, 31, 33, 34, 37–39, 92, 96–98, 106, 113, 126, 129, 130, 139, 142 Littorella, 38 Lobelia, 33, 39, 40 Loss of retention areas, 194 M Macrophytes, 29, 30, 34, 37–43, 83, 92, 97–99, 106, 123–144, 153, 157, 158, 160–162, 164, 165, 167, 168, 179, 194, 213, 214, 220, 227, 230–232 Magnesium, 6, 115, 125, 279, 280, 336, 342 Management measures, 64, 217, 325 Matter cycles, 5–7, 9, 12, 16–21, 162 flow, 4, 12–13, 20, 21, 77, 166 losses, 7, 8, 11–13, 15–17, 19–22 recycling, 2, 8, 20 Meadow lakes, 177, 178 Mesotrophic, 41, 114, 214 Methane, production of, 79–81, 134, 141, 302–304, 306–309, 313, 323, 324, 365 Methanogenic bacteria, 302, 303, 306, 307 Microbial activity, 7 Microcystis, 42, 107, 109, 115, 120 Micro-landscape, 262 Microtopography, 248–249, 334, 335 Mineralisation, 12, 27, 29, 78, 79, 93, 94, 124, 238, 303, 356, 364 Minerogenic bottom, 26, 27, 29, 30 Minerotrophic, 248, 333, 337 Mire restoration, methods of, 332 Molinia caerulea, 292, 293, 298, 300, 311, 312, 314–317, 320 Monitoring, 11–12, 18–20, 22, 51, 67, 72, 74, 111–113, 203, 209, 210, 217, 219, 220, 270–271, 332, 335–338, 349–352, 362 Monitoring programme, 11–12, 20, 22, 72, 111–113, 219, 332, 335–336, 349
Index Moraine, 26, 33 Morphodynamical processes, 186 Mosses, 38, 168, 289, 290, 294, 296, 298, 299, 306, 310, 312, 316, 318–321, 342, 358 Myricaria germanica, 193 Myriophyllum, 38, 98, 124, 128 Myriophyllum spicatum, 98 N Nasturtium, 158 Netherlands Bargerveen bog, 287, 291–300, 317 Fochteloërveen bog, 287, 291, 314–318 Haaksbergerveen bog, 287, 291, 300–302, 306 Mariapeel bog, 287, 291, 295, 310–314, 317 New-grown peat, 272, 273 Nitrate sink, 176 Nitrogen deposition, 290, 292, 297–300, 318, 325 filter, 298 fixation, 231 Nitrophilous, 298, 314 Northern European lakes, 26–34 Nuphar, 106, 107 Nurse crop, 321, 325 Nutrient(s) cycling, 3, 27, 92, 125–126, 232, 234, 237, 240 external loading, 27, 71, 73, 74, 78, 86, 94, 113, 114, 121 internal loading, 41, 71, 74, 75, 86, 121 leaching of, 21, 26, 27, 91, 99 recycling, 2, 3, 16, 20, 186, 199, 207 retention, 12, 18, 21, 79, 86, 96, 165–168, 186, 187, 207, 226, 227, 234, 238 sink, 74, 75, 314 source, 74, 86, 91, 121, 232, 282 Nutrients-phytoplankton interactions, 47–49 Nymhaea, 38, 106 O Ochre, 177 Oligotrophic, 26, 27, 37–41, 98, 107, 108, 143, 315 Ombrotrophic raised bogs, 332, 334, 335, 337 Open systems, 3 Order from disorder, 3 Organic matter dissolved, 16, 20, 151, 162–164, 167
371 oxidation of, 16, 78, 79 particulate, 151, 162–163, 167, 236 Organogenic sediment, 26, 27, 29, 30 Oscillatoria, 42 Oxygen availability of, 159–161 concentration, 39–41, 43, 50, 60, 63, 78, 80, 94, 115, 159, 160 fluctuations of, 114, 160 P Palaeochannels, 151 Palaeolimnology, 26, 29–32, 72, 73, 108 Peak flow, 228, 232, 234, 236, 238, 239 Peat chemistry, 278–283 decomposition, 271 humification, 305 stratigraphy, 272 thickness, 272 Peat pits, hydrochemical stratification, 278, 280 Perca fluviatilis, 111 Periphyton, 15, 29, 39, 41, 138, 161, 164, 165, 168 pH, fluctuations of, 114 Phosphate, 345, 346 Phosphorous budget, 115, 119 concentration of, 48–50, 54, 55, 62, 91, 93, 97, 108 dissolved organic, 87 particulate, 87, 177 precipitation of, 86, 94, 108 Phosphorus binding, 78, 81, 85–87, 92, 94, 95, 98, 119 Phragmites, 29, 38, 107, 125–128, 131, 133–135, 137–139, 141 Phragmites australis, 125, 126, 131 Phreatic level, 249, 251–254 Phytoplankton biomass, 51, 111–113 Pinus rotundata, 333 Pinus sylvestris, 273, 276, 282 Pinus x pseudopumilio, 334, 335, 343 Plankton, 29, 39, 41, 46, 50–52, 57, 60, 62, 63, 65, 66, 72, 73, 79, 89, 98, 101, 108, 109, 124, 164 Plant form factor, 13 Plaur formation, 106, 108, 124, 134 Plecoptera, 151, 159 Pollen analysis, 31
372 Polutant’s transformation, 8, 30, 58, 59, 164, 166, 168, 175, 186 Polytrichum strictum, 342, 343, 358 Pool, 13, 55, 56, 66, 68, 151, 153, 155, 158, 163, 167, 171, 173, 178, 179, 227, 235, 245, 246, 248, 249, 253, 260, 294, 334 Poplar monocultures, 193, 207, 213, 216, 218 Postglacial development, 26 Potamogeton, 38, 98, 106, 124, 128, 161, 214, 216 Potamogeton berchtoldii, 98, 216 Potamogeton friesii, 214 Potamogeton pectinatus, 214 Potassium, 78 Potential evapotranspiration, 227 Primary production, 8, 41, 46, 59–62, 77, 79, 83, 85, 138, 162 Primary productivity, 21, 26, 29, 73, 74 Profundal, 38 Pyrite, oxidation of, 177 R Redox conditions, 5, 8, 78, 81 Restoration limno-technical plan, 72–74 of the landscape, 1–23 of lowland streams, 240 methods, 73, 74, 127–129, 139, 244–246 objectives, 199 planning, 32 pre-investigation, 92 principles, 71–75 project design, 73–74, 134–142 project evaluation, 74, 97–99, 109–111, 116–119 Retention, of toxic substances, 186 Rewetting, 283, 286, 290–293, 301, 305, 306, 311–313, 315–319, 323, 349, 351, 352 Rheotrophic, 248 Rhizomes, 106, 125, 126, 128, 135, 137, 138 Riccia, 38 Riffle-pool sequence, 153–155, 171, 235–237 Riparian ecotone, 228–230 Riparian habitat, 156–157 Riparian vegetation, 156, 164, 227, 228, 231 Riparian wetland, 226, 228, 230, 233–234, 237–240 River braided zone, 187 cross-section, 13, 188
Index dynamics, 186, 190, 199, 201, 203 embankment, 193 engineering measures, 187, 188 lateral view, 151 longitudinal view, 148–150 meandering zone, 187, 188, 214 navigability, 195 section, 13 temporal view, 151–152 transition zone, 165, 187, 228 vertical view, 151 River-bed dredging, 195 River Continuum Concept, 164, 165 Rock weathering, 12 Romania Calarasi-Raul island, 206, 207 Cama and Dinu islands, 207 Danube Delta Biosphere, 204, 219 ‘Iron Gate,’ 195, 197, 204 Nature Park, 204 Reserve, 204 Root-felt, accumulation of, 134 removal of, 135–136, 138 r-strategy, 3, 22 Rubus chamaemorus, 273, 358, 363 Rutilus rutilus, 111 S Salinisation, 198 Salix, 29, 131, 193, 211 Salix eleagnos, 193 Salmonids, 158, 160, 163, 171 Satellite imagery, 12 Schoenoplectus, 29, 126–128, 134, 135, 138, 139 Schoenoplectus lacustris, 126 Sediment deposition rates, 28 growth rate, 27, 29, 31, 73, 74, 108 heavy metals content, 86, 115, 186 organic matter content, 115–117 organic sediment thickness, 33, 93, 104, 124, 126 phosphorus release, 86, 93, 98, 102 removal, 75, 101–121, 229–230 stratigraphy, 71, 73, 104, 271, 272, 280 transport, 117, 148, 149, 157, 186, 227, 234 treatment, 75, 77–84, 86 Sedimentation, 18, 19, 27, 49, 78, 82, 87, 88, 93, 94, 98, 114, 151, 158, 167, 214, 227, 236, 239 Self-organisation, 3, 4 Sewage diversion, 99, 107, 126, 239
Index Shading, 41, 160, 161, 230–232, 298, 299, 312, 314, 318 Shallow inundation, 275, 318–321, 324 Shallow lakes, 26, 27, 29, 37–43, 81, 86, 94, 102, 106, 123–144, 177 Sialis, 163 Side slope, 234–237 Size-efficiency hypothesis, 45, 51–53 Soak, 248, 249, 282, 285, 318–321, 324 Softwood, 193, 199, 211, 212, 220 Soil degradation, 356 fertility, 4, 8, 9, 15, 16, 18, 20, 22 organic matter, 16 subsidence, 124, 176 Solar energy, dissipation of, 2, 12 Sparganium emersum, 158 Spawning grounds, 157, 174, 198, 216 Sphagnum, 21, 30, 244, 245, 249, 262, 269, 270, 272–277, 280–282, 286, 287, 289, 290, 292–301, 305, 306, 309–313, 315–317, 319, 322–325, 337, 338, 342, 343, 356, 358 Sphagnum cuspidatum, 249, 273, 290, 292–295, 310, 312, 313, 315, 317, 322, 342, 343 Sphagnum fallax, 290, 319, 321, 343 Sphagnum magellanicum, 249, 285, 290, 296, 297, 310, 322, 323 Sphagnum papillosum, 285, 290, 317, 319, 322, 323 Sphagnum peat, 30, 245, 272, 282, 306, 312, 323 Sphagnum rubellum, 285, 317, 323 Sphagnum, submerged, 293–296, 324 Spirodela, 38, 312, 313 Stephanodiscus hantzschi, 114 Stream classification, 149, 150 Stream habitat, 153–155, 173, 178, 227 Stream hydrograph, 221–229 Sturgeon, 197 Sublittoral, 38 Submersed plants, 124, 127, 128, 157, 161 Subsidence, 124, 244, 248, 252, 258, 287, 301, 351 Subsurface flow, 166, 228–230 Sulphate, 80, 86–89, 92, 108, 307, 308, 344–346, 352 Sulphate reduction, 92, 352 Surface run-off, 176, 228, 230, 249, 342, 352 Surface slope, 250–252, 255, 256, 258, 260, 261, 287, 288, 301
373 Suspended matter, 15, 108, 195, 214 Sustainability (of catchment), 3, 4, 8, 15–18, 20, 22, 23 Sustainable land use/management, criteria of, 16, 18, 20 Sustainable restoration, 1–23, 75, 133, 141 Sweden Björnmossen bog, 279 Hornborga lake, 123–144 Hultamossen bog, 267, 277–279 Läsarmossen bog, 267, 268, 270, 272, 279 Lillesjön lake, 79, 81, 82 Trummen Lake, 28, 32, 101–121 T Target water table concept, 336, 349 Temperature attenuation, 12 See also Damping Temperature distribution, 12, 19, 20 Terrestrialisation, 29, 102, 248, 338, 351 Thermocline, 92 Top-down control, 50 Transmissivity, 250–258, 260–262 Trapa natans, 108, 215 Trichophorum caespitosum, 334–336, 339, 340, 342, 343 Typha, 29, 38, 107–108 U Upstream-downstream continuity, 173 V Vaccinium oxycoccus, 246 Vaccinium uliginosum, 273, 334–336, 338, 343 Vascular plants, 294, 298, 299, 313, 314, 321, 325, 336 Vegetation, 8, 12, 13, 15, 16, 18–23, 26, 27, 29, 31, 33, 34, 39–42, 72, 73, 75, 79, 87, 92, 97, 99, 102, 106–108, 124–134, 137–139, 142, 143, 147, 153, 156, 161, 164, 166, 167, 172, 174–176, 186, 194, 199–212, 214, 219, 220, 227–234, 238, 245, 248–249, 267, 272, 274, 277, 280, 282, 286–290, 292, 297–301, 307, 309, 312–314, 316, 317, 320, 321, 323, 325, 332, 334–336, 338, 341, 356, 362, 363, 365 Vegetation succession, 29 Virgin forest, 4–8, 15, 20
374 W Water chemistry, 43, 279, 280, 286 Water conservation, 287 Water cycle long, 20 short, 5, 19, 186 Water cycling, 9, 232, 234, 237, 240 Water flow dynamics, disturbance of, 13–15 Waterfowl, 72, 73, 109, 111, 124, 127, 129, 131, 138, 141, 293 Water level lowering, 31, 129, 133, 142 Water level raising, 131, 134, 135, 138–141, 144, 176–177 Waterlogged spruce forest, 333–336, 339, 340 Waterlogging, 319 Water pollution, 21, 204 Water-retention capacity, 5, 12 Water run off, 5, 21, 103, 104, 108, 109, 228, 343–346, 352
Index Water-saturated condititions, 2, 5, 238 Water table fluctuation of, 186, 276 oscillations, 7, 8, 12 raising of, 131, 134, 138, 139, 340 Water transparency, 41, 42, 91, 97 Weichselian, 90, 248 Weichselian glaciation, 90, 248 Wetland ageing, See also Lake ageing, Landscape ageing, 26, 128 Wet meadows, 169, 176, 177, 211 Z Zannichellia palustris, 214 Zoobenthos biomass, 50 Zooplankton biomass, 50, 57, 65 Zooplankton-phytoplankton ratio, 51, 55, 67 Zooplankton size, 57